Effects of substituents on aryl groups during the reaction of triarylphosphine radical cation and oxygen

Article abstract of DOI:10.1039/b606857j

In a previous report (S. Yasui, S. Tojo and T. Majima, J. Org. Chem., 2005, 70, 1276), we presented the results from the laser flash photolysis (LFP) and product analysis of the 9,10-dicyanoanthracene (DCA)-photosensitized oxidation of triarylphosphine (Ar₃P) in acetonitrile under air, which showed that the photoreaction results in the oxidation of Ar₃P to give the corresponding phosphine oxide (Ar₃PO) in a nearly quantitative yield, and that the reaction is initiated by the electron transfer (ET) from Ar ₃P to DCA in the singlet excited state (¹DCA*), producing the triarylphosphine radical cation Ar₃P⁺. This radical cation decays through radical coupling with O₂ to afford the peroxy radical cation Ar₃P⁺-O-O, which we proposed to be the intermediate leading to the product Ar₃PO. We now examined this photoreaction in more detail using ten kinds of Ar₃P with various electronic and steric characteristics. The decay rate of Ar₃P ⁺ measured by the LFP was only slightly affected by the substituents on the aryl groups of Ar₃P. During the photolysis of trimesitylphosphine (Mes₃P), the peroxy radical cation intermediate (Mes₃P⁺-O-O) had a lifetime long enough to be spectrophotometrically detected. The quantum yields of Ar₃PO increased with either electron-withdrawing or -releasing substituents on the aryl groups, suggesting that a radical center is developed on the phosphorus atom during the step when the quantum yield is determined. In addition, the o-methyl substituents in Ar₃P decreased the quantum yield. These results clearly indicated that Ar₃P⁺-O-O undergoes radical attack upon the parent phosphine Ar₃P that eventually produces the final product, Ar₃PO. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b606857j

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Effects of substituents on aryl groups during the reaction of triarylphosphine
radical cation and oxygen
a
b
b
Shinro Yasui,* Sachiko Tojo and Tetsuro Majima*
Received 15th May 2006, Accepted 14th June 2006
First published as an Advance Article on the web 30th June 2006
DOI: 10.1039/b606857j
In a previous report (S. Yasui, S. Tojo and T. Majima, J. Org. Chem., 2005, 70, 1276), we presented the
results from the laser flash photolysis (LFP) and product analysis of the 9,10-dicyanoanthracene
(
DCA)-photosensitized oxidation of triarylphosphine (Ar
that the photoreaction results in the oxidation of Ar P to give the corresponding phosphine oxide
Ar
P=O) in a nearly quantitative yield, and that the reaction is initiated by the electron transfer (ET)
from Ar
3
P) in acetonitrile under air, which showed
3
(
3
1
3
P to DCA in the singlet excited state ( DCA*), producing the triarylphosphine radical cation
P . This radical cation decays through radical coupling with O to afford the peroxy radical cation
P –O–O , which we proposed to be the intermediate leading to the product Ar
P=O. We now
P with various electronic and steric
measured by the LFP was only slightly affected by the
P. During the photolysis of trimesitylphosphine (Mes P), the
•
+
Ar
Ar
3
2
+
•
3
3
examined this photoreaction in more detail using ten kinds of Ar
3
•
+
characteristics. The decay rate of Ar
substituents on the aryl groups of Ar
peroxy radical cation intermediate (Mes
spectrophotometrically detected. The quantum yields of Ar
electron-withdrawing or -releasing substituents on the aryl groups, suggesting that a radical center is
developed on the phosphorus atom during the step when the quantum yield is determined. In addition,
3
P
3
3
+
•
3
P –O–O ) had a lifetime long enough to be
P=O increased with either
3
the o-methyl substituents in Ar P decreased the quantum yield. These results clearly indicated that
3
+
•
Ar
3
P –O–O undergoes radical attack upon the parent phosphine Ar
3
P that eventually produces the
final product, Ar
3
P=O.
7
undergoes an ionic reaction with water to produce Ar
3
P O. On
=
Introduction
the other hand, the photoreaction under aerobic conditions in
Trivalent phosphorus compounds Z
and are easily converted to the corresponding radical cations Z
when treated with a one-electron acceptor.
can act either as a cation or as a radical, but this class of radical
cations is in fact quite labile toward a nucleophile mainly showing
the reactivity as a cation. Thus, Z
reaction with a small amount of alcohol or water in the solvent to
eventually afford the corresponding pentavalent oxo-compounds
3
P are good electron donors
the absence of nucleophiles affords Ar
3
P=O through the radical
•
+
3
P
P
coupling of Ar
P
•+
with molecular oxygen (O ) in the solvent.
