Electron-transfer properties of active aldehydes of thiamin coenzyme models, and mechanism of formation of the reactive intermediates

Article abstract of DOI:10.1002/(SICI)1521-3765(19991001)5:10<2810::AID-CHEM2810>3.0.CO;2-F

The active aldehydes 2a-c^- derived from the reaction of 3-benzylthiazolium salts (1a⁺: 3-benzyl-4-methylthiazolium bromide, 1b⁺: 3-benzyl-4,5-dimethylthiazolium bromide, 1c⁺: 3-benzylthiazolium bromide) with o-t

Full text of DOI:10.1002/(SICI)1521-3765(19991001)5:10<2810::AID-CHEM2810>3.0.CO;2-F

FULL PAPER
Electron-Transfer Properties of Active Aldehydes of Thiamin Coenzyme
Models, and Mechanism of Formation of the Reactive Intermediates
Ikuo Nakanishi, Shinobu Itoh, and Shunichi Fukuzumi*[a]
Abstract: The active aldehydes 2a ± c
and leads to formation of the corre-
sponding radical intermediates (2a ± n ),
outer-sphere one-electron oxidant,
[CoII(phen)3]2‡
(phen ˆ 1,10-phenan-
throline), whose one-electron reduction
potential (Eored
0.97 V) is about the
same as the one-electron oxidation po-
tential of 2a (Eoox
0.96 V), occurs
.
derived from the reaction of 3-benzylth-
‡
iazolium salts (1a : 3-benzyl-4-methyl-
which have been characterized by elec-
tron spin resonance (ESR) spectroscopy.
The rapid rates of electron exchange
‡
thiazolium bromide, 1b : 3-benzyl-4,5-
ˆ
‡
dimethylthiazolium
bromide,
1c :
.
3-benzylthiazolium bromide) with o-tol-
ualdehyde in the presence of DBU (1,8-
diazabicyclo[5.4.0]undec-7-ene) are sta-
ble in deaerated MeCN at 298 K be-
cause of the steric bulkiness of the o-
methyl group, which prohibits benzoin
condensation with a second aldehyde
molecule. The one-electron oxidation of
fourteen different active aldehydes
(2a ± n ) derived from various alde-
hydes occurs at 0.98 to 0.77 V vs.
SCE in deaerated MeCN at 298 or 233 K
between 2a ± c and 2a ± c were deter-
ˆ
mined by the linewidth variations of the
efficiently to yield the corresponding
.
ESR spectra of 2a ± c in the presence of
CoI complex. The observed rate con-
‡
different concentrations of 2a ± c , dem-
onstrating the efficient electron-transfer
properties of the active aldehydes. The
electron transfer from 2a to an
stants for formation of [CoI(phen)3]
agree with those for formation of the
active aldehyde examined independent-
ly. This agreement indicates that rate-
determining formation of 2a , which is a
very strong reductant, precedes the
highly efficient electron transfer from
2a to [CoII(phen)3]2‡.
Keywords: coenzymes ´ cyclic vol-
tammetry
´ ESR spectroscopy ´
radicals ´ thiamin
enzymatic reactions.[7±17] However, the generated active
aldehyde readily undergoes acyloin-type condensation with
a second pyruvate or aldehyde molecule in the absence of the
oxidizing agents.[18, 19] This instability of the active aldehydes
has so far precluded the direct detection of the intermediates
or determination of fundamental redox properties of the
active aldehydes such as the one-electron redox potentials and
the intrinsic barrier for the electron-transfer reactions.[20]
We report herein the direct detection of radical intermedi-
Introduction
Thiamin diphosphate (ThDP) is the coenzyme for a number
of important biochemical reactions, catalysing the decarbox-
ylation of a-keto acids and the transfer of acyl groups.[1±3] The
conjugate base of 2-(a-hydroxyethyl)ThDP, which is an acyl
carbanion equivalent, called an active aldehyde, is known to
play an essential role in the catalysis of ThDP-dependent
enzymes. The active aldehyde has ability to mediate an
efficient electron transfer to various physiological electron
acceptors, such as, for example, lipoamide in pyruvate
dehydrogenase multienzyme complex,[4] flavin adenine dinu-
cleotide (FAD) in pyruvate oxidase,[5] and the Fe4S4 cluster in
pyruvate ± ferredoxin oxidoreductase.[6] In this context, chem-
ical models of thiamin coenzyme have been studied exten-
sively by means of simple thiazolium ions, providing valuable
information about the elementary steps of ThDP-dependent
.
ates (2a ± l ) of active aldehydes (2a ± l ) derived from three
‡
different 3-benzylthiazolium salts (1a ± c ) and eight different
aldehydes in the presence of 1,8-diazabicyclo[5.4.0]undec-7-
ene (DBU) in acetonitrile (MeCN) by using electron spin
resonance (ESR), and the first and second one-electron
oxidation potentials of fourteen different active aldehydes
(2a ± n : Scheme 1).[21] This has been made possible by using
active aldehydes stabilized by steric bulk, or by lowering the
reaction temperature. The rapid electron exchange rates
.
between 2a ± c and 2a ± c can be determined by the line-
[a] Prof. Dr. S. Fukuzumi, I. Nakanishi, Assoc. Prof. Dr. S. Itoh
Department of Material and Life Science
Graduate School of Engineering, Osaka University
Suita, Osaka 565-0871 (Japan)
width variations of the ESR spectra depending on different
concentrations of the active aldehydes. This study also reports
the kinetic investigation of an efficient electron transfer from
an active aldehyde to an outer-sphere one-electron oxidant,
[CoII(phen)3]2‡ (phen ˆ 1,10-phenanthroline), as well as the
Fax : (‡ 81)6-6879-7370
E-mail: fukuzumi@chem.eng.osaka-u.ac.jp
2810
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2810 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 10

2810±2818
‡
zylthiazolium salts (1a ± c ) with o-tolualdehyde (o-MeC6H4-
CHO) in the presence of DBU are stable enough for it to be
possible to determine redox potentials by cyclic voltammetry
(CV) measurements in MeCN at 298 K. This is because the
steric bulk of the o-methyl group prevents the benzoin
condensation with a second aldehyde molecule. Thus, a cyclic
voltammogram recorded for 2a prepared in situ by addition
of neat DBU (1.0 Â 10 2 m) to an MeCN solution containing
‡
1a (5.0 Â 10 3 m), o-MeC6H4CHO (0.25m), and tetra-n-
butylammonium perchlorate (TBAP) (0.10m) as a supporting
electrolyte at 298 K exhibits two reversible one-electron
redox peaks at Eooxꢀ1†
ˆ
0.96 and Eooxꢀ2†
ˆ
0.52 V vs. SCE
(Figure 1).[22] The reversible CV waves can be observed only
Scheme 1. Formation of active aldehyde from 3-benzylthiazolium salts
‡
1a ± c and aldehydes (RCHO) in the presence of DBU.
