Determination of the C4-H bond dissociation energies of NADH models and their radical cations in acetonitrile

Article abstract of DOI:10.1002/chem.200390108

Heterolytic and homolytic bond dissociation energies of the C4-H bonds in ten NADH models (seven 1,4-dihydronicotinamide derivatives, two Hantzsch 1,4-dihydropyridine derivatives, and 9,10-dihydroacridine) and their radical cations in acetonitrile were evaluated by titration calorimetry and electrochemistry, according to the four thermodynamic cycles constructed from the reactions of the NADH models with N,N,N′,N′-tetramethyl-p-phenylenediamine radical cation perchlorate in acetonitrile (note: C9-H bond rather than C4-H bond for 9,10-dihydroacridine; however, unless specified, the C9-H bond will be described as a C4-H bond for convenience). The results show that the energetic scales of the heterolytic and homolytic bond dissociation energies of the C4-H bonds cover ranges of 64.2-81.1 and 67.9-73.7 kcal mol^-1 for the neutral NADH models, respectively, and the energetic scales of the heterolytic and homolytic bond dissociation energies of the (C4-H)^.+ bonds cover ranges of 4.1-9.7 and 31.4-43.5 kcal mol^-1 for the radical cations of the NADH models, respectively. Detailed comparison of the two sets of C4-H bond dissociation energies in 1-benzyl-1,4-dihydronicotinamide (BNAH), Hantzsch 1,4-dihydropyridine (HEH), and 9,10-dihydroacridine (AcrH₂) (as the three most typical NADH models) shows that for BNAH and AcrH₂, the heterolytic C4-H bond dissociation energies are smaller (by 3.62 kcal mol^-1) and larger (by 7.4kcal mol^-1), respectively, than the corresponding homolytic C4-H bond dissociation energy. However, for HEH, the heterolytic C4-H bond dissociation energy (69.3 kcal mol^-1) is very close to the corresponding homolytic C4-H bond dissociation energy (69.4 kcal mol^-1). These results suggests that the hydride is released more easily than the corresponding hydrogen atom from BNAH and vice versa for AcrH₂, and that there are two almost equal possibilities for the hydride and the hydrogen atom transfers from HEH. Examination of the two sets of the (C4-H)^.+ bond dissociation energies shows that the homolytic (C4-H)^.+ bond dissociation energies are much larger than the corresponding heterolytic (C4-H)^.+ bond dissociation energies for the ten NADH models by 23.3-34.4 kcal mol^-1; this suggests that if the hydride transfer from the NADH models is initiated by a one-electron transfer, the proton transfer should be more likely to take place than the corresponding hydrogen atom transfer in the second step. In addition, some elusive structural information about the reaction intermediates of the NADH models was obtained by using Hammett-type linear free-energy analysis.

Full text of DOI:10.1002/chem.200390108

FULL PAPER
À
Determination of the C4 H Bond Dissociation Energies of NADH Models
and Their Radical Cations in Acetonitrile
Xiao-Qing Zhu*, Hai-Rong Li,Qian Li,Teng Ai,Jin-Yong Lu,Yuan Yang,
and Jin-Pei Cheng*[a]
Abstract: Heterolytic and homolytic
ranges
of
4.1 9.7
and
31.4
that the hydride is released more easily
than the corresponding hydrogen atom
from BNAH and vice versa for AcrH2,
and that there are two almost equal
possibilities for the hydride and the
hydrogen atom transfers from HEH.
Examination of the two sets of the
bond dissociation energies of the C4 H
43.5 kcalmolÀ1 for the radical cations
of the NADH models, respectively. De-
tailed comparison of the two sets
À
bonds in ten NADH models (seven 1,4-
dihydronicotinamide derivatives, two
Hantzsch 1,4-dihydropyridine deriva-
tives, and 9,10-dihydroacridine) and
their radical cations in acetonitrile were
evaluated by titration calorimetry and
electrochemistry, according to the four
thermodynamic cycles constructed from
the reactions of the NADH models with
N,N,N',N'-tetramethyl-p-phenylenedi-
amine radical cation perchlorate in ace-
À
of C4 H bond dissociation energies
in
1-benzyl-1,4-dihydronicotinamide
(BNAH), Hantzsch 1,4-dihydropyridine
(HEH), and 9,10-dihydroacridine
(AcrH2) (as the three most typical
NADH models) shows that for BNAH
.
‡
À
(C4 H)
shows that the homolytic (C4 H)
bond dissociation energies
.
‡
À
bond dissociation energies are much
larger than the corresponding heterolyt-
À
and AcrH2, the heterolytic C4 H bond
.
‡
À
dissociation energies are smaller (by
ic (C4 H) bond dissociation energies
for the ten NADH models by 23.3
34.4 kcalmolÀ1; this suggests that if the
hydride transfer from the NADH mod-
els is initiated by a one-electron transfer,
the proton transfer should be more
likely to take place than the correspond-
ing hydrogen atom transfer in the sec-
ond step. In addition, some elusive
structural information about the reac-
tion intermediates of the NADH models
was obtained by using Hammett-type
linear free-energy analysis.
3.62 kcalmolÀ1)
and
larger
(by
tonitrile (note: C9 H bond rather than
7.4 kcalmolÀ1), respectively, than the
À
À
À
C4 H bond for 9,10-dihydroacridine;
corresponding homolytic C4 H bond
dissociation energy. However, for
HEH, the heterolytic C4 H bond dis-
sociation energy (69.3 kcalmolÀ1) is very
À
however, unless specified, the C9 H
bond will be described as a C4 H bond
for convenience). The results show that
the energetic scales of the heterolytic
and homolytic bond dissociation ener-
À
À
close to the corresponding homolytic
À
C4 H
bond dissociation energy
gies of the C4 H bonds cover ranges of
(69.4 kcalmolÀ1). These results suggests
À
64.2 81.1 and 67.9 73.7 kcalmol À1 for
the neutral NADH models, respectively,
and the energetic scales of the hetero-
lytic and homolytic bond dissociation
Keywords: bond energy ¥ calorime-
try
¥ effective charge ¥ NADH
models ¥ radical ions
.
‡
À
energies of the (C4 H) bonds cover
NH2
NH2
H
Introduction
H
H
CONH2
N
N
N
N
CONH2
N
N
O
O
O
O
- H
+ H
N
H
N
N
N
H
P
O
O
P
O
P
O
O
P
O
The reduced form of the nicotina-
mide adenine dinucleotide coen-
zyme (NADH) plays an important
role in many bioreductions by
transferring a hydride ion or elec-
O
O
O
O
(1)
H
H
H
H
H
H
H
H
OH
OH
OH
OH
H
H
H
H
OH
H
H
OH
OH OH
OH OH
OH OH
NAD+
NADH
tron to the surrounding substrates [Eq. (1)].[1] The mecha-
nistic details of the hydride transfer has been a subject of
great interest and is increasingly drawing the attention
of researchers.[2, 3] One of the issues in focus is whether
the hydride transfer proceeds by direct one-step or multi-
step sequence: electron/proton/electron, electron/hydrogen
atom, or hydrogen atom/electron (Scheme 1). Based on
published experimental results, a direct one-step mecha-
[a] Prof. X.-Q. Zhu, Prof. J.-P. Cheng, H.-R. Li, Dr. Q. Li, T. Ai, J.-Y. Lu,
Y. Yang
Department of Chemistry
State Key Laboratory of Elemento-Organic Chemistry
Nankai University, Tianjin 300071 (China)
Fax : (‡86)022-2350-2458
E-mail: xqzhu@nankai.edu.cn
Chem. Eur. J. 2003, 9, No. 4
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0871 $ 20.00+.50/0
871

X.-Q. Zhu, J.-P. Cheng et al.
FULL PAPER
‡
NADH
NADH
NADH
NADH
!
!
!
!
NAD ‡ HÀ
DHhet(NADH) ˆ ?
DHhomo(NADH) ˆ ?
(2)
(3)
(4)
(5)
.
.