11
3
2
1–10
•+
In principle, Z
3
Thus, we performed the product analysis and the laser flash
photolysis (LFP) on the DCA-photosensitized oxidation of Ar
under aerobic conditions to show the mechanism which includes
3
P
•
+
•+
3
P
easily undergoes an ionic
the radical coupling of Ar
P
with O
. This was the first finding
3
2
•+
that the trivalent phosphorus radical cation Z
radical coupling with O
suggests the peroxy radical cation Ar
but it is not clear how this intermediate produces the observed
product Ar
P=O. To determine this, it is quite informative to
examine the effect of the substituents of aryl groups in Ar P on
the quantum yield of Ar
P=O.
Now, with ten electronically and sterically tuned Ar
3
P
undergoes
2
in the ground state. This mechanism
•
+
Z
3
P=O. One of the few examples of the radical reactions by Z
3
P
+
•
3
P –O–O as an intermediate,
•
+
•
is the radical coupling of Z
3
P
with the aryl radical Ar taking
place in competition with the ionic reaction with an alcohol upon
3
1
treatment of Z P with a diazonium salt in an alcoholic solvent.
3
3
We have found that the 9,10-dicyanoanthracene (DCA)-
photosensitized oxidation of triarylphosphine (Ar P) quantita-
tively affords the corresponding triarylphosphine oxide (Ar
P=O),
in which the reaction is initiated by the electron transfer (ET) from
3
3
P (1)
3
3
available, we analyzed the DCA-photosensitized oxidation of 1
based on the results from LFP and product analysis (Scheme 1).
Our major findings in the present study are that (1) both
electron-withdrawing and -releasing substituents on the aryl
1
Ar
3
P to DCA in the singlet excited state ( DCA*) to generate the
•
+ 7,11
triarylphosphine radical cation Ar
mechanism depends on the reaction conditions. When the pho-
toreaction is carried out in aqueous acetonitrile (MeCN) under
3
P . Quite interestingly, the
ligands increased the relative quantum yield of Ar
3
P=O, and
•+
that (2) on the other hand, the decay rate of Ar P was not
3
•
+
an argon atmosphere, the initially generated radical cation Ar P
3
affected by the substituents in this way. These findings along
with other evidence strongly suggest that the peroxy radical
a
Laboratory of Biology and Chemistry, Tezukayama University, 3-1 Gakuen-
Minami, Nara 631-8585, Japan. E-mail: yasui@tezukayama-u.ac.jp
+
•
cation Ar
3
P –O–O undergoes radical attack on 1 to produce
+
•
a dimeric radical cation Ar
3
P –O–O–P Ar
3
. Our mechanism is
-mediated oxidation of
), for which the ET from Ar P to
b
The Institute of Scientific and Industrial Research, Osaka University,
different from that proposed for the O
Ar P in dichloromethane (CH Cl
2
8
-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan. E-mail: majima@
sanken.osaka-u.ac.jp
3
2
2
3
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2969–2973 | 2969

View Article Online
under an argon atmosphere, 1• spontaneously decayed at a slower
+
rate. Table 1 also summarizes the pseudo-first order rate constants
•
+
k
decay(Ar) for the decay under an argon atmosphere. Since 1
11
reacts with O as reported in a previous article, the second-order
2
•
+
rate constant kO2 for the reaction of 1 with O
2
was calculated
according to eqn (1).
k
O2 = (kdecay(air) − kdecay(Ar))/[O
2
]
(1)
The LFP for trimesitylphosphine 1h provided useful informa-
tion with respect to the fate of the transient species during the
photo-oxidation of Ar P. When 1h was likewise subjected to the
3
photolysis under an argon atmosphere, a transient absorption
spectrum with a maximum at 600 nm assignable to the trime-
•
+
sitylphosphine radical cation (1h ) was observed at 100 ns after the
laser flash, which disappeared on the time scale of a few ls (Fig. 1).