‡
Figure 1. Cyclic voltammograms of 2a , derived from 1a (5.0 Â 10 3 m), o-
formation of active aldehydes. The highly negative oxidation
potentials of active aldehydes and the spin distribution of the
intermediate radicals determined for the first time in this
study provide comprehensive and corroborative understand-
ing of the ThDP-dependent electron-transport systems as well
as valuable mechanistic insight into the enzymatic reactions.
tolualdehyde (0.25m), and DBU (1.0 Â 10 2 m), at 298 K (solid curve), and
of 2 f , derived from 1a (5.0 Â 10 3 m), benzaldehyde (0.25m), and DBU
‡
(1.0 Â 10 2 m), at 233 K (dashed curve) in deaerated MeCN containing
1
0.10m TBAP. Sweep rate 0.10 Vs
.
‡
in the presence of all the components together, that is, 1a , o-
MeC6H4CHO, and DBU. This indicates that it is not the
parent compound but the active aldehyde 2a that undergoes
the electrochemical redox reactions. The reversible one-
electron redox waves were also observed at 298 K for active
Results and Discussion
One-electron oxidation potentials of active aldehydes: The
active aldehydes 2a ± c derived from the series of 3-ben-
‡
aldehydes 2d and 2e , derived from 1a with 1-naphthalde-
hyde (R ˆ Naph) and 9-anthraldehyde (R ˆ An), respective-
ly; the Eooxꢀ1† and Eooxꢀ2† values are listed in Table 1. However,
no CV peaks can be observed for active aldehydes derived
from other aldehydes, which are unstable at 298 K.
Abstract in Japanese:
The benzoin condensation which causes disappearance of
the active aldehyde may be retarded by lowering the temper-
ature to 233 K. In fact, the reversible CV waves for active
aldehydes derived from other aldehydes can be observed at
233 K. A typical example of the CV wave of 2 f at 233 K
‡
derived from 1a and benzaldehyde (PhCHO) in the presence
of DBU is shown in Figure 1. The first oxidation potentials
Eooxꢀ1† and the second oxidation potentials Eooxꢀ2† of various
active aldehydes (2a ± n ) thus obtained are summarized in
Table 1.
As shown in Table 1, the parent aldehydes have little effect
on the Eooxꢀ1† values. The Eooxꢀ1† values of the active aldehydes
are nearly the same ( 0.93 to 0.98 V) irrespective of the
parent aldehyde, except for those from PhCHO (2 f ± h ) and
p-cyanobenzaldehyde (p-CNC6H4CHO) (2l ), which are
somewhat less negative than the others.
Chem. Eur. J. 1999, 5, No. 10
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2811 $ 17.50+.50/0
2811

S. Fukuzumi et al.
FULL PAPER
Table 1. Potentials (vs. SCE) of active aldehydes (2 )[a] in deaerated MeCN determined by cyclic
voltammetry, and adiabatic ionization potentials (Ip) and negative change densities (1) on the
oxygen and C2 carbon atoms calculated by the PM3 methods.
value of 2.0054 was detected successfully at
298 K as shown in Figure 2a. The signal
disappeared when the solution was electro-
lysed at 1.20 and 0.30 V vs. SCE, values
which correspond to the one-electron oxida-
tion and reduction potentials, respectively, of
2a . The radical species 2d and 2e , derived
from 1-naphthaldehyde and 9-anthralde-
hyde, were also observed at 298 K in the
same manner (Figure 2e and 2f, respective-
ly). The radicals derived from unstable active
aldehydes at 298 K could also be detected
after the controlled-potential electrolysis of
the corresponding active aldehydes by keep-
ing the reaction system at a low temperature
(233 K) as in the case of CV measurements.
The ESR spectra thus obtained are also
shown in Figure 2.
.
.
.
R1
R2
H
R
Eooxꢀ1† [V][b] Eooxꢀ2† [V][b] Ip [eV][c] 1O
1C2
0.65
2a
2b
2c
2d
2e
2 f
2g
2h
2i
2j
2k
2l
2m
2n
Me
o-MeC6H4
0.96
0.52
1.78
1.68
1.70
1.72
1.87
1.72
1.73
1.72
1.57
1.56
1.57
1.96
1.83
1.72
0.56
0.54
0.56
0.54
0.53
0.54
0.54
0.55
0.56
0.56
0.56
0.54
0.53
0.54
Me Me o-MeC6H4
H
Me
Me
Me
Me Me Ph
H
Me
Me Me Me
H
Me
Me
Me
0.97
0.56
0.69
0.64
0.68
0.68
0.69
0.69
0.65
0.71
0.71
0.70
0.68
0.68
0.69
H
H
H
H
o-MeC6H4
Naph
An
0.95
0.50
0.96
0.53
0.97
0.42
Ph
0.78[d]
0.79[d]
0.77[d]
0.98[d]
0.96[d]
0.93[d]
0.78[d]
0.93[d]
0.96[d]
0.44[d]
0.45[d]
0.42[d]
0.74[d]
0.73[d]
0.73[d]
0.44[d]
0.41[d]
0.64[d]
H
H
Ph
Me
H
H
H
H
Me
p-CNC6H4
2,4-Cl2C6H3
p-MeOC6H4
The observed ESR spectra of active alde-
.
hyde radicals (2a ± l ) can be simulated with
[a] Active aldehydes 2a ± n were prepared by adding neat DBU (1.0 Â 10 2 m) to deaerated
the hyperfine splitting (hfs) values listed in
Table 2, as shown in Figure 2. The hfs values
indicate clearly that the active aldehydes
exist in the anion form 2a ± l where depro-
tonation of the hydroxy group occurs in the
presence of a strong base such as DBU, since
‡
MeCN solution containing 1a ± c (5.0 Â 10 3 m), RCHO (0.25m), and TBAP (0.10m). Working
1
electrode: Pt. Sweep rate: 0.10 V
s
.
[b] Measured at 298 K unless otherwise noted.
.
[c] Calculated from the difference in the heat of formation (DHf) between 2a ± n and 2a ± n .
[d] Measured at 233 K.
The adiabatic ionization potentials Ip and the negative
charge densities 1 on the oxygen and C2 carbon atoms of 2a ±
n
were calculated by using the semiempirical PM3 MO
method (see Experimental Section);[23] these too are listed in
Table 1. Both the Ip and 1 values are also insensitive to the
parent aldehydes, indicating that the HOMO levels and the
solvation energies are about the same regardless of the parent.