NAD ‡ H
.
.
.
.
‡
‡
‡
‡
NAD ‡ H
DHhet(NADH ) ˆ ?
.
.
‡
‡
NAD ‡ H
DHhomo(NADH ) ˆ ?
we have successfully developed four new thermodynamic
cycles according to the reactions of the N,N,N',N'-tetramethyl-
p-phenylenediamine radical cation with some NADH models
in acetonitrile; these can be used to directly evaluate the bond
À
dissociation energies of the C4 H bonds in NADH models
and their radical cations. Here we report the details.
Scheme 1. Possible pathways of the hydride transfer from NADH.
nism was suggested for the reductions of ketones, aldehydes,
imines, etc.[4] However, at the same time, a multi-step
sequence mechanism was also proposed for the reductions
Results
N,N,N',N'-Tetramethyl-p-phenylenediamine radical cation
‡
of FeIII, Fc , and many other single-electron oxidants.[5]
.
‡
À
perchlorate (TMPA ClO4 ) is a very strong single-electron
organic oxidant,[14] which can readily oxidise NADH models
such as 1-benzyl-1,4-dihydronicotinamide (BNAH), Hantzsch
1,4-dihydropyridine (HEH), 10-methyl-9,10- dihydroacridine
(AcrH2), and many other 1,4-dihydropyridine derivatives in
acetonitrile to give N,N,N',N'-tetramethyl-p-phenylenedi-
amine and the corresponding pyridinium cation derivatives,
respectively. The stoichiometry of the reactions was deter-
mined by using spectrophotometric titration (see Figure 1).
Moreover, much conflicting evidence can be found for the
same NADH mimic reaction, such as the reduction of quinone
derivatives.[6] In short, the mechanistic details of the hydride
transfer is still unclear at present; no common view yet
explains all the diverse and confusing experimental results.
Clearly, a detailed thermodynamic analysis on each mecha-
nism-step for the hydride transfer from NADH and its models
must be performed to completely solve this important
problem.[7]
Examination of past reports on this subject shows that
although there are some chemists, such as Saveant and co-
workers,[8] who devoted much time to the study of the
thermodynamics of NADH and its models, their main
attention was limited to the electrochemistry and acid-base
‡
equilibria of NADH/NAD analogues. Rather scant attention
has been paid to the detailed thermodynamic analysis on each
mechanistic step for the hydride transfer from NADH and its
models. From Scheme 1, it is clear that lack of knowledge
about the energies of the related bond dissociation of NADH
and its models prevents us making a detailed thermodynamic
analysis on each mechanistic step for the hydride transfer.
However, despite the clear importance and necessity of the
bond energy data of NADH and its model compounds,
.
‡
Figure 1. Plot of the concentration of TMPA consumed in the electron-
.
‡
transfer reactions of BNAH with TMPA versus the concentration of
BNAH.
À
essentially no direct determination of the C4 H bond
dissociation energies either for the neutral NADH models
or for their corresponding radical cations have yet been
reported, except for some relative dissociation energies of the
.
‡
The results showed that two moles of TMPA were required
to thoroughly consume one mole of NADH models such as
C4 H bond for a few NADH models[9] and estimation for the
À
À
absolute dissociation energy of the C4 H bond for only a
BNAH in dry acetonitrile [Eq. (6)]. The reaction mechanism
limited few NADH models.[10] To access the thermodynamic
data of each mechanistic step for the hydride transfer from
NADH and its models, as shown in Scheme 1, the key work is
was proposed to be an eÀ
transfer (Scheme 2).[15]
H
eÀ sequence of hydride
‡
On the basis of the reaction [Eq. (6)], four thermodynamic
cycles were constructed as shown in Schemes 3 and 4, from
which four formulae to evaluate the heterolytic and homolytic
À
to determine the heterolytic and homolytic C4 H bond
dissociation energies in NADH and its radical cation
.
‡
À
NADH [Eqs. (2) (5)]. Since it is impossible to obtain them
directly by using the well-known thermodynamic approaches
introduced by Bordwell,[11] Arnett,[12] and others,[13] it is
necessary to develop a new method suitable for the determi-
C4 H bond dissociation energies in NADH models and their
radical cations were derived [Eqs. (7) and (8) from the cycles
in Scheme 3 and Eqs. (9) and (10) from the cycles in
Scheme 4].[16] From Equations (7) (10), the heterolytic and
À
À
nation of the C4 H bond scission energies in solution; this has
been a goal in our research program for a long time. Recently,
homolytic bond dissociation energies of the C4 H bonds in
NADH models and their radical cations can be easily
872
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0872 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 4

NADH Models
871 880
dropyridine derivatives (2), and
H
H
NMe2
NMe2
NHMe2
H
N
9,10-dihydroacridine
deriva-
CONH2
CONH2
+
tives (3) (Scheme 5).
(6)
+
+
2
N
The titration calorimetry was
performed in acetonitrile at
258C on a CSC4200 isothermal
titration calorimeter taking
10 automatic injections; the re-
NMe2
NMe2
NMe2
CH2Ph
CH2Ph
BNA+
TMPA•+
TMPA
TMPAH+
BNAH
action heats were obtained by area integration of each peak
except the first one (calibration graph, see Figure 2). The ten
enthalpy changes of the reaction [Eq. (6)] (DHrxn) and
‡
the deprotonation enthalpy change of TMPAH
.
‡
Scheme 2. Reaction mechanism of BNAH with TMPA
.
‡
[DHa(TMPAH )] in acetonitrile, as well as redox potentials
.
‡
of TMPA
,
NADH, and
NAD models are listed in
Table 1. Heterolytic and homo-
‡
À
lytic C4 H bond dissociation
energies for the ten NADH
models and the corresponding
ten radical cations are summar-
ized in Table 2; these data pro-
vide a good platform for the
comparison of the relative ease
À
with which the C4 H bond
.
‡
Scheme 3. Two thermodynamic cycles constructed from the reactions of TMPA with NADH Models.
.
.
‡
scission processes among the
different types of NADH mod-
els and their radical cations
‡
‡
i) DHi ˆ DHrxn
.
ii) DGii ˆ ÀF[E1/2(NAD ) À E1/2(TMPA )].
iii) DGiii ˆ ÀF[E1/2(TMPA ) À E1/2(NADH)].
. .
‡ ‡
iv) DHiv ˆ DHa(TMPAH ). v) DGv ˆ ÀF[E1/2(H‡/0) À E1/2(TMPA )]. vi) DHvi ˆ DHhet(NADH ). vii) DHvii
ˆ
‡
.
‡
DHhomo(NADH
)
Scheme 5. 1: G1 ˆ PhCH2 (BNAH), 4-CH3OPh (CH3OPNAH), 4-CH3Ph
(CH3PNAH), Ph (PNAH), 4-ClPh (ClPNAH), 4-BrPh (BrPNAH),
4-CF3Ph (CF3PNAH). 2: G2 ˆ H (HEH), CH3 (MeHEH). 3: AcrH2.
Scheme 4. Two thermodynamic cycles were made on the basis of the
.
‡
À
heterolytic and homolytic (C4 H) bond dissociations in the radical
cations of NADH models.
obtained only if the enthalpy change of the reaction [Eq. (6)]
‡
(DHrxn), the enthalpy change of deprotonation of TMPAH
‡
[DHa(TMPAH )], and the redox potentials of the related
species are available. Evidently, the enthalpy changes can be
obtained from the corresponding heats of reaction, which can
be directly determined by titration calorimetery. The redox
potentials can be measured by cyclic voltammetry.
.
.
‡
‡
‡/0
DHhomo(NADH
)
ˆ
F[E1/2(TMPA ) À E1/2(NADH)] À F[E1/2(H
)
(7)
(8)
.
‡
‡
À E1/2(TMPA )] ‡ DHa(TMPAH ) ‡ DHrxn
.
.
‡
‡
DHhet(NADH
)
ˆ F[E1/2(TMPA ) À E1/2(NADH)]
.