−
3
Under air (where [O
2
] = 1.82 × 10 M in MeCN), the transient
•
+
absorption of 1h disappeared on the time scale of 300 ns, and
the absorption with a maximum at 560 nm appeared on the time
scale of 1 ls. Furthermore, the LFP under an O atmosphere gave
2
Scheme 1
the absorption spectrum with a maximum at 560 nm at 100 ns
after the laser flash. The peak at 600 nm was observed only as a
shoulder peak, which disappeared in the time scale of several tens
of ls (Fig. 2). The spectral change seen in Fig. 2 is very similar to
that reported for the reaction of an aromatic olefin radical cation
+
•
Ar
3
P –O–O has been assumed to construct a chain-reaction
12
producing Ar
3
P=O. The difference in the stability of the
intermediate well explains the diversity of the mechanism.
Results and discussion
Laser flash photolysis
When a mixture of 1, biphenyl (BP) and DCA in MeCN was
irradiated with a 355 nm laser, the transient absorption spectrum
•
+
7
•+
characteristic of the radical cation 1 was observed. Clearly, 1
is generated from the ET from 1 to the biphenyl radical cation
(
•
+
BP ). It is well established that the photoinduced ET from BP
to the excited DCA occurs to give the DCA radical anion and
•
+
•+
BP , and the hole transfer from BP to 1 is favorable because 1
has a lower oxidation potential than BP. We have found that the
quantum yield of the product from 1 is one order lower when BP
7,11
is absent.
The absorption of 1•+ disappeared on the basis of the first-order
Fig. 1 Transient absorption spectra recorded at 100 ns (᭹), 1 ls (᭺),
and 5 ls (ꢀ) after a laser flash during the 355 nm LFP of 1h under an
argon atmosphere. Inset: time profile of the transient absorption at 600 nm
•
+
kinetics in air. For each radical cation 1a–f , the disappearance of
the absorption at 525–600 nm was analyzed in order to determine
the decay rate constant kdecay(air) as summarized in Table 1. Even
•
+
showing the decay of 1h .
•
+
a
Table 1 Rate constants for the decay of 1 determined by transient absorption measurements
•
+
b
6
−1
c
5
−1
d
8
−1 −1
e
Radical cation (1 )
k
decay(air) /10 s
k
decay(Ar) /10 s
k
O2 /10 M
s
k
max /nm
•
+
+
+
+
+
1
1
1
1
1
1
1
1
1
1
a
b
1.82
1.57
5.15
2.39
2.30
1.95
1.82
2.57
2.78
1.89
2.19
3.65
6.54
2.39
4.55
2.86
4.35
3.25
4.29
3.12
8.80
6.65
24.7
14.8
10.1
9.12
7.60
12.3
12.9
8.69
525
530
550
600
550
550
550
600
535
535
•
•
c
•
d
•
e
•
+
f
•
•
+
+
g
h
•
+
+
i
•
j
a
3+
−2
c
−1
The LFP of 1 with the third harmonic generation (355 nm, 5 ns) from a Q-switched Nd :YAG laser: [1] = 1.00 × 10 M, [BP] = 1.00 × 10 M and
−
5
b
−3
[
DCA] = 5.00 × 10 M in MeCN. Pseudo-first order rate constant determined in air, [O
2
] = 1.82 × 10 M in MeCN. Pseudo-first order rate constant
d
e
•+
determined under an argon atmosphere. Second order rate constant. Wavelength monitored for the 1 transient absorption.
2
970 | Org. Biomol. Chem., 2006, 4, 2969–2973
This journal is © The Royal Society of Chemistry 2006

View Article Online
Table 2 Relative quantum yield Qrel of the DCA-photosensitized oxida-
a
tion of 1 in air
Rr•
b
Q
Ar
3
P (1)
rel
1
1
1
1
1
1
1
1
1
1
a
b
c
d
e
f
0
1
0.36
0.54
1.26
1.17
d1.17
1.242
1.561
2.803
0.564
2.076
0.989
0.289
1.197
1.367
c
g
h
i
c
c
3.51
0.39
0.39
j
a
−2
−1
−5
[
1] = 1.00 × 10 M, [BP] = 1.00 × 10 M and [DCA] = 5.00 × 10 M;
under aerobic conditions in MeCN; irradiated with visible light (>390 nm)
from a Xe lamp. The sum of the r values of the three substituents. The
Fig. 2 Transient absorption spectra recorded at 100 ns (᭹), 1 ls (᭺),
b
•
•
c
•
•
d
and 5 ls (ꢀ) after a laser flash during the 355 nm LFP of 1h under an O
2
r values are taken from ref. 17. Assuming r (o-Me) = r (p-Me). Not
•
+
atmosphere, showing the reaction of 1h with O
2
. Inset: time profiles of the
available.
transient absorptions at 600 (black) and 550 nm (grey), showing the decay
•
+
+
•
3
of 1h and the formation of the peroxy radical cation (Mes P –OO ).