This may be the reason why the Eooxꢀ1† values are insensitive to
the parent aldehydes, although the slight difference in the
solvation may be reflected in the somewhat different Eooxꢀ1†
values. The Eooxꢀ1† values in Table 1 support the proposal by
Jordan et al. that the one-electron oxidation potential of the
active aldehyde, which can reduce a flavin analogue, must be
more negative than 0.67 V vs. SCE.[7e]
Scheme 2. Oxidation of active aldehydes 2a ± n by two electron-transfer
steps.
no hyperfine splitting due to the hydroxy proton of the one-
.
electron-oxidized radicals (2a ± l ) is observed. Deuterium
substitution at appropriate known sites may permit exper-
imental verification of the assignment of the observed radical
species, since a single deuteron gives a triplet (instead of
doublet) hyperfine pattern and the deuteron splitting should
decrease by the magnetogyric ratio of proton to deuterium
(0.143).[24, 25] In fact, deuterium substitution of the two hydro-
The second oxidation potentials Eooxꢀ2† of 2a ± n are more
sensitive to the parent aldehydes than are the first oxidation
potentials Eooxꢀ1† (Table 1). The more negative Eooxꢀ2† of 2i ± k ,
derived from acetaldehyde, compared with those for other
active aldehydes (Table 1) may be ascribed to the electron-
releasing effect of methyl group that would stabilize the
.
gen atoms at the benzylic position of 2a resulted in a drastic
change in the splitting pattern from the spectrum in Figure 2a
‡
to that in Figure 2b, where 1a is replaced by 3-([a,a'-
‡
corresponding cations 2i ± k .
2H2]benzyl)-4-methylthiazolium ion. The hfs value of 2.35 G
.
resulting from PhCH2 protons of 2a is decreased by the factor
Detection and characterization of radical intermediates: The
observation of well-defined one-electron redox couples of
active aldehydes (2a ± n ) indicates that radical intermediates
of the magnetogyric ratio of proton to deuterium (0.143) to
0.35 G due to the PhCD2 deuterons of the corresponding
deuterated radical (2a (PhCD2)). The change in the splitting
.
.
(2a ± n ) are formed in the first one-electron oxidation of the
pattern is also observed on deuterium substitution of two
.
.
active aldehydes. Therefore, the ESR spectrum of a radical
hydrogen atoms at the benzylic position of 2i (2i (PhCD2)) as
.
intermediate (2a , Scheme 2) generated by the controlled-
well as that of the three hydrogen atoms of the acetyl moiety
.
.
potential electrolysis of 2a was measured in deaerated
MeCN containing 0.10m TBAP at 298 K (see Experimental
Section). When the solution containing 2a was electrolysed
at 0.70 V vs. SCE, which is between the first and second one-
electron oxidation potentials, a radical species having a g
of 2i (2i (CD3CO)) shown in parts k ± m of Figure 2. The hfs
values of 2.34 and 3.46 G due to PhCH2 and CH3CO protons
.
of 2d are decreased by factors of 0.143 to 0.33 and 0.51 G due
to PhCD2 and CD3CO deuterons, respectively, while the other
hfs values remain identical. The substitution of one hydrogen
2812
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2812 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 10

Active Aldehydes
2810±2818
.
Figure 2. ESR spectra of active aldehyde radicals (2 ) produced by the electrochemical oxidation of 2 in deaerated MeCN containing 0.10m TBAP and the
.
.
.
.
.
.
.
.
.
.
.
.
.
computer simulation spectra: a) 2a , b) 2a (PhCD2), c) 2b , d) 2c , e) 2d , f) 2e , g) 2 f , h) 2 f (PhCD2), i) 2g , j) 2h , k) 2i , l) 2i (PhCD2), m) 2i (CD3CO),
.
.
.
n) 2j , o) 2k , and p) 2l . The hfs values used for the simulation are listed in Table 2.
.
atom with a methyl group at the C4 position of the thiazolium
ring also causes a change in the splitting pattern (Figure 2).
The above results confirm the assignment of the observed
and Oesterhelt.[6b] The radical species 2a ± n loses one more
electron in the second oxidation to form the 2-acylthiazolium
‡
ions 2a ± n as depicted in Scheme 2.
.
ESR signals to 2a ± l in Table 2.
An important point to note from the results in Table 2 is no
or very small hfs values of the aromatic ring of active aldehyde
radicals 2a ± h and 2l derived from aromatic aldehydes. This
indicates no or little delocalization of spin on the aromatic
ring. This is confirmed by the spin densities of 2a and 2 f ,
Thus, the one-electron oxidation of each active aldehyde
leads to formation of the corresponding neutral radical, whose
structure is similar to that suggested for the oxidized active
aldehyde in pyruvate ± ferredoxin oxidoreductase by Kerscher
.
.
.
.
Chem. Eur. J. 1999, 5, No. 10
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2813 $ 17.50+.50/0
2813

S. Fukuzumi et al.
FULL PAPER
Table 2. g Values and hyperfine splitting (hfs) values of active aldehyde radicals.
hfs [G]
aH(C5)
Radical
g
aN(N)
aH(PhCH2)
aH(C4)
aH(CH3CO)
aH(C2')
aH(C4')
.
[a]
[a]
[a]
2a
2.0054
4.71
(4.73)[b]
4.71
4.75
4.89
4.68
4.74
4.53
(4.89)[b]
4.53
4.70
4.66
4.74
4.74
4.74
5.02
4.87
4.48
2.35
2.89
(2.59)[b]
2.89
2.64
2.66
2.95
2.89
3.10
(2.73)[b]
3.10
2.64
2.83
3.12
3.12
3.12
2.85
3.00
2.93
±
(1.55)[b]
0.35[c]
2.07
(0.05)[b]
(<0.02)[b]
(0.00007)[b]
.
[a]
[a]
[a]
[a]
[a]
[a]
[a]
2a (PhCD2)
2.0054
2.0051
2.0053
2.0055
2.0055
2.0057
±
±
±
±
±
±
.
[a]
[a]
[a]
[a]
[a]
2b
2c
2d
2e
.
2.66
2.38
2.35
2.40
(1.90)[b]
0.35[c]
2.20
.
0.42
0.50
0.42
(0.03)[b]
0.42
0.45
0.42
0.64
0.64
0.64
0.50
0.38
0.50
.
±
.
2 f
0.24
(0.12)[b]
0.24
0.24
0.21
±
±
±
±
0.48
(0.14)[b]
±
±
±
3.46
3.46
0.51[c]
3.54
3.45
±
.
2 f (PhCD2)
2.0057
2.0051
2.0057
2.0052
2.0052
2.0052
2.0052
2.0055
2.0055
0.48
0.48
0.43
±
±
±
.
2g
.
2h
2.48
.