‡ ‡
‡
À F[E1/2(NAD ) À E1/2(TMPA )] ‡ DHa(TMPAH ) ‡ DHrxn
Figure 2. Titration calibration graph: BNAH in acetonitrile was titrated
.
.
‡
DHhet(NADH)
ˆ
DHhomo(NADH ) À F[E1/2(H0/À) À E1/2(NADH)] (9)
‡
into N,N,N',N'-tetramethyl-p-phenylenediamine radical cation (TMPA
)
.
‡
in acetonitrile solution ([BNAH] ˆ 2.01mm, [TMPA ] ˆ 10.87mm).
.
DHhomo(NADH)
ˆ
DHhet(NADH ) À F[E1/2(H‡/0) À E1/2(NADH)] (10)
‡
Three types of 1,4-dihydropyridine derivatives were chosen
as NADH models in this work; these include seven 1,4-
dihydronicotinamide derivatives (1), two Hantzsch 1,4-dihy-
take place. The stabilities of the related reactive intermediates
can also be visualized on the basis of the derived energetic
data.
Chem. Eur. J. 2003, 9, No. 4
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0873 $ 20.00+.50/0
873

X.-Q. Zhu, J.-P. Cheng et al.
FULL PAPER
Table 1. Reaction enthalpy changes of the NADH models with N,N,N',N'-tetramethyl-p-phenylenediamine
bis(diethylphosphino)propane,
dmpp ˆ bis(dimethylphosphi-
no)propane);[19] this suggests
that the NADH models with
the 1,4-dihydropyridine struc-
ture would be a good hydride
donor reducing agent, which
fits well with many other ex-
perimental results.[4] The reason
could be that the pyridinium
salts formed by the heterolytic
.
‡
radical cation perchlorate (TMPA ) and the redox potentials [V verus Fc‡/0] of the relative species in acetonitrile.
‡
‡
NADH Models
DHrxn[a]
Eox(NADH)[b]
Ered(NAD )[b]
E1/2(TMPA)[b]
DHa(TMPAH )[c]
1
BNAH
CH3OPNAH
CH3PNAH
PNAH
ClPNAH
BrPNAH
CF3PNAH
2
À 38.0
À 35.3
À 34.6
À 33.4
À 351.90.39
À 31.7
À 29.6
0.259
0.281
0.313
0.363
À 1.297
À 1.247
À 1.209
À 1.154
À 1.087
À 1.082
À 1.003
0.296
0.296
0.296
0.296
0.296
0.296
0.296
9.1
9.1
9.1
9.1
9.1
9.1
9.1
0.400
0.471
HEH
MeHEH
3
À 32.90.507
À 32.3
À 1.142
À 1.089
0.296
0.296
9.1
9.1
À
C4 H bond dissociation are
0.498
0.492
more stable than the primary
benzylic carbenium ions in ace-
tonitrile. This result may pro-
vide a reason why NADH coen-
zyme chooses 1,4-dihydropyri-
dine as its reaction active center
in vivo. Detailed examination
of the heterolytic bond dissoci-
AcrH2
À 21.1
À 0.819
0.296
9.1
[a] DHrxn obtained from the reaction heat of Equation (6) by switching the sign of the heat value, the latter was
measured by titration calorimetry in MeCN at 258C. The data given in kcalmolÀ1 were average values of at least
two independent runs, each of which was again an average value of ten consecutive titrations, except the first. The
‡/0
reproducibility is 0.5 Æ 0.2 kcalmolÀ1. [b] Measured in MeCN at 258C in volts by mV versus Fc
and
‡
‡
reproducible to 5 mV or better. [c] DHa(TMPAH ) is the enthalpy change of the deprotonation of TMPAH
equal to the reaction heat of TMPAwith HClO4 in acetonitrile, the latter was measured by titration calorimetry at
258C. The data [kcalmolÀ1] obtained were average values of two independent runs, each of which was again an
average value of 10 consecutive titrations except the first. Uncertainty is smaller than 0.6 kcalmolÀ1
.
À
ation energies of the C4 H
bonds in BNAH, HEH, and
AcrH2 (as three typical NADH
models) reveals that the magni-
tude of the heterolytic bond
dissociation energies of the
À
Table 2. Heterolytic and homolytic C4 H bond dissociation energies of the NADH models and their radical
cations in acetonitrile [kcalmolÀ1].[a]
.
.
‡
‡
NADH Models
DHhet(NADH)
DHhomo(NADH)
DHhet(NADH
)
DHhomo(NADH )
À
C4 H bonds increases in the
1
following order of NADH mod-
els: BNAH (64.2 kcalmolÀ1) <
HEH (69.3 kcalmolÀ1) < AcrH2
(81.1 kcalmolÀ1); this is an in-
dication that BNAH should be
the strongest hydride donor,
while AcrH2 should be the
weakest hydride contributor
among the three NADH mod-
els. This has been supported by
BNAH
CH3OPNAH
CH3PNAH
PNAH
ClPNAH
BrPNAH
CF3PNAH
2
64.2
66.9
67.5
68.8
70.2
70.4
72.6
67.98.7
69.4
69.2
69.2
69.1
32.0
9.7
8.8
7.6
6.8
34.2
34.1
34.2
34.9
35.0
35.5
69.2
69.5
6.8
5.4
HEH
MeHEH
3
69.3
69.9
69.4
68.8
4.5
4.1
31.4
32.2
AcrH2
81.1
73.7
9.2
43.5
6]
many experiments.[3
.
.
‡
‡
[a] DHhet(NADH), DHhomo(NADH), DHhet(NADH and DHhomo(NADH
) were estimated from Equa-
To verify the heterolytic
bond dissociation energies of
tions (7) (10), taking E1/2(H‡/0) ˆ À2.307 (V versus Fc‡/0), E1/2(H0/À) ˆ À1.137 [V versus Fc‡/0] (the values from
Parker's work[24] adjusted to versus Fc‡/0 by adding À0.537 V). Standard deviations were estimated to be smaller
than 1 kcalmolÀ1
.
[17]
À
the C4 H bonds in the NADH
models obtained using the pres-
‡
ent method, the reaction heats of AcrH with NADH models
Discussion
1 and 2 in acetonitrile were determined. The results are
summarized in Table 3. Since a hydride exchange completed
Heterolytic and homolytic bond dissociation energies of the
‡
the reactions of AcrH with 1 and 2 [Eq. (11)],[20] the reaction
À
C4 H bonds in the NADH models and the relative stabilities
heats (Qrxn) should be equal to the difference of the
of the corresponding pyridinium cation and pyridinyl radical:
À
heterolytic C4 H bond dissociation energies between AcrH2
The second column in Table 2 shows that the energetic scale
À
and 1 and 2. Evidently, the differences between DHhet(1,2)
and DHhet(AcrH2) in Table 3 are well matched with the
corresponding reaction heats; this suggests that the hetero-
of the heterolytic bond dissociation of the C4 H bonds in the
ten NADH models in acetonitrile ranges from 64.2 kcalmolÀ1
for BNAH to 81.1 kcalmolÀ1 for AcrH2. This is much smaller
than the hydride affinities of the primary benzylic carbenium
ion in acetonitrile (e.g., the hydride affinities of 118, 106, 112,
À
lytic bond dissociation energies of the C4 H bonds in the
NADH models obtained by the present method should be
reliable, and also supports the efficiency of the present
method.
122, 121, and 122 kcalmolÀ1 for PhCH2 , 4-CH3OC6H4CH2 ,
‡
‡
‡
‡
‡
4-MeC6H4CH2 ,
4-FC6H4CH2 ,
4-ClC6H4CH2 ,
and
4-CNC6H4CH2 , respectively),[18] but close to the hydride
affinities of some transition-metal hydrides in acetonitrile
(e.g., the hydride affinities of 66.2 and 60.4 kcalmolÀ1 for
‡
From the third column in Table 2, the energetic scale of the
À
homolytic bond dissociation of the C4 H bonds in the ten
NADH models in acetonitrile ranges from 67.9kcalmol À1 for
BNAH to 73.7 kcalmolÀ1 for AcrH2, which is smaller than the
‡
‡
[HNi(depp)2] and [HNi(dmpp)2] , respectively; depp ˆ
874
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0874 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 4

NADH Models
871 880
Table 3. The differences between DHhet(AcrH2) and DHhet(NADH) along
with the reaction heats (Qrxn) of AcrH with the other NADH models in
acetonitrile [kcalmolÀ1].