16
quantum yield of 2. Importantly, both the electron-withdrawing
and -releasing substituents increase the Qrel value.
with O
corresponding peroxy radical cation (R R C –CR R –OO ).
Therefore, the spectral change observed in the LFP for 1h supports
2
, in which the olefin radical cation is converted to the
1
2
+
3
4
•
13,14
•
+
Substituent effect
our proposal that 1 decays through radical coupling with oxygen
to produce the peroxy radical cation (3 in Scheme 2). The second
•
In Fig. 3, the logarithm of Qrel was plotted versus Rr , the sum of
•
+
order rate constant for the bimolecular reaction between 1h and
was also calculated according to eqn (1) as listed in Table 1.
•
the Hammett substituent constants r of the substituents on the
O
2
aryl groups of Ar
3
P (1). The linear correlation with the positive
•
+
This is the first direct observation of the reaction of Ar
to give Ar
3
P
and O
2
•
value of the slope (q = 0.33; correlation coefficient r = 0.97)
+
•
15
3
P OO .
•
was found with the exception of 1e and 1h. The r scale has
been defined for the homolytic cleavage of dibenzylmercury as a
measure of the spin density developing on the benzyl carbon atom
17
in the transition state. That is, the correlation in Fig. 3 strongly
suggests that the spin is developed on the phosphorus atom in
the transition state of the Qrel-determining step. The phosphorus
atom of 1 (or any intermediates) occupies the position equivalent
to that of the benzyl carbon of dibenzylmercury with respect to
•
the aryl moieties. It should be noted that the r scale has been
Scheme 2
obtained by subtracting the polar effects from the total effects of
the substituents for measuring the pure effect used to stabilize a
radical center. The analysis has assumed that the meta-substituents
Product analysis
18
exert only a polar effect on the radical center. In our system, the
reaction of tris(m-tolyl)phosphine (1g) gives Qrel nearly identical
to that for the unsubstituted phosphine (1a), which may indicate
that there is a slight stabilization by the substituents through the
polar effect in all stages during the reaction sequence.
A mixture of 1, DCA and BP in MeCN was irradiated with light
from a xenon lamp (>390 nm with a glass filter) under aerobic
conditions. Theirradiated solution was analyzed byGC, indicating
that 1 was gradually converted to the corresponding phosphine
oxide (2) without other products in an appreciable amount. As
has been previously observed, the plot of the amount of the
conversion of 1 versus the irradiation time was linear in each
experiment. The slopes were sometimes slightly different values in
11
the different experiments even for the same Ar
due to the instability of the instrument and/or subtle differences
in the concentration of the dissolved O . To more reliably perform
the examination of the substituent effect, we prepared sample
solutions of 1a (as a standard) and of the particular Ar P from
3
P. This is probably
2
3
the common mother solution and then the resulting solutions
were subjected to the photoreaction one after the other (see
Experimental section). The value Qrel was obtained by dividing
the slope for the reaction of the particular Ar P by that for 1a as
listed in Table 2. The value Qrel represents the relative value of the
3
•
Fig. 3 Plot of logQrel vs. Rr .
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2969–2973 | 2971

View Article Online
•
The r values of the ortho- and meta-methyl groups are not
and obtained the degree of pyramidalization of these phosphines
•
◦
23
◦
available in the literature. Therefore, we take the r value of
(a ; see Scheme 3). The a -values are almost unchanged upon
introduction of the para- or meta-methyl substituents on the aryl
groups. Meanwhile, the introduction of one ortho-methyl on each
the para-methyl group as that of the ortho-methyl group based
on the reasonable assumption that the ortho-methyl group has
almost the same electronic influence on the phosphorus as the
para-methyl substituent. The Qrel value for the reaction of tris(o-
tolyl)phosphine 1e was plotted versus the Rr of 1f. The point was
significantly deviated downward from the line. Such an inhibitory
effect by the ortho-methyl groups on Qrel could be attributed to
a steric factor. 1h having six ortho-methyl groups produced much
smaller Qrel values than those expected from the electronic effects
of the substituents as shown in Table 2 (not plotted in Fig. 3).