2i
2.34
0.33[c]
2.34
2.24
2.65
.
2i (PhCD2)
.
2i (CD3CO)
.
2j
±
.
2k
±
±
.
[a]
[a]
2l
2.30
[a] Too small to be determined. [b] The values in parentheses are those evaluated by the PM3 method. [c] Deuterium splitting value.
which were calculated by the PM3 method.[23] The calculated
hfs values based on the spin densities, which are given in
parentheses in Table 2,[26] agree reasonably well with the
experimental values determined from the ESR spectra. The
Reorganization energies of active aldehyde intermediates:
The linewidth variations of the ESR spectra of 2a ± c , which
.
are stable at 298 K, were used to investigate the electron-
.
transfer exchange reactions between 2a ± c and 2a ± c
.
optimized structure of 2a with the SOMO orbitals is shown in
[Eq. (1)].[27] The active aldehydes 2a ± c were prepared by
.
Figure 3. The benzene ring of 2a derived from o-tolualde-
ˆ
hyde is nearly perpendicular (878) to the plane of the C C
double bond. This may be the reason why no spin is
delocalized to the benzene ring in Figure 3. This is also
consistent with the Eooxꢀ1† values being insensitive to the parent
aldehydes.
‡
the addition of neat DBU to an MeCN solution of 1a ± c and
o-tolualdehyde, keeping the ratio of the components constant
‡
(1a ± c :o-tolualdehyde:DBU ˆ 1:50:2). Then, the solution
was partially electrolysed at 0.70 V vs. SCE, to give the
.
corresponding radical species 2a ± c . The maximum slope
linewidths (DHmsl) were determined from the computer
simulation of the ESR spectra (Figure 4). The DHmsl values
.
of 2a ± c thus determined increase linearly with an increase in
the concentration of 2a ± c as illustrated in Figure 5. The rate
constants (kex) of the self-exchange reactions [Eq. (1)] were
determined by means of Equation (2), where DHmsl and DHmo sl
.
kex ˆ 1.52  107 (DHmsl DHmo sl)/{(1 Pi) [2a ± c ]}
(2)
Figure 3. SOMO orbitals of 2a calculated by the PM3 method.
2814
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2814 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 10

Active Aldehydes
2810±2818
also listed in Table 3. No prominent change in the kex or l
values is observed dependent on the presence of methyl
substituents on the thiazolium rings. The l values are as small
as those of fast electron-transfer exchange systems such as p-
1
benzoquinone/semiquinone radical anion (13.1 kcalmol
in DMF) and naphthalene/naphthalene radical anion
(12.0 kcalmol 1 in DMF).[30]
The reorganization energies for the electron exchange
.
between 2a and 2a were theoretically evaluated with
semiempirical PM3 MO calculations (see Experimental
Section).[23] The PM3-optimized structure of 2a was com-
.
pared with that of the corresponding radical 2a , shown in
Figure 6. Little structural change associated with the electron-
.
transfer oxidation of 2a to 2a is visible in Figure 6, although
.
Figure 4. ESR spectra of 2a in the presence of different concentrations of
2a : a) 1.5 Â 10 3 and b) 3.0 Â 10 2 m in deaerated MeCN containing 0.10m
TBAP at 298 K, and the corresponding computer-simulated spectra from
the hfs values shown in Table 4 and linewidths (DHmsl) of a) 0.50 and
b) 1.55 G.
Figure 6. Optimized structures of 2a and the corresponding radical
intermediate 2a calculated by using the PM3 method.
.
.
ˆ
the C C double bond in 2a (1.50 Š) becomes somewhat
longer than that in 2a (1.39 Š). The difference between DHf
for 2a with the unchanged structure from 2a and DHf for 2a
.
.
with the optimized structure can be regarded as the reorgan-
ization energy of the inner coordination spheres (li) associ-
ated with the structural change upon electron-transfer oxida-
.
tion in the gas phase. Thus, the li values of the 2a ± c /2a ± c
system were calculated; the values are also listed in Table 3.
The small li values are consistent with little structural change
accompanying the electron transfer. The difference between l
and li indicates that solvent reorganization also plays a role in
determining the intrinsic barrier to the electron transfer
.
between 2a ± c and 2a ± c .
.
.
.
~
&
*
Figure 5. Plots of DHmsl of ESR spectra of 2a ( ), 2b ( ), 2c ( ) vs. [2a ±
Electron transfer from active aldehydes to an electron
acceptor: The small reorganization energies and the highly
negative oxidation potentials of the active aldehydes indicate
the high reducing ability to mediate electron transfer to an
electron acceptor. This is confirmed by using tris(1,10-
phenanthroline)cobalt(ii) complex, [CoII(phen)3]2‡ (phen ˆ
1,10-phenanthroline), as an electron acceptor. This electron
acceptor was chosen since the one-electron reduction poten-
c ] in deaerated MeCN containing 0.10m TBAP at 298 K.
are the maximum slope linewidths of the ESR spectra in the
presence and absence of 2a ± c , respectively, and Pi is a
statistical factor which can be taken as nearly zero.[28] The kex
values thus determined are listed in Table 3.
The reorganization energies (l) of the self-exchange
reactions [Eq. (1)] were obtained from the kex values by
tial (Eored
ˆ
0.97 V vs. SCE)[31] is about the same as the one-
1
means of Equation (3) (Z ˆ 1011m 1 s );[29] the l values are
electron oxidation potential of 2a (Eoox
ˆ
0.96 V vs. SCE)
and the reorganization energy of the [CoII(phen)3]2‡/
kex ˆ Zexp( l/4RT)
(3)
‡
[CoI(phen)3] system is small.[31±33]
Table 3. Electron-exchange rate constants (kex) between 2a ± c and 2a ±
c , and experimental (l) and calculated reorganization energies (li) by the
PM3 method.
Upon addition of DBU (0.03m) to a deaerated MeCN
.
‡
solution of 1a (2.0 Â 10 4 m), o-tolualdehyde (0.03m), and
[CoII(phen)3]2‡ (1.0  10 3 m) at 298 K, a new absorption band
8
1
1
1
10 kex [m 1 s
]
l [kcalmol
]
li [kcalmol ]
at 1415 nm due to the corresponding CoI complex,
Active aldehyde
‡
2a
2b
2c
5.6
6.6
4.5
12.4
12.0
12.9
7.3
11.4
9.6
[CoI(phen)3] , appeared rapidly (Figure 7).[34] In the absence
‡
‡
of 1a and/or o-MeC6H4CHO, no formation of [CoI(phen)3]
was observed. The formation of [CoI(phen)3] demonstrates
‡
Chem. Eur. J. 1999, 5, No. 10
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2815 $ 17.50+.50/0
2815

S. Fukuzumi et al.
FULL PAPER
Figure 7. Visible ± NIR spectrum observed after addition of DBU (0.03m)
‡
Figure 8. Plots of kobs vs. [o-MeC6H4CHO] for formation of [CoI(phen)3]
‡
to a deaerated MeCN solution of 1a (2.0 Â 10 4 m), o-tolualdehyde
in the electron-transfer reduction of 2a with [CoII(phen)3]2‡
(
) and for
*
(0.03m), and [CoII(phen)3](PF6)2 (1.0 Â 10 3 m) at 298 K. The peak marked
with an asterisk is due to an artifact in the instrument. Inset: plot of a2/(1
a) vs. ([CoII]0/[2a ]0 a); see text.