1-aryl-1,4-dihydronicotinamide GPNAH, it is interesting to
find that when the substituent G is an electron-donating group
(EDG), the heterolytic bond dissociation energies of the
‡
[a]
NADH Models
DHhet(AcrH2) À DHhet(NADH)
Qrxn
À
C4 H bonds are smaller than the corresponding homolytic
À
bond dissociation energies of the C4 H bonds, but when the
BNAH
16.916.6
14.2
13.6
CH3OPNAH
CH3PNAH
PNAH
14.7
13.6
12.0
substituent G is an electron-withdrawing group (EWG), the
DHhet(NADH) becomes larger than the corresponding
DHhomo(NADH). This result suggests that when the substitu-
ent G is an EDG, the NADH model has better hydride-
donating properties than when G is an EWG, but when the G
is an EWG, the NADH model become a better donor of
hydrogen atoms than when the substituent is an EDG. Clearly,
whether the hydride or the hydrogen atom is more likely to be
released from 1-aryl-1,4-dihydronicotinamide can be control-
led by changing the remote substituent G.
12.3
ClPNAH
BrPNAH
CF3PNAH
HEH
10.910.5
10.7
8.5
11.8
11.2
10.3
8.9
11.6
11.0
MeHEH
‡
[a] Qrxn denoting the reaction heat of AcrH with the NADH models 1, 2
[Eq. (11)] measured by titration calorimetry in CH3CN at 258C. The data
obtained were average values of at least two independent runs, each of
which was again an average value of ten consecutive titrations except the
first. The reproducibility is 0.5 Æ 0.2 kcalmolÀ1
.
Heterolytic and homolytic
bond dissociation energies of
H
H
N
H
H
H
.
H
‡
À
the (C4 H)
bonds in the
CONH2
CONH2
CH3CN
.
‡
NADH models: As stated in
the introduction section, there
are many chemical mimic ex-
periments that support the idea
that the reactions of the NADH
(11)
+
+
+ Qrxn
N
N
N
CH2Ph
CH3
CH2Ph
CH3
models were initiated by a single-electron transfer.[5, 28] It is
well recognized that if the reaction of NADH is initiated by a
single-electron transfer, an incipient radical cation intermedi-
bond dissociation energies (BDE) of 9-substituted fluorenes
(Fl-H) in dimethyl sulfoxide[21] (e.g., the BDEs of 76.0, 74.0,
and 82.1 kcalmolÀ1 for 9-CO2Me-Fl-H, 9-Ph-Fl-H, and Fl-H,
.
‡
respectively),[22, 23] but close to the M H BDE of some metal
À
ate NADH should be formed from which a proton transfer
or a hydrogen-atom transfer is believed to follow in the
second reaction step. So, the question arises as to which
species (hydride or hydrogen atom) is more likely to transfer
in the second reaction step. Clearly, it is necessary to evaluate
and compare the heterolytic and homolytic bond dissociation
hydrides in acetonitrile, which is a good hydrogen-atom donor
(such as the BDE×s of 61.5, 69.2, 72.3, 57.1, and 64.9 kcalmolÀ1
for ([(h5-C5H5)Cr(CO)3H], [(h5-C5H5)Mo(CO)3H], [(h5-
C5H5)W(CO)3H], [(h5-C5H5)Fe(CO)2H], and [(h5-C5H5)Ru-
(CO)2H], respectively).[24] These results suggest that the three
types of NADH models (1, 2, 3) are not only good hydride
donors, but also good hydrogen-atom contributors. This is
supported by their reactions with some organic bromides[25]
.
‡
À
energies of the (C4 H) bonds in the radical cations of the
NADH models. From columns 4 and 5 in Table 2, we find that
the energetic scale of the heterolytic bond dissociation of the
.
.
‡
‡
and peroxides.[26] Meanwhile, the low homolytic C4 H bond
dissociation energies hint that the pyridinyl radicals NAD
À
À
(C4 H) bonds in the nine NADH models in acetonitrile
ranges from 4.1 kcalmolÀ1, for the radical cation of N-
methylated Hantzsch 1,4-dihydropyridine, to 9.7 kcalmolÀ1,
for the radical cation of 1-(para-methoxylphenzyl)-1,4-dihy-
dronicotinamide. The energetic scale of the corresponding
.
À
formed by homolytic C4 H bond dissociation from the
NADH models should not be too unstable in acetonitrile
solution: they would be able to be captured by diamagnetic
radical scavengers (spin traps) or other reactive radicals; this,
in fact, has been confirmed by extensive experimental
results.[27] Comparison of columns 2 and 3 in Table 2 shows
that when the NADH model is BNAH, the DHhomo of the
.
‡
À
homolytic bond dissociation of the (C4 H) bonds ranges
.
from 31.4 kcalmolÀ1 for HEH
to 43.5 kcalmolÀ1 for
‡
.
‡
.
AcrH2 . Clearly, a simple comparison of the heterolytic
‡
À
(C4 H) bond cleavage energies with the corresponding
.
‡
À
À
homolytic (C4 H) bond cleavage energies reveals that the
heterolytic (C4 H) bond cleavage energies are quite a bit
smaller than the corresponding homolytic (C4 H) bond
cleavage energies, in fact by 23.3 34.3 kcalmolÀ1. This
indicates that the proton transfer from the NADH models
should be much easier than the hydrogen-atom transfer from
the same NADH models. From the very low heterolytic
bond dissociation energies of the (C4 H) bonds (4.1
9.7 kcalmolÀ1), it is conceivable that the proton transfer
C4 H bond is larger than the corresponding DHhet by
.
‡
3.7 kcalmolÀ1, but when the NADH model is AcrH2, the
À
.
‡
À
À
DHhomo of the C9 H bond is smaller than the corresponding
DHhet by 7.4 kcalmolÀ1. This suggests that if BNAH is used as
the NADH model to react with an oxidant, initial hydride
transfer should be more likely than the initial hydrogen atom
transfer, but if AcrH2 is used as the NADH model, the case is
reversed; this has been also supported by many other
reports.[2 6] Examining the DHhet (69.3 kcalmolÀ1) and DHhomo
.
‡
.
‡
.
‡
À
À1
À
should be extremely fast and is possibly diffusion-controlled
(69.4 kcalmol ) of the C4 H bond for HEH shows that the
‡
in solution. Here it is concluded that the eÀ
H
eÀ sequence
DHhet and DHhomo are close to each other; this indicates that
the hydride and hydrogen-atom transfer from HEH is almost
equally possibile. In addition, by detailed inspection of the
relationship of DHhet(NADH) and DHhomo(NADH) for the
of hydride transfer should be the most likely among all the
possible multi-step hydride-transfer mechanisms of NADH
and its models.
Chem. Eur. J. 2003, 9, No. 4
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0875 $ 20.00+.50/0
875

X.-Q. Zhu, J.-P. Cheng et al.
FULL PAPER
.
‡
Examination of the heterolytic bond dissociation energies
oxidation of the nine NADH models by TMPA would
readily take place and quickly go to completion. The
experimentally observed clean and rapid reactions in this
work do in fact support this expectation based on the
energetic analysis. Detailed inspection of the last three
columns in Table 4 shows that the free-energy changes of
electron transfer in the third reaction step for the nine
reactions are all very large negative values (À25.7 to
À36.7 kcalmolÀ1), which indicate that the electron transfer
in the third reaction step for the nine reactions would be
diffusion-controlled. The rate-limiting step is therefore either
the initial electron transfer or the second proton transfer. By
comparing the free-energy changes in the first two reaction
steps, it is clear that when the NADH model is BNAH and
CH3OPNAH (group I in Table 4), the driving forces of the
initial electron transfer are larger than the corresponding
driving forces of the proton transfer in the second step. This
suggests that the proton transfer in the second step could be
the rate-limiting step. However, for group II, the driving
forces of the initial electron transfer are quite a bit smaller
than those of the corresponding proton transfer in the second
step; this suggests that the initial electron transfer could be
the rate-limiting step. Though these suggestions have not been
verified by other experimental results, the mechanistic guide-
lines suggested here should be valuable in predicting or
differentiating the mechanistic possibilities of NADH model
reactions.