◦
◦
aryl group results in a small but explicit change in a (27.2 for 1a
19
◦
and 25.7 for 1e). That is, the C–P–C angle is flattened, increasing
•
the p-character of the lone pair on the phosphorus atom. As a
result, 1e becomes less reactive toward the radical attack by 3.
This may be the reason why the reaction of 1e produces the lower
Qrel value than expected from the electronic effect of the methyl
substituents (Fig. 3). Such a steric effect is much more significant
with the phosphine having two ortho-methyl groups on each aryl
◦
group; thus, 18.8 for 1h. In fact, the Qrel value for the reaction of
1
h is much smaller than expected based on the electronic effect of
Reaction mechanism
the methyl substituents.
The previous study based on LFP and product analysis has shown
that the photoreaction is initiated by the ET from 1 to DCA in the
1
11
singlet excited state, DCA*. The ET generates the radical cation
•
+
•+
1
, which decays through the reaction of 1 with O
2
to give the
intermediate peroxy radical cation 3 (Scheme 2). Our next task is
to determine the fate of 3 leading to the final product 2.
•
+
It is highly unlikely that 3 undergoes radical coupling with 1 ,
Scheme 3
•
+
because the concentrations of 3 and 1 are so low that there is very
little possibility for both species to encounter each other. Much
more likely, 3 attacks the phosphorus in 1 under our experimental
We are now convinced that the DCA-photosensitized oxidation
of 1 proceeds according to the mechanism shown in Scheme 2.
The intermediate 4 affords the product 2 along with the radical
cation 5. Yet, how 5 yields the final product 2 remains unsolved in
the present study, and will be clarified in due course.
20–22
conditions where 1 exists in a high concentration.
This reaction
affords the dimeric radical cation 4 in which one of the phosphorus
atoms has an unpaired electron on it. The observed substituent
effect supports the participation of 4 in the present photoreaction.
•
Thus, the positive value of q given in Fig. 3 shows clearly that the
Experimental
present photoreaction takes place via an intermediate in which the
phosphorus atom has an unpaired electron. The step from 3 to 4
Instruments
•
+
does determine Qrel. On the other hand, the reaction of 1 with O
cannot be the Qrel-determining step because the results from the
LFP reveal no correlation between kO2 and Qrel
The ET from 1 to 3, instead of the radical attack by 3 on 1, has
2
The instruments used in the present study were described in a
11
previous article.
.
Materials
12b
been reported for the O
2
-mediated oxidation of 1 to 2 in CH
2
2
Cl .
The ET mechanism is unlikely at least under our experimental
conditions where the solvent is MeCN. The ET would promote
a chain reaction, which is against our observations in the present
study. Thus, the quantum yields of the formation of 2 from 1
were smaller than 0.1, indicating no chain reaction was involved.
The phosphonium center in 3 is thought to be more stabilized by
Phosphines 1, biphenyl (BP) and 9,10-dicyanoanthracene (DCA)
were commercially available and purified through recrystallization
11
as described.
Laser flash photolysis
−
2
−1
A solution of 1 (1.00 × 10 M), BP (1.00 × 10 M) and DCA
solvation in MeCN than in CH
the oxygen atom is not very solvated. If this is true, 3 could accept
an electron more easily in CH Cl than in MeCN.
2
2
Cl , whereas the radical center on
−
5
(
5.00 × 10 M) in MeCN was photolyzed. The decrease in the
resulting absorption was monitored at the appropriate wavelengths
Table 1).
2
2
(
In the present photoreaction, a steric effect has been observed
in the step where 3 attacks 1. Importantly, the LFP for 1h in
General procedure
A 5 ml solution containing BP (1.00 × 10 M) and DCA (5.00 ×
the O atmosphere produced a spectrum that strongly suggests
2
−
1
the formation of 3h (Fig. 2). This finding further supports the
significance of the steric effect on the radical attack by 3 on 1.
The two ortho-methyl groups of 1h prevent the radical attack
by 3h on 1h, making 3h long-lived enough for the spectroscopic
observation. This is also the reason for the lower Qrel in the photo-
oxidation of 1h.
Tris(o-tolyl)phosphine 1e, which has ortho-methyl groups, pro-
duced a low Qrel value. This finding is also in line with the steric
inhibition by the ortho-methyl groups on the radical attack. Tordo
et al. performed an X-ray analysis on several triarylphosphines
−
5
10 M) in MeCN was prepared in air. 2 ml portions from the
solution were separately added to two square quartz cells (1 cm ×
1 cm), one containing 2.00 mmol of 1a (as a standard) and the
other containing 2.00 mmol of the particular phosphine 1. These
solutions were subjected to the photoreaction as described below.