‡
*
formation of 2a
(
) in deaerated MeCN at 298 K. [1 ] ˆ 1.0  10 3 m;
[DBU] ˆ 0.01 or 0.05m.
centrations were maintained at more than a tenfold excess of
that the active aldehyde 2a , having a strong reducing ability,
is formed in situ and acts as a good electron donor to reduce
[CoII(phen)3]2‡ [Eq. (4)].
‡
the 1a concentration. The observed pseudo-first-order rate
‡
constants for the formation of [CoI(phen)3] agree with those
for formation of the active aldehyde (2a ) shown in Figure 8.
Such an agreement demonstrates that the highly efficient
electron transfer from 2a to [CoII(phen)3]2‡ occurs following
the rate-determining formation of 2a , which is very strongly
reducing. Since the kobs values for formation of 2a are
constant with respect to changing o-MeC6H4CHO concen-
tration and increase with increasing DBU concentration
(Figure 8), the rate-determining step for formation of 2a
and the electron transfer from 2a to [CoII(phen)3]2‡ may be
K
.
‡
2a ‡ [CoII(phen)3]2‡ „ 2a ‡ [CoI(phen)3]
(4)
The equilibrium constant (K) for the electron transfer
between 2a and the CoII complex [Eq. (5)] is determined as
0.29 from the slope of the plot of a2/(1 a) vs. ([CoII]0/
[2a ]0 a) (in the inset of Figure 7) according to Equa-
tion (5), where a is the yield of the CoI complex which is equal
‡
the deprotonation of 1a by DBU as sketched in Scheme 3.
a2/(1 a) ˆ K([CoII]0/[2a ]0 a)
(5)
‡
The rate constants (k1) for deprotonation of 1a ± c were
determined from the spectral change at 300 nm without
aldehydes. The k1[DBU] values were then compared with the
kobs values for formation of active aldehyde (2a ± c ) (Ta-
ble 4). The kobs values for formation of the active aldehyde
2a ± c at different DBU concentrations are essentially the
‡
to [CoI]/[2a ]0, [CoI] is the concentration of [CoI(phen)3] ,
and [2a ]0 and [CoII]0 are the initial concentrations of
[CoII(phen)3]2‡ and 2a , respectively. From the K value, the
one-electron oxidation potential of 2a is determined as
0.94 V using Equations (6) and (7). This value agrees well
with that obtained from the cyclic voltammogram ( 0.96 V, in
Table 1).
‡
same as the k1[DBU] values for deprotonation of 1a ± c by
DBU in the absence of o-tolualdehyde. This agreement
DGoet
ˆ
RTlnK
(6)
(7)
DGoet ˆ F(Eoox Eroed
)
‡
The rates of formation of [CoI(phen)3] in the electron
transfer from 2a to [CoII(phen)3]2‡ were monitored by means
‡
of the increase of the absorbance due to [CoI(phen)3] at l ˆ
426 nm. The formation rates obeyed pseudo-first-order
kinetics under conditions where the o-MeC6H4CHO and
DBU concentration was maintained at more than a tenfold
‡
excess of the 1a and [CoII(phen)3]2‡ concentrations. The
observed pseudo-first-order rate constant (kobs) for forma-
‡
tion of [CoI(phen)3] is constant with the change in the
o-MeC6H4CHO concentration used in excess (Figure 8).
‡
The rates of formation of [CoI(phen)3] were compared
with those of formation of the active aldehyde intermediates
determined from the increase in absorbance at lmax ˆ 380 nm
due to 2a .[7d±g] The rates obeyed pseudo-first-order kinetics
under conditions where the o-MeC6H4CHO and DBU con-
Scheme 3. Mechanism of electron transfer from active aldehyde 2a to
[CoII(phen)3]2‡
.
2816
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2816 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 10

Active Aldehydes
2810±2818
Table 4. Observed pseudo-first-order rate constants (kobs) for formation of
active aldehydes (2a ± c ), and forward rate constants k1[DBU] for
deprotonation of (1a ± c ) in deaerated MeCN at 298 K.
atmospheric pressure. The rates of formation of thiazolium ylides and
active aldehydes were determined by monitoring the increase in the
intensity of the absorption bands at 300 and 380 nm, due to the ylides and
active aldehydes, respectively, under pseudo-first-order conditions where
the concentrations of DBU and/or o-tolualdehyde were maintained at
more than tenfold excess of the thiazolium salt concentration. Pseudo-first-
order rate constants were determined by a least-squares curve fit carried
out by means of a Macintosh personal computer. The first-order plots of
ln(A1 A) vs. time (A1 and A are the final absorbance and the absorbance
during the reaction, respectively) were linear for three or more half-lives
with the correlation coefficient 1 > 0.999.
‡
1
1
Thiazolium salt
[DBU] [m]
kobs [s
]
k1[DBU] [s ]
‡
1a
0.01
0.05
0.05
0.05
0.41
1.7
1.6
0.41
2.0
2.0
‡
1b
‡
1c
2.8
3.1
between the kobs and k1[DBU] confirms that the active
Reaction of the active aldehyde 2a with [CoII(phen)3](PF6)2 was started by
addition of DBU (13 mL, 8.7 Â 10 5 mol) to a quartz cuvette (10 mm i.d.)
aldehydes are formed by rate-determining deprotonation of
‡
which contained 1a (2.0 Â 10 4 m), o-tolualdehyde (0.03m), and [CoII-
‡
1a ± c by DBU to afford the ylids 1a ± c, followed by the
(phen)3](PF6)2 (1.0 Â 10 3 m) in deaerated MeCN (3.0 mL) at 298 K.
Visible ± NIR spectral change associated with the reduction of [CoII-
(phen)3](PF6)2 was monitored with a Hewlett Packard 8453 diode array
spectrophotometer or a Shimadzu UV-3100PC spectrophotometer. Kinetic
subsequent rapid addition of 1a ± c to the aldehydes and the
further deprotonation by DBU as depicted in Scheme 3 for
‡
the case of 1a .