.
‡
À
of the (C4 H) bonds among the three extensively used
NADH models (BNAH, HEH, and AcrH2) shows that
.
‡
HEH (4.5 kcalmolÀ1) is the strongest acid among the three
.
‡
NADH models. Inspection of the homolytic bond dissoci-
.
‡
À
ation energies of the (C4 H) bonds in BNAH, HEH, and
.
AcrH2 reveals that HEH (31.4 kcalmolÀ1) is the strongest
hydrogen donor, whereas AcrH2 (43.5 kcalmolÀ1) is the
‡
.
‡
weakest hydrogen contributor, which is also in line with many
.
‡
experimental results. In fact, the radical cation AcrH2 can be
directly detected and characterized by EPR spectroscopy,[29]
.
.
‡
‡
but so far the radical cations BNAH and HEH cannot.
Driving-force analysis on each mechanistic step of hydride
.
‡
transfer from the NADH models oxidized by TMPA
:
Extensive chemical mimic experimental results show that
‡
the 4-hydrogen was transferred by the eÀ
H
eÀ mechanism
in the reactions of NADH models with one-electron oxidants.
However, which of the three elementary steps is the rate-
limiting step is still a disputed question. Some reports insist
that the initial electron transfer should be the rate-limiting
step according to the observation of no remarkable kinetic
isotopic effect,[4d, 30] others argue that the second proton
transfer should be the rate-limiting step on the basis of general
base catalysis.[5f, 31] Clearly, it is necessary to invoke thermo-
dynamic driving force analysis on each mechanistic step to aid
solution of the mechanistic problem. Regarding the reaction
.
‡
mechanism of the nine NADH models with TMPA as
Evaluation of the effective charge distribution on the pyridine
ring in the models of NADH,NADH ,NAD ,and NAD
.
.
mentioned above, the 4-hydrogen transfer from the NADH
‡
‡
:
‡
models should proceed in an eÀ
H
eÀ three-step se-
.
.
‡
NADH and NAD are important intermediates formed
during the course of NADH reduction initiated by one-
electron transfer; however, very little information about the
structure and the charge distribution is available. In order to
elucidate the structure and the charge distribution on the
pyridine ring, the dependencies of DHhet(GPNAH),
quence.[15] However, it is difficult to determine which is the
rate-limiting step. In order to solve this problem, the driving
force for each elementary step in Scheme 2 has been
calculated according to the thermodynamic data obtained in
this work; the details are listed in Table 4.
.
.
‡
‡
DHhomo(GPNAH ), and DHhet(GPNAH ) on the substituent
constant (s) were examined (see Figure 3). This exhibits three
The most eye-catching feature of the data in Table 4 is that the
energy changes of all the elementary steps are either
exothermic or only slightly endothermic, indicating that
Table 4. Energetics (kcalmolÀ1) of each mechanistic step for the reactions
.
‡
of TMPA with the NADH models in acetonitrile shown in Scheme 2.
‡
Models
eÀ
H
eÀ
‡
DG(eT)[a]
DH(H T)[b]
DG'(eT)[c]
Group I
BNAH
CH3OPNAH
Group II
CH3PNAH
PNAH
ClPNAH
BrPNAH
CF3PNAH
HEH
MeHEH
AcrH2
À 0.9
À 0.4
À 0.4
À 36.7
À 35.6
0.6
0.4
1.6
2.3
2.4
4.0
4.9
4.7
4.5
À 0.3
À 1.5
À 2.3
À 2.3
À 3.7
À 4.6
À 5.0
À 0.1
À 34.7
À 33.4
À 31.9
À 31.8
À 30.0
À 33.2
À 31.9
À 25.7
.
‡
!
&
Figure 3. Hammett plots of DHhet(NADH) ( ), DHhomo(NADH ) ( ),
.
‡
~
^
DHhet(NADH ) ( ), DG(eT) ( ), and DG'(eT) ( ) versus s. NADH ˆ
E
GPNAH (G ˆ 4-OCH3, 4-CH3, 4-H, 4-Cl, 4-Br, 4-CF3). DHhet(NADH),
.
.
‡
‡
[a] Calculated according to the equation of DG(eT) ˆ À23.06[E1/2
DHhet(NADH ), and DHhomo(NADH ) are listed in Table 2; DG(eT) and
DG'(eT) were calculated according to the equations DG(eT) ˆ À23.06
.
‡
(TMPA ) À Eox(NADH)]. [b] Calculated from the equation of
.
.
.
‡
‡
‡
‡
‡
DH(H T) ˆ DHhet(NADH ) À DHa(TMPAH ). [c] Calculated according
[E1/2(TMPA ) À Eox(GPNAH)] and DG'(eT) ˆ À23.06[E1/2(TMPA ) À
.
‡
‡
‡
to the equation of DG'(eT) ˆ À23.06[E1/2(TMPA ) À Ered(NAD )].
Ered(GPNA )], respectively.
876
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0876 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 4

NADH Models
871 880
.
‡
lines with slopes of 7.04 Æ 0.13 (equivalent to 1 of À1.235 for
effective charge of 0.741 on the N1 atom in GPNAH
[32]
À
the heterolytic C4 H bond scission reaction), 1.82 Æ 0.28
ascertained above. In order to further support the evaluations
made above, the Hammett-type linear free-energy correla-
tions of DG(eT) and DG'(eT) against the substituent constant
(s) were also performed (see Figure 3). The line slopes are
5.20 Æ 0.34 and 6.91 Æ 0.24 for DG(eT) and DG'(eT), respec-
tively, which is equivalent to the positive charge increases of
0.739 Æ 0.048 and 0.982 Æ 0.034 on the N1 atom on going from
.
‡
À
(equivalent to 1 of À0.319for the homolytic (C4 H) bond
scission reaction), and À5.11 Æ 0.41 (equivalent to the
.
‡
À
1 of 0.896 for the heterolytic (C4 H) bond scission
reaction). The positive slopes for DHhet(GPNAH) and
DHhomo(GPNAH ) against the substituent constant s reflect
.
‡
that the positive charge increases at the N1 atom on the
.
.
‡
‡
‡
pyridine ring in going from GPNAH to GPNA , and from
GPNAH to GPNAH and from GPNA to GPNA , respec-
tively (Scheme 6). Clearly, the effective charge changes of
0.739and 0.982 are close to the differences of the correspond-
ing effective charges on the N1 atom between PNAH (0.000)
.
‡
‡
GPNAH to GPNA , respectively. The negative slope for
.
‡
DHhet(GPNAH ) reflects the positive charge decrease on the
.
.
‡
N1 atom in going from GPNAH to GPNA . The magnitude
of the line slope values is a measure of the effective charge
change on the N1 atom. Comparing the magnitudes of the
.
.
‡
and GPNAH (‡0.741), and between GPNA (0.015) and
‡
GPNA (‡1.000), respectively. The deviations are all within
.
‡
three slope values for DHhet(GPNAH), DHhomo(GPNAH ),
the experimental errors, suggesting that the evaluations of the
.
.
.
‡
‡
and DHhet(GPNAH ) shows that the absolute values of the
three slopes are jslope for DHhet(GPNAH) j>j slope for
effective charge on the N1 in the GPNAH and GPNA
derived from the slope values of the Hammett-type free-
energy linear plots are reasonable and reliable. In fact, similar
evaluations of effective charge distribution on an active atom
in a molecule or a transition-state structure by using the
Hammett linear free-energy relationship have been exten-
sively reported in the literature.[33] According to the evalua-
tion of the effective charge (‡0.741) on the 1-position
.