Each cell was irradiated with light from a xenon arc short lamp
through a sharp-cut filter (irradiation at k > 390 nm). At specific
intervals, a 50 ll aliquot was taken and diluted with 25 ll of
MeCN with benzyl ether (for 1a, b, e and f) or hydroquinone
2
972 | Org. Biomol. Chem., 2006, 4, 2969–2973
This journal is © The Royal Society of Chemistry 2006

View Article Online
dibenzyl ether (for 1c, d and g–j) as the external standard, and the
resulting mixture was analyzed by GC.
12 (a) Z. B. Alfassi, P. Neta and B. Beaver, J. Phys. Chem. A, 1997, 101,
2
153; (b) B. Beaver, D. Rawlings, P. Neta, Z. B. Alfassi and T. N. Das,
Heteroat. Chem., 1998, 9, 133.
1
1
3 S. Tojo, K. Morishima, A. Ishida, T. Majima and S. Takamuku, J. Org.
Chem., 1995, 60, 4684.
4 M. Fujita, A. Shindo, A. Ishida, T. Majima, S. Takamuku and S.
Fukuzumi, Bull. Chem. Soc. Jpn., 1996, 69, 743.
References
1
2
3
4
(a) S. Yasui, K. Shioji and A. Ohno, Tetrahedron Lett., 1994, 35, 2695;
b) S. Yasui, K. Shioji and A. Ohno, Heteroat. Chem., 1995, 6, 223.
S. Yasui, K. Shioji, A. Ohno and M. Yoshihara, J. Org. Chem., 1995,
0, 2099.
S. Yasui, M. Tsujimoto, K. Shioji and A. Ohno, Chem. Ber./Recl., 1997,
30, 1699.
(
15 We have obtained further evidence for the participation of 3 in the
present photoreaction based on pulse radiolysis experiments, which
will be published elsewhere.
16 Even if there were no fluctuations in determining the value of the slope,
it is still impossible to determine the absolute value of the quantum
yield since we did not make the light monochromatized.
17 S. Dincturk, R. A. Jackson, M. Townson, H. Agirbas, N. C. Billingham
and G. March, J. Chem. Soc., Perkin Trans. 2, 1981, 1121.
18 In ref. 17, the thermal homolytic cleavage of dibenzylmercury has been
6
1
(a) S. Yasui, K. Shioji, M. Tsujimoto and A. Ohno, J. Chem. Soc.,
Perkin Trans. 2, 1999, 855; (b) S. Yasui, K. Itoh, A. Ohno and N.
Tokitoh, Chem. Lett., 2001, 1056.
5
6
7
S. Yasui, M. Tsujimoto, K. Itoh and A. Ohno, J. Org. Chem., 2000, 65,
0
0
4
715.
analyzed based on the dual-parameter equation, log(k/k
0
) = r q +
•
•
S. Yasui, K. Itoh, M. Tsujimoto and A. Ohno, Bull. Chem. Soc. Jpn.,
002, 75, 1311.
M. Nakamura, M. Miki and T. Majima, J. Chem. Soc., Perkin Trans.
, 2000, 1447.
r q , where the first term is fixed by the reactions with a series of meta-
2
substituted substrates.
19 Since this assumption is not valid for the meta-substituent, Fig. 3 does
not include the plot for 1g.
20 G. B. Watts and U. K. Ingold, J. Am. Chem. Soc., 1972, 94, 2528.
21 W. G. Bentrude, Acc. Chem. Res., 1982, 15, 117.
22 S. Fukuzumi, K. Shimoosako, T. Suenobu and Y. Watanabe, J. Am.
Chem. Soc., 2003, 125, 9074.
2
8
9
R. L. Powell and C. D. Hall, J. Am. Chem. Soc., 1969, 91, 5403.
G. Pandey, S. Hajra and M. K. Ghorai, Tetrahedron Lett., 1994, 35,
7
837.
1
1
0 S. Takagi, T. Okamoto, T. Shiragami and H. Inoue, J. Org. Chem., 1994,
9, 7373.
1 S. Yasui, S. Tojo and T. Majima, J. Org. Chem., 2005, 70, 1276.
5
23 M. Culcasi, Y. Berchadsky, G. Gronchi and P. Tordo, J. Org. Chem.,
1991, 56, 3537.
This journal is © The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 2969–2973 | 2973