‡
As demonstrated above, thiazolium salts serve as efficient
redox catalysts by forming the active aldehyde intermediates
derived from various aldehydes in the presence of DBU. The
largely negative one-electron oxidation potentials of the
active aldehydes and the small reorganization energies for
the electron-transfer reactions indicate that the active alde-
hydes have strong electron-donor abilities and that they are
suitable for fast electron-transfer systems where they can act
as efficient electron-transfer catalysts.
measurements for formation of [CoI(phen)3] were carried out by means of
a Union RA-103 stopped-flow spectrophotometer under argon at atmos-
‡
pheric pressure. The rates of formation of [CoI(phen)3] were determined
by an increase in the absorption band intensity at 426 nm due to the CoI
complex.
Cyclic voltammetry: Typically, DBU (7.5 mL, 5.0 Â 10 5 mol) was added to
‡
an electrochemical cell which contained 1a (5.0 Â 10 3 m), o-tolualdehyde
(0.25m), and TBAP (0.10m) in deaerated MeCN (5.0 mL) under argon at
atmospheric pressure. The first and second one-electron redox potentials of
the active aldehyde (2a ) thus generated were determined at 298 K by the
cyclic voltammograms measured under deaerated conditions with a three-
electrode system and a BAS 100B electrochemical analyser. Similar
experimental conditions were employed for the cyclic voltammetry (CV)
measurements of 2b ± e . An MeCN bath containing solid CO2 was used to
keep the reaction temperature at 233 K for the CV measurements of 2 f ±
n . The working and counterelectrodes were platinum, while Ag/AgNO3
(0.01m) was used as the reference electrode. All potentials are reported in
V vs. SCE. The E1/2 value of ferrocene used as a standard is 0.37 V vs. SCE
in MeCN under our solution conditions.[39]
Experimental Section
‡
Materials: Thiazolium salts (1a : 3-benzyl-4-methylthiazolium bromide,
‡
‡
1b : 3-benzyl-4,5-dimethylthiazolium bromide, and 1c : 3-benzylthiazo-
lium bromide) were prepared from the corresponding thiazole (4-
methylthiazole, 4,5-dimethylthiazole, and thiazole, respectively) by reac-
tion with benzyl bromide at 808C for 20 h, and purified by recrystallization
from ethanol or acetone as described in the literature.[18a] The deuterated
compound 3-([a,a'-2H2]benzyl)-4-methylthiazolium bromide was prepared
by the reaction of thiazole and [a,a'-2H2]benzyl bromide, which was
obtained by reaction of [a,a'-2H2]benzyl alcohol with HBr.[35] [a,a'-
2H2]Benzyl alcohol was prepared by reduction of benzoic acid with
LiAlD4, obtained from Aldrich. Tris(1,10-phenanthroline)cobalt(ii) hexa-
fluorophosphate, [CoII(phen)3](PF6)2, was prepared by adding 3 equiv of
1,10-phenanthroline (monohydrate) to cobalt(ii) chloride (hexahydrate) in
ethanol followed by the addition of KPF6 according to literature
procedures.[31, 36, 37] Aldehydes used in this study [o-tolualdehyde (o-
MeC6H4CHO), 1-naphthaldehyde (NaphCHO), 9-anthraldehyde (An-
CHO), benzaldehyde (PhCHO), acetaldehyde (MeCHO), p-cyanobenz-
aldehyde (p-CNC6H4CHO), 2,4-dichlorobenzaldehyde (2,4-Cl2C6H3CHO),
ESR measurements: Typically, DBU (7.5 mL, 5.0 Â 10 5 mol) was added to
‡
an electrolysis cell which contained 1a (5.0 Â 10 3 m), o-tolualdehyde
(0.25m), and TBAP (0.10m) in deaerated MeCN (5.0 mL) under argon at
.
atmospheric pressure. The active aldehyde radical 2a was generated
electrochemically at an applied potential of 0.7 V vs. SCE. The solution
containing the radical was transferred to an ESR tube by means of a syringe
which had earlier been purged with a stream of argon. The ESR spectra of
.
2a and other active aldehyde radicals generated electrochemically were
measured at 298 K or 233 K with a JEOL X-band spectrometer (JES-
RE1XE). It was confirmed that the ESR signals due to the active aldehyde
radicals disappeared when the latter were reduced or oxidized at 1.2 or
0.3 V, respectively. The ESR spectra were recorded under nonsaturating
microwave power conditions. The magnitude of modulation was chosen to
optimize the resolution and the signal-to-noise (S/N) ratio of the observed
spectra. The g values were calibrated with a Mn2‡ marker, and the
hyperfine splitting (hfs) values were determined by computer simulation on
a Calleo ESR Version 1.2 program (Calleo Scientific Software Publishers)
on a Macintosh personal computer.
p-methoxybenzaldehyde
(p-MeOC6H4CHO),
[2H4]acetaldehyde
(CD3CDO)] and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) were obtained
from Tokyo Chemical Industry Co., Ltd. (Japan) and purified by the
standard methods.[38] Tetra-n-butylammonium perchlorate (TBAP) used as
a supporting electrolyte was purchased from Sigma Chemical Co., purified
by successive recrystallizations from ethanol, and dried under vacuum at
408C. Acetonitrile (MeCN) used as solvent was purchased from Wako Pure
Chemical Ind. Ltd. (Japan), and purified by successive distillation over
CaH2 prior to use.
Theoretical calculations: The theoretical studies used the PM3 molecular
orbital method.[23] The calculations were performed by using the MOL-
MOLIS program Ver. 2.8 (Daikin Industries, Ltd). Final geometries and
energies were obtained by optimizing the total molecular energy with
respect to all structural variables. The geometries of the radicals were
optimized using the unrestricted Hartree ± Fock (UHF) formalism. The
heat of formation (DHf) values of the radicals were calculated with the
UHF-optimized structures by the half-electron (HE) method with the
restricted Hartree ± Fock (RHF) formalism.[40] The adiabatic ionization
potentials (Ip) were calculated as the difference in DHf between the radical
and the corresponding anion form. The reorganization energies of the inner
coordination spheres (li) associated with the structural change of active
aldehydes upon the electron-transfer oxidation were calculated as the
difference in DHf of the radicals with the same structures as the anion forms
and DHf with the structures optimized by the UHF formalism.