.
‡
‡
DHhet(GPNAH ) j>j slope for DHhomo(GPNAH ) j , indicat-
ing that the effective change in charge on the N1 atom would
decrease in the following order: GPNAH (heterolytic) >
.
.
‡
‡
GPNAH (heterolytic) > GPNAH (homolytic). In order
to quantitatively evaluate the relative effective charge
À
changes on the N1 atom for the three different C4 H bond
.
‡
dissociation processes, we arbitrarily defined that the effective
charge on the N1 atom in GPNAH is zero, and the effective
nitrogen in GPNAH , it is conceived that the center of the
radical spin in GPNAH should be located at the carbon
atoms on the pyridine ring. The resonance structures of
.
‡
‡
charge on the N1 atom in the GPNA is a positive one. This is
.
‡
from our basic knowledge of the electronic structures of the
neutral GPNAH and the corresponding quaternary ammo-
GPNAH (Scheme 7) show that the 3- or 5-position carbon
on the pyridine ring should be the spin center, but only the
latter was supported by the formation of the [2‡3] cyclo-
adduct of BNAH with p-benzoquinone initiated by single-
‡
nium salts GPNA ; these indicate that the slope value of 7.04
À
for the heterolytic C4 H bond scission process is equivalent
.
to a positive charge increase of one unit on the N1 atom in
electron transfer.[34] Concerning the pyridinyl radical GPNA ,
‡
going from GPNAH to GPNA . According to this definition,
the evaluation of 0.015 positive effective charge on the N1
atom suggests that the lone pair of electrons remains on N1 in
GPNA , indicating that the single electron should appear
mainly on the 4-position carbon atom (see the resonance
structures for GPNA in Scheme 7). That is, radical GPNA
.
‡
it is clear that the slope of 1.82 Æ 0.28 for DHhomo(GPNAH )
is equivalent to the positive charge increase of 0.259 Æ 0.039
.
.
‡
‡
on the N1 atom on going from GPNAH to GPNA . Since
.
.
‡
the effective charge on the N1 atom in GPNA has been
defined as a positive one, the effective charge on the N1 atom
should be essentially due to a carbon radical rather than a
hetero-atom-centered radical; this has been supported by
many experimental results.[35] According to the effects of the
.
‡
in GPNAH should be 0.741 Æ 0.039(Scheme 6). Similarly,
.
‡
the slope of À5.11 Æ 0.41 for DHhet(GPNAH ) should be
À
equivalent to the negative effective charge increase of
para-substituents (G) on the BDEs of the C4 H bond in
.
‡
À0.726 Æ 0.058 on the N1 atom on going from GPNAH to
GPNAH (see the third column in Table 2), it is interesting to
.
.
.
GPNA . This suggests that the N1 atom in GPNA should
find that the pyridinyl radical of GPNA is destabilized not
possess an effective charge of 0.015 Æ 0.058 according to the
only by the remote electron-donating substituents (EDG), but
also by the remote electron-
withdrawing
substituents
(EWG); this differs substantial-
ly from the previously reported
behaviors of carbon radicals in
the literature in which both the
EDGs and EWGs were found
to be radical-stabilizing (i.e.,
the so-called ™Class S∫ radi-
cal).[36] This also differs from
the observations on various
kinds of heteroatom-centered
radicals so far investigated,
which show ™O (opposite)-
type∫ substituent behavior, that
is, EDG is radical-stabilizing
.
.
‡
Scheme 6. Relative effective charge on the 1-position nitrogen atom (N-1) in GPNAH, GPNAH , GPNA , and
GPNA in acetonitrile.
‡
Chem. Eur. J. 2003, 9, No. 4
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0877 $ 20.00+.50/0
877

X.-Q. Zhu, J.-P. Cheng et al.
FULL PAPER
5) Hammett-type linear free-
energy analyses show that
the N1 atom on the pyridine
ring carries the relative ef-
fective charges of 0.000,
‡0.741, 0.015, and ‡1.000
.
‡
in GPNAH, GPNAH
,
.
‡
GPNA , and GPNA , re-
spectively, which, to our best
knowledge, appears to be
the first work to quantita-
tively and experimentally
evaluate the relative effec-
tive charge distribution on
the N1 atom in the various
forms of NADH model
GPNAH.
.
.
‡
Scheme 7. Resonance structures of GPNAH and GPNA .
Experimental Section
but EWG is radical-destabilizing (so-called ™Class O∫ radi-
cal).[37] On the basis of the earlier so-called class S/O radical
Materials: 1-Benzyl-1,4-dihydronicotinamide (BNAH) was prepared ac-
cording to the literature method.[38] Hantzsch 1,4-dihydropyridine was
prepared by depositing the mixture of aldehyde, ammonia, and ethyl
acetoacetate in methanol. N-Methylated Hantzsch 1,4-dihydropyridines
were prepared according to the literature.[39] 10-Methyl-9,10-dihydroacri-
dine (AcrH2) was prepared according to the literature.[40] 1-(4-Substituted
phenyl)-1,4-dihydronicotinamides were prepared according to the follow-
ing general methods: the appropriate aniline (1 mmol) dissolved in dry
methanol (15 mL) was added to a solution of 1-(2,4-dinitrophenyl)nico-
tinamide chloride (1 mmol) in methanol (100 mL). The resulting red
solution was then heated gently overnight or until the red color faded to
yellow, indicating the formation of 2,4-dinitroaniline. The solution was
cooled, and the precipitated side product was removed by filtration. The
filtrate was then evaporated in vacuum, and the residue was dissolved in
H2O (100 mL). The aqueous phase was then exhaustively washed with
diethyl ether. The water layer was then evaporated under reduced pressure
to give a crude product, which was recrystallized from methanol/diethyl
ether. Reduction of the pyridinium salt was performed in aqueous basic
sodium dithionite to give the appropriate reduced pyridine derivatives.
N,N,N',N'-tetramethyl-p-phenylenediamine radical cation perchlorate
.
definition of Walter,[37a] the pyridinyl radical GPNA should
be a class counter-S type, which suggests that not all carbon
radicals have S-type behavior.
Conclusions
À
Four sets of C4 H bond dissociation energies for ten NADH
models and their radical cations were determined in acetoni-
trile by a new, developing method. It can be concluded that:
À
1) The heterolytc and homolytic C4 H bond scission energies
for the ten NADH models are significantly smaller than
3
À
those of general sp C H bonds; this suggests that the
NADH models should be not only good hydride donors
but also good hydrogen-atom contributors in solution.
2) When the NADH model is 1-benzyl-1,4-dihydronicotina-
.
(TMPA ClO4À) was obtained from the oxidation of N,N,N',N'-tetrameth-
yl-p-phenylenediamine by Br2 in aqueous NaClO4.[41] Reagent grade
acetonitrile was distilled from P2O5 being passed through a column of
active neutral alumina to remove water and protic impurities.
‡
À
mide (BNAH), the heterolysis of the C4 H bond to
release a hydride is more likely to take place than the
corresponding homolysis to remove a hydrogen atom in
solution; when the NADH model is 9,10-dihydroacridine
Titrated calibration experiments: The titration experiments were per-
formed on a CSC4200 isothermal titration calorimeter in acetonitrile at
258C. Prior to use, the instrument was calibrated against an internal heat
pulse, and the functional response was verified by determination of the heat
of dilution of a concentrated sucrose solution.[42] Data points were collected
every 2 s. The reaction heat was determined following 10 automatic
injections from a 250 mL injection syringe (containing 2.01mm of NADH
model) into the reaction cell (1.00 mL) containing 10.87mm N,N,N',N'-
tetramethyl-p-phenylenediamine radical cation perchlorate. Injection vol-
umes (10 mL) were delivered at 0.5 s time intervals with 500 s between
every two injections. The calibration graph is shown in Figure 2. The
reaction heat was obtained by area integration of each peak except for the
first.