Spectral and kinetic measurements: Typically, DBU (2 mL, 1.3 Â 10 5 mol)
was added to a quartz cuvette (10 mm i.d.) which contained 3-benzyl-4-
‡
methylthiazolium bromide (1a : 4.4 Â 10 4 m) and o-tolualdehyde (1.2 Â
10 2 m) in deaerated MeCN (3.0 mL) at 298 K. The corresponding active
aldehyde (2a ) was formed. Similar experimental conditions were em-
ployed for formation of other active aldehydes. UV/Vis spectral change
associated with formation of the active aldehyde was monitored with a
Hewlett Packard 8453 diode array spectrophotometer. Kinetic measure-
ments for formation of thiazolium ylides and active aldehydes were carried
out with a Union RA-103 stopped-flow spectrophotometer under argon at
Chem. Eur. J. 1999, 5, No. 10
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2817 $ 17.50+.50/0
2817

S. Fukuzumi et al.
FULL PAPER
[16] M. Tazaki, M. Kumakura, S. Nagahama, M. Takagi, J. Chem. Soc.
Chem. Commun. 1995, 1763.
Acknowledgments
[17] F. J. Leeper, D. H. C. Smith, J. Chem. Soc. Perkin Trans. 1 1995, 861.
[18] a) T. Ugai, R. Tanaka, T. Dokawa, J. Pharm. Soc. Jpn. 1943, 63, 296;
b) W. Tagaki, H. Hara, J. Chem. Soc. Chem. Commun. 1973, 891;
c) J. A. Zoltewicz, J. K. OꢁHalloran, J. Org. Chem. 1978, 43, 1713; d) F.
Diederich, H.-D. Lutter, J. Am. Chem. Soc. 1989, 111, 8438; e) Y.-T.
Chen, G. L. Barletta, K. Haghjoo, J. T. Cheng, F. Jordan, J. Org. Chem.
1994, 59, 7714; f) K. Motesharei, D. C. Myles, J. Am. Chem. Soc. 1997,
119, 6674.
This work was partially supported by
a Grant-in-Aid for Scientific
Research (No. 101008) from the Japan Society for the Promotion of
Science and by a Grant-in-Aid for Scientific Research Priority Area
(Nos. 10149230, 10131242, 10146232, and 10125220) from the Ministry of
Education, Science, Culture and Sports, Japan.
[1] a) R. Breslow, J. Am. Chem. Soc. 1958, 80, 3719; b) L. O. Krampitz,
Annu. Rev. Biochem. 1969, 38, 213; c) R. Kluger in The Enzymes,
Vol. 20 (Ed.: D. S. Sigman), 3rd ed., Academic Press, New York, 1992,
pp. 271 ± 318; d) R. Kluger, Chem. Rev. 1987, 87, 863; e) A. Schellen-
berger, Angew. Chem. 1967, 79, 1050; Angew. Chem. Int. Ed. Engl.
1967, 6, 1024; f) The Biochemistry and Physiology of Thiamin
Diphosphate Enzymes (Eds.: H. Bisswanger, A. Schellenberger),
Intemann, Prien, Germany, 1996; g) The Biochemistry and Physiology
of Thiamin Diphosphate Enzymes (Eds.: H. Bisswanger, J. Ullrich),
VCH, Weinheim, 1991; h) Thiamin Pyrophosphate Biochemistry
(Eds.: A. Schellengerger, R. L. Schowen), CRC, Boca Raton, FL,
1988; i) Thiamin: Twenty Years of Progress (Eds.: H. Z. Sable, C. J.
Gubler), New York Academy of Sciences, New York, 1982.
[2] a) F. Dyda, W. Furey, S. Swaminathan, M. Sax, B. Farrenkopf, F.
Jordan, Biochemistry 1993, 32, 6165; b) X. Zeng, B. Farrenkopf, S.
Hohmann, F. Dyda, W. Furey, F. Jordan, Biochemistry 1993, 32, 2704;
c) I. Baburina, Y. Gao, Z. Hu, F. Jordan, S. Hohmann, W. Furey,
Biochemistry 1994, 33, 5630; d) A. Brown, N. Nemeria, J. Yi, D. Zhang,
W. B. Jordan, R. S. Machado, J. R. Guest, F. Jordan, Biochemistry
1997, 36, 8071; e) F. Guo, D. Zhang, A. Kahyaoglu, R. S. Farid, F.
Jordan, Biochemistry 1998, 37, 13379.
[19] a) R. Breslow, E. Kool, Tetrahedron Lett. 1988, 29, 1635; b) R.
Breslow, R. Kim, Tetrahedron Lett. 1994, 35, 699; c) R. Breslow, C.
Schmuk, Tetrahedron Lett. 1996, 37, 8241.
[20] The oxidation peak potentials of O-methylated active aldehyde
analogues have been reported to be in the range of
169 to
‡144 mV vs. SCE in DMSO. The formation of a radical cation
intermediate upon the electrochemical oxidation is suggested based
on chemical demonstration of formation of a dimer at the C2a carbon;
see: G. Barletta, A. C. Chung, C. B. Rios, F. Jordan, J. M. Schlegel, J.
Am. Chem. Soc. 1990, 112, 8144.
[21] Preliminary reports have appeared: a) I. Nakanishi, S. Itoh, T.
Suenobu, S. Fukuzumi, Angew. Chem. 1998, 110, 1040; Angew. Chem.
Int. Ed. 1998, 37, 992.
[22] I. Nakanishi, S. Itoh, T. Suenobu, H. Inoue, S. Fukuzumi, Chem. Lett.
1997, 707.
[23] J. J. P. Stewart, J. Comput. Chem. 1989, 10, 209, 221.
[24] J. E. Wertz, J. R. Bolton, Electron Spin Resonance, Elementary Theory
and Practical Applications, McGraw ± Hill, New York, 1972.
[25] S. Fukuzumi, Y. Tokuda, T. Kitano, T. Okamoto, J. Otera, J. Am.
Chem. Soc. 1993, 115, 8960.
[3] a) S. Sun, G. S. Smith, M. H. OꢁLeary, R. L. Schowen, J. Am. Chem.
Soc. 1997, 119, 1507; b) F. J. Alvarez, J. Ermer, G. Hübner, A.
Schellenberger, R. L. Schowen, J. Am. Chem. Soc. 1995, 117, 1678; c) S.
Sun, R. G. Duggleby, R. L. Schowen, J. Am. Chem. Soc. 1995, 117,
7317.
[4] L. J. Reed, Acc. Chem. Res. 1974, 7, 40.
[5] a) L. P. Hager, J. Biol. Chem. 1957, 229, 251; b) Y. A. Müller, G. E.
Schulz, Science 1993, 259, 965.
[26] The hfs values of aH(PhCH2) were evaluated from the 2s spin densities
.
.
of 2a and 2 f calculated by the PM3 method with the UHF
formalism. The hfs values of aN(N), aH(C4), aH(C5), aH(C2'), and
aH(C4') were estimated from the 2pz spin densities (1i) calculated with
the RHF formalism by using the relation aN ˆ 251i and aH ˆ 22.51i.[24]
[27] I. Nakanishi, S. Itoh, T. Suenobu, S. Fukuzumi, Chem. Commun. 1997,
1927.
[28] a) R. L. Ward, S. I. Weissman, J. Am. Chem. Soc. 1957, 79, 2086; b) R.
Chang, J. Chem. Educ. 1970, 47, 563; c) M. T. Watts, M. L. Lu, R. C.
Chen, M. P. Eastman, J. Phys. Chem. 1973, 77, 2959; d) K. S. Cheng, N.
Hirota, Investigation of Rates and Mechanisms of Reactions, Vol. VI
(Ed.: G. G. Hammes), Wiley-Interscience, New York, 1974, p. 565.
[29] a) R. A. Marcus, Annu. Rev. Phys. Chem. 1964, 15, 155; b) R. A.
Marcus, Angew. Chem. 1993, 105, 1161; Angew. Chem. Int. Ed. Engl.
1993, 32, 1111.
[6] a) K. Uyeda, J. C. Rabinowitz, J. Biol. Chem. 1971, 246, 3120; b) L.
Kerscher, D. Oesterhelt, Eur. J. Biochem. 1981, 116, 595.
[7] a) F. Jordan, Z. H. Kudzin, C. B. Rios, J. Am. Chem. Soc. 1987, 109,
4415; b) F. G. Bordwell, A. V. Satish, F. Jordan, C. B. Rios, A. C.
Chung, J. Am. Chem. Soc. 1990, 112, 792; c) X. Zeng, A. Chung, M.
Haran, F. Jordan, J. Am. Chem. Soc. 1991, 113, 5842; d) G. Barletta,
W. P. Huskey, F. Jordan, J. Am. Chem. Soc. 1992, 114, 7607; e) C. C.
Chiu, K. Pan, F. Jordan, J. Am. Chem. Soc. 1995, 117, 7027; f) C. C.
Chiu, A. Chung, G. Barletta, F. Jordan, J. Am. Chem. Soc. 1996, 118,
11026; g) G. L. Barletta, Y. Zou, W. P. Huskey, F. Jordan, J. Am.
Chem. Soc. 1997, 119, 2356.
[30] L. Eberson, Adv. Phys. Org. Chem. 1982, 18, 79 and references therein.
[31] C. V. Krishnan, B. S. Brunschwig, C. Creutz, N. Sutin, Adv. Phys. Org.
Chem. 1985, 107, 2005.
[32] R. Berkoff, K. Krist, H. D. Gafney, Inorg. Chem. 1980, 19, 1.
[33] C. Creutz, H. A. Schwartz, N. Sutin, J. Am. Chem. Soc. 1984, 106, 3036.
[8] D. Hilvert, R. Breslow, Bioorg. Chem. 1984, 12, 206.
‡
[34] [CoI(phen)3] was generated in the reduction of the CoII complex by
[9] a) S. Shinkai, T. Yamashita, Y. Kusano, O. Manabe, Tetrahedron Lett.
1980, 21, 2543; b) S. Shinkai, T. Yamashita, Y. Kusano, O. Manabe, J.
Org. Chem. 1980, 45, 4947; c) S. Shinkai, T. Yamashita, Y. Kusano, O.
Manabe, J. Am. Chem. Soc. 1982, 104, 563.
[10] a) S. Ohshima, N. Tamura, T. Nabeshime, Y. Yano, J. Chem. Soc.
Chem. Commun. 1993, 712; b) A. Takaki, K. Utsumi, T. Kajiki, T.
Kuroi, T. Nabeshima, Y. Yano, Chem. Lett. 1997, 75; c) K. Utsumi, Y.
Nishihara, K. Hoshino, S.-I. Kondo, T. Nabeshima, Y. Yano, Chem.
Lett. 1997, 1081.
sodium amalgam in deaerated MeCN. Visible ± NIR (MeCN), lmax
(nm) [10 3e (m 1 cm 1)]: 426 [1.7], 1415 [4.7].
[35] a) K. M. Doxsee, M. Feigel, K. D. Stewart, J. W. Canary, C. B.
Knobler, D. J. Cram, J. Am. Chem. Soc. 1987, 109, 3098; b) V.
Boekelheide, G. K. Vick, J. Am. Chem. Soc. 1956, 78, 653; c) W. G.
Brown, Org. React. 1951, 469; d) H. C. Brown, P. M. Weissmann, N. M.
Yoon, J. Am. Chem. Soc. 1966, 88, 1458; e) G. J. Karabatsos, R. L.
Shone, J. Org. Chem. 1968, 33, 619.
[11] a) H. Inoue, K. Higashiura, J. Chem. Soc. Chem. Commun. 1980, 549;
b) H. Inoue, S. Tamura, J. Chem. Soc. Chem. Commun. 1985, 141; c) H.
Inoue, S. Tamura, J. Chem. Soc. Chem. Commun. 1986, 858.
[12] a) S.-W. Tam, L. Jimenez, F. Diederich, J. Am. Chem. Soc. 1992, 114,
1503; b) W.-W. Tam-Chang, L. Jimenez, F. Diederich, Helv. Chim.
Acta 1993, 76, 2616; c) P. Mattei, F. Diederich, Angew. Chem. 1996,
108, 1434;Angew. Chem. Int. Ed. Engl. 1996, 35, 1341; d) P. Mattei, F.
Diederich, Helv. Chim. Acta 1997, 80, 1555.
[36] a) D. K. Liu, B. S. Brunschwig, C. Creutz, N. Sutin, J. Am. Chem. Soc.
1986, 108, 1749; b) D. J. Szalda, C. Creutz, D. Mahajan, N. Sutin, N.
Inorg. Chem. 1983, 22, 2372.
[37] F. H. Burstall, R. S. Nyholm, J. Chem. Soc. 1952, 3570.
[38] D. D. Perrin, W. L. F. Armarego, D. R. Perrin, Purification of Labo-
ratory Chemicals, Pergamon, Elmsford, 1966.
[39] C. K. Mann, K. K. Barnes, Electrochemical Reactions in Nonaqueous
Systems, Marcel Dekker, New York, 1990.
[13] J. Grosby, R. Stone, G. E. Lienhard, J. Am. Chem. Soc. 1970, 92, 2891.
[14] Y. Murakami, J.-I. Kikuchi, Y. Hisaeda, O. Hayashida, Chem. Rev.
1996, 96, 721.
[40] T. Clark, A Handbook of Computational Chemistry, Wiley, New York,
1985, p. 97.
[15] L. M. Ferreira, H. T. Chaves, A. M. Lobo, S. Prabhakar, H. S. Rzepa, J.
Chem. Soc. Chem. Commun. 1993, 133.
Received: February 15, 1999 [F1618]
2818
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0510-2818 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 10