À
(AcrH2), the homolysis of the C9 H bond to remove a
hydrogen atom is more likely to take place than the
À
corresponding heterolysis of the C9 H bond to release a
hydride in solution. However, when the NADH model is
Hantzsch 1,4-dihydropyridine (HEH), the hydride and
hydrogen-atom transfers from HEH occur with almost
equal probability.
.
‡
3) Since the DHhet(NADH ) for the ten NADH models are
.
‡
much smaller than the corresponding DHhomo(NADH ), it
‡
is conceived that the eÀ
H
eÀ sequence of hydride
Electrochemical experiments: All electrochemical experiments were
carried out by CV (sweep rate, 100 mVsÀ1) with a BAS-100B electro-
chemical apparatus in acetonitrile under an argon atmosphere as described
previously.[2i] Bu4NPF6 (0.1m) was employed as the supporting electrolyte.
A standard three-electrode assembly consisted of a glassy carbon disk as
the working electronic, a platinum wire as counter electrode, and 0.1m
AgNO3/Ag as reference. All sample solutions were 1.0mm. The ferroce-
nium/ferrocene redox couple (Fc‡/0) was taken as an internal standard.
Reproducibility is generally smaller than 5 mV.
transfer should be the most likely among all the possible
multi-step hydride transfer mechanisms of NADH and its
models.
.
4) The pyridinyl radical GPNA should be due to a counter-S
type carbon radical rather than a nitrogen radical and the
lone single electron mainly appears at the C4-position on
the pyridine ring.
878
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0878 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 4

NADH Models
871 880
6711; c) G. Kiss, K. Zhang, S. L. Mukerjee, C. D. Hoff, G. C. Roper, J.
Am. Chem. Soc. 1990, 112, 5657.
Acknowledgement
[14] a) T. Yao, S. Musha, M. Munemori, Chem. Lett. 1974, 939; b) E. P.
Platonova, V. G. Kashechko, V. Y. Atamanyuk, V. D. Pokhodenko,
Elektrokhimiya 1980, 16, 674.
This work was supported by the National Outstanding Youth Fund (Grant
No. 20125206) and the Natural Science Foundation of China (Grant Nos.
20272027 and 29972028), which are gratefully acknowledged.
[15] Z.-L. Guo, Y.-C. Liu, Liu, Chin. Sci. Bull. 1988, 18, 1388.
[16] In Equations (7) (10), free-energy change DG was used to replace
the enthalpy change DH for the electron transfer processes. This
treatment is reasonable because the entropies associated with electron
transfer have been verified by Arnett×s work (see E. M. Arnett, K.
Amarnath, N. G. Harvey, J. Cheng, J. Am. Chem. Soc. 1990, 112, 344).
The main reason is that the molecular structures of the products and
the corresponding reactants are similar for a single electron transfer.
[1] a) Y. Murakami, J. Kikuchi, Y. Hisaeda, O. Hayashida, Chem. Rev.
1996, 96, 721; b) M. E. Brewster, A. Simay, K. Czako, D. Winwood, H.
Farag, N. Bodor, J. Org. Chem. 1989, 54, 3721; c) F. Friedlos, R. J.
Knox, Biochem. Pharmacol. 1992, 44, 631.
[2] a) X.-Q. Zhu, Y. Liu, B. J. Zhao, J.-P. Cheng, J. Org. Chem. 2001, 66,
370; b) X.-Q. Zhu, H.-Y. Wang, J.-S. Wang, Y.-C. Liu, J. Org. Chem.
2001, 66, 344; c) X.-Q. Zhu, B.-J. Zhao, J.-P. Cheng, J. Org. Chem. 2000,
65, 8158; d) X.-Q. Zhu, Y.-C. Liu, J.-P. Cheng, J. Org. Chem. 1999, 64,
8980; e) X.-Q. Zhu, Y.-C. Liu, H.-Y. Wang, W. Wang, J. Org. Chem.
1999, 64, 8982; f) X.-Q. Zhu, Y.-C. Liu, J. Org. Chem. 1998, 63, 2786;
g) X.-Q. Zhu, H.-L. Zou, P.-W. Yuan, L. Cao, J.-P. Cheng, J. Chem.
Soc. Perkin Trans. 2 2000, 1857; h) X.-Q. Zhu, Y. Liu, B.-J. Zhao, J.-P.
Cheng, J. Org. Chem. 2001, 66, 370.
[3] a) T. Bally, P. Bednarek, J. Phys. Chem. A 2000, 104, 724; b) P.
Bednarek, T. Bally, J. Phys. Chem. A 2000, 104, 718; c) N. Kanomata,
T. Nakata, J. Am. Chem. Soc. 2000, 122, 4563; d) A. Marcinek, J.
Adamus, K. Huben, J. Gbicki, T. J. Bartczak, P. Bednarek, T. Bally, J.
Am. Chem. Soc. 2000, 122, 437; e) S. Fukuzumi, H. Imahori, K.
Okamoto, H. Yamada, M. Fujitsuka, O. Ito, D. M. Guldi, J. Phys.
Chem. A 2002, 106, 1903; f) I. Nakanishi, S. Fukuzumi, T. Konishi, K.
Ohkubo, M. Fujitsuka, O. Ito, N. Miyata, J. Phys. Chem. B 2002, 106,
2372; g) H. C. Lo, C. Leiva, O. Buriez, J. B. Kerr, M. M. Olmstead,
R. H. Fish, Inorg. Chem. 2001, 40, 6705; h) J. Luo, T. C. Bruice, J. Am.
Chem. Soc. 2001, 123, 11952; i) S. Fukuzumi, Y. Fujii, T. Suenobu, J.
Am. Chem. Soc. 2001, 123, 10191; j) B. Nidetzky, M. Klimacek, P.
Mayr, Biochem. 2001, 40, 10371.
[4] a) M. F. Powell, T. C. Bruice, J. Am. Chem. Soc. 1982, 104, 5834;
b) M. F. Powell, T. C. Bruice, J. Am. Chem. Soc. 1983, 105, 1014;
c) J. W. Verhoeven, W. van Gerresheim, F. M. Martiens, S. M. van der
Kerk, Tetrahedron 1986, 42, 975; d) L. L. Miller, J. R. Valentine, J. Am.
Chem. Soc. 1988, 110, 3982; e) D. Ostovic, R. M. G. Roberts, M. M.
Kreevoy, J. Am. Chem. Soc. 1983, 105, 7629.
[5] a) F. Powell, S. C. Wu, T. C. Bruice, J. Am. Chem. Soc. 1984, 106, 3850;
b) B. W. Carlson, L. L. Miller, J. Am. Chem. Soc. 1983, 105, 7453; c) S.
Fukuzumi, S. Koumitsu, K. Hironaka, T. Tanaka, J. Am. Chem. Soc.
1987, 109, 305; d) A. Sinha, T. C. Bruice, J. Am. Chem. Soc. 1984, 106,
7291; e) C. C. Lai, A. K. Colter, J. Chem. Soc. Chem. Commun. 1980,
1115.
[6] a) L. L. Miller, J. R. Valentine, J. Am. Chem. Soc. 1988, 110, 3982, and
references therein; b) B. W. Carlson, L. L. Miller, J. Am. Chem. Soc.
1985, 107, 479; c) S. Fukuzumi, S. Koumitsu, K. Hironaka, T. Tanaka, J.
Am. Chem. Soc. 1987, 109, 305, and references therein; d) S.
Fukuznmi, N. Nishizawa, T. Tanaka, J. Org. Chem. 1984, 49, 3571;
e) C. A. Coleman, J. G. Rose, C. J. Murray, J. Am. Chem. Soc. 1992,
114, 9755.
‡
[17] In these estimates, irreversible redox potentials of NADH and NAD
models in Table 1 were used because only irreversible redox potentials
‡
of NADH and NAD models are available by CV. The uncertainties
were estimated generally not to be larger than 1.0 kcalmolÀ1 according
to the work of Fukuzumi, in which he obtained the standard potentials
of NADH model BNAH (a typical model of NADH) in acetonitrile
(0.60 V) by using the fluorescence quenching method; this value is
only larger than the irreversible CV peak value of BNAH (0.57 V) by
30 mV (ꢀ0.7 kcalmolÀ1, ref: S. Fukuzumi, K. Hironaka, N. Nishizawa,
T. Tanaka, Bull. Chem. Soc. Jpn. 1983, 56, 2220).
[18] J.-P. Cheng, K. L. Handoo, V. D. Parker, J. Am. Chem. Soc. 1993, 115,
2655.
[19] C. J. Curtis, A. Miedaner, W. Ellis, D. L. DuBois, J. Am. Chem. Soc.
2002, 124, 1918.
[20] a) A. van Laar, H. J. van Ramesdonk, J. W. Verhoeven, Recl. Trav.
Chim. Pays-Bas 1983, 102, 157; b) W. van Gerresheim, J. W. Verho-
even, Recl. Trav. Chim. Pays-Bas 1983, 102, 339.
À
[21] Unlike heterolytic C H bond scission energies, generally the BDE of
À
the C H bond is not very dependent on the nature of solvents.
[22] X.-M. Zhang, F. G. Bordwell, J. Am. Chem. Soc. 1992, 114, 9787.
[23] V. D. Parker, J. Am. Chem. Soc. 1992, 114, 7458.
[24] D. D. Watner, V. D. Parker, Acc. Chem. Res. 1993, 26, 287.
[25] D. D. Tanner, H. K. Singh, A. Kharrat, A. R. Stein, J. Org. Chem. 1987,
52, 2142.
[26] E. S. Huyser, J. A. K. Harmony, F. L. McMillian, J. Am. Chem. Soc.
1972, 94, 3176.
[27] a) S. Fukuzumi, T. Kitano, M. Ishikawa, J. Am. Chem. Soc. 1990, 112,
5631; b) P. J. Elving in Topics in Bioelectrochemistry and Bioenergetics,
Vol 1 (Ed. G. Milazzo), Wiley, New York, 1976, p. 179; c) V. Carelli, F.
Liberatore, A. Casini, R. Mondelli, A. Arnone, I. Carelli, G. Retilio, I.
Mavelli, Bioorg. Chem. 1980, 9, 342; d) S. Fukuzumi, T. Kitano, K.
Mochida, J. Am. Chem. Soc. 1990, 122, 3246; e) X.-Q. Zhu, Y.-C. Liu,
J. Li, H. Y. Wang, J. Chem. Soc. Perkin Trans. 2 1997, 2191; f) E. S.
Huyser, J. A. K. Harmony, F. L. McMillian, J. Am. Chem. Soc. 1972,
94, 3176.
[28] a) Y.-C. Liu, B. Li, Q. X. Guo, Tetrahedron Lett. 1994, 35, 8429; b) Y.-
C. Liu, B. Li, Q. X. Guo, Tetrahedron 1995, 51, 9671; c) L. L. Miller,
J. R. Valentine, J. Am. Chem. Soc. 1988, 110, 3982.
[29] S. Fukuzumi, T. Kitano, Chem. Lett. 1990, 1275.
[30] B. W. Carlson, L. L. Miller, P. Neta, J. Grodkowski, J. Am. Chem. Soc.
1984, 106, 7233.
[7] a) P. Varnai, A. Warshel, J. Am. Chem. Soc. 2000, 122, 3849; b) A.
Yadav, R. M. Jackson, J. J. Holbrook, A. Warshel, J. Am. Chem. Soc.
1991, 113, 4800.
[31] A. Sinha, T. C. Bruice, J. Am. Chem. Soc. 1984, 106, 7291.
[32] According to the Hammett equation logK ˆ 1s and DH ꢀÀ RT logK,
the line slope of DH against s should be approximately equal to
À5.7 1.
[8] a) A. Anne, P. Hapiot, J. Moiroux, P. Neta, J.-M. Saveant, J. Am.
Chem. Soc. 1992, 114, 4694; b) A. Anne, S. Fraoua, V. Grass, J.
Moiroux, J.-M. Saveant, J. Am. Chem. Soc. 1998, 120, 29 51; c) A.
Anne, S. Fraoua, J. Moiroux, J.-M. Saveant, J. Am. Chem. Soc. 1996,
118, 3938; d) P. Hapiot, J. Moiroux, J.-M. Saveant, J. Am. Chem. Soc.
1990, 112, 1337; e) A. Anne, J. Moiroux, J. Org. Chem. 1990, 55, 4608.
[9] J. Klippenstein, P. Arya, D. D. Wayner, J. Org. Chem. 1991, 56, 6736.
[10] K. L. Handoo, J.-P. Cheng, V. D. Parker, J. Am. Chem. Soc. 1993, 115,
5067.
[11] W. S. Matthews, J. E. Bares, J. E. Bartmess, F. G. Bordwell, F. J.
Cornforth, G. E. Drucker, Z. Margonlin, R. J. McCallum, G. J.
McCollum, N. R. Vanier, J. Am. Chem. Soc. 1975, 97, 7006.
[12] E. Arnett, K. Amarnath, N. G. Harvey, J.-P. Cheng, J. Am. Chem. Soc.
1990, 112, 344.
[33] a) M. Page, A. Williams, Organicand Bio-organicMehcanism
,
Longman, 1997, Chapter 3, pp. 52 79; b) A. Williams, Acc. Chem.
Res. 1984, 17, 425; c) A. Williams, Acc. Chem. Res. 1989, 22, 387; d) A.
Williams, Adv. Phys. Org. Chem. 1991, 27, 1; e) A. Williams, J. Am.
Chem. Soc. 1985, 107, 6335; f) X.-Q. Zhu, Q. Li, W.-F. Hao, J.-P. Cheng,
J. Am. Chem. Soc. 2002, 124, 9887.
[34] S. Fukuzumi, Y. Fujii, T. Suenobu, J. Am. Chem. Soc. 2001, 123,
10191.
.
[35] C4 C4 dimers of radical BNA and its derivatives have been obtained
.
from several experiments, which indicate that BNA should result
from the C4 radical rather than the N1 radical.
[36] a) D. F. McMillen, D. M. Golden, Annu. Rev. Phys. Chem. 1982, 33,
493; b) J. M. Dust, D. R. Arnold, J. Am. Chem. Soc. 1983, 105, 1221,
and references therein.
[13] a) D. E. Berning, B. C. Noll, D. L. DuBois, J. Am. Chem. Soc. 1999,
121, 11432; b) M. Tilset, V. D. Parker, J. Am. Chem. Soc. 1989, 111,
Chem. Eur. J. 2003, 9, No. 4
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0879 $ 20.00+.50/0
879

X.-Q. Zhu, J.-P. Cheng et al.
FULL PAPER
[37] a) R. I. Walter, J. Am. Chem. Soc. 1966, 88, 1923; b) F. G. Bordwell,
J.-P. Cheng, J. Am. Chem. Soc. 1991, 113, 1736, and references therein;
c) J.-P. Cheng, Y. Lu, B. Liu, Y. Zhao, D. Wang, Y. Sun, J. Mi, Sci.
China Ser. B 1998, 41, 215, and references therein.
[38] A. C. Anderson, G. Berkelnammer, J. Am. Chem. Soc. 1958, 80,
992.
[40] R. M. G. Roberts, D. Ostovic, M. M. Kreevoy, Faraday Discuss. Chem.
Soc. 1982, 74, 257.
[41] L. Michaelis, S. Granick, J. Am. Chem. Soc. 1943, 65, 1747.
[42] F. T. Gucker, H. B. Pickard, R. W. Planck, J. Am. Chem. Soc. 1939, 61,
459.
[39] J. L. Kutz, R. Hutton, F. H. Westheimer, J. Am. Chem. Soc. 1961, 83,
584.
Received: June 17, 2002
Revised: November 7, 2002 [F4188]
880
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0904-0880 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 4