Kinetics of the base-catalyzed permanganate oxidation of benzaldehyde

Article abstract of DOI:10.1021/jo991540l

The kinetics of the base-catalyzed permanganate oxidation of benzaldehyde have been reexamined. The rate is proportional to the first power of the aldehyde and permanganate concentrations, and there are terms that are zero order, first order, and second order in hydroxide ion. The reaction has an isotope effect, and the effect of substituents gives ρ = +1.58. The possible mechanisms for the reaction are discussed in the context of ab initio calculations at the B3P86/6-311+G** and MP2/6-31G* theoretical levels, and both one- and two-electron processes are possible. Benzaldehyde hydrate dianion is calculated to have a remarkably small C-H bond dissociation energy.

Full text of DOI:10.1021/jo991540l

J . Org. Chem. 2000, 65, 573-576
573
Kin etics of th e Ba se-Ca ta lyzed P er m a n ga n a te Oxid a tion of
Ben za ld eh yd e
Kenneth B. Wiberg* and Fillmore Freeman1
Department of Chemistry, Yale University, New Haven, Connecticut 06520-8107
Received October 1, 1999
The kinetics of the base-catalyzed permanganate oxidation of benzaldehyde have been reexamined.
The rate is proportional to the first power of the aldehyde and permanganate concentrations, and
there are terms that are zero order, first order, and second order in hydroxide ion. The reaction
has an isotope effect, and the effect of substituents gives F ) +1.58. The possible mechanisms for
the reaction are discussed in the context of ab initio calculations at the B3P86/6-311+G** and
MP2/6-31G* theoretical levels, and both one- and two-electron processes are possible. Benzaldehyde
hydrate dianion is calculated to have a remarkably small C-H bond dissociation energy.
Some years ago, we reported an investigation of the
kinetics of the permanganate oxidation of benzaldehyde.2
The pH-rate profile had three regions corresponding to
an acid-catalyzed reaction below about pH 5, an uncata-
lyzed reaction from about pH 6 to pH 9, and a base-
catalyzed reaction above pH 10. The kinetics of the latter
reaction could not be determined precisely because of the
high rate constants and the analytical method used. We
now wish to report an extended investigation of the base-
catalyzed reaction using a stopped-flow reactor to mea-
sure spectrometrically the change in permanganate
concentration with time.
Ta ble 1. Effect of P er m a n ga n a te Con cen tr a tion on th e
Ra te of Oxid a tion a
-
4
k1 × 10-2 s-1
[MnO4 ] × 10
k (av)
2
.5
6.02
6.13
6.05
6.29
6.08 ( 0.05
3
.0
6
6
6
6.38
6.41
.37
.40
.25
6.33 ( 0.06
5.0
6
6
.29
.61
6.42 ( 0.09
6.54 ( 0.16
The reaction was studied in basic aqueous solutions
under pseudo-first-order conditions with permanganate
as the limiting reagent. Under these conditions, manga-
nate is the inorganic product. Good first-order behavior
was noted through better than 75% reaction, and the rate
constant did not change significantly with a 3-fold change
in permanganate concentration (Table 1). For the re-
mainder of the experiments, a permanganate concentra-
7
.5
6.44
.33
6
6.82
a
-
λ ) 522 nm, T ) 25.0 °C, [HO ] ) 0.25 M, pH ) 13.4, µ ) 0.4,
-3
[
PhCHO] ) 9.8 × 10 M.
-
4
tion of 3 × 10 M was used.
The effect of benzaldehyde concentration on the rate
of reaction was determined using a 10-fold change in
concentration. The data are summarized in Figure 1. A
small intercept is found on extrapolation to zero benzal-
dehyde concentration. This in part results from the base-
3
catalyzed decomposition of permanganate. However, it
appears somewhat larger than might be expected from
this source.
The effect of hydroxide ion concentration also was
determined, giving the data summarized in Figure 2. A
linear relationship was not obtained. The curvature
suggests a kinetic form of the type
-
- 2
kobs ) k + k [HO ] + k [HO ]
1
2
3
A least-squares fit of the experimental data to the above
F igu r e 1. Effect of benzaldehyde concentration on the pseudo-
first-order rate constants for the disappearance of permanga-
nate ion.
-
2
-2
equation gave k
1
) 2.06 × 10 , k
2
) 14.9 × 10 , and
-
2
k
3
) 11.1 × 10 for an aldehyde concentration of 9.8 ×
10- M. The solid line in the figure is calculated using
3
(
1) Department of Chemistry, University of California, Irvine, CA.
(2) (a) Wiberg, K. B.; Stewart, R. J . Am. Chem. Soc. 1955, 77, 1786.
these constants and can be seen to give an excellent fit.
The dotted line represents k
oxidation of benzaldehyde may now be expressed as
(
b) Tompkins, F. C. Trans. Faraday Soc. 1943, 39, 280.
3) J ezowska-Trzebiatowska, B.; Kalecinski, J . Chem. Abstr. 1960,
4, 19122.
2
alone. The kinetics of the
(
5
1
0.1021/jo991540l CCC: $19.00 © 2000 American Chemical Society
Published on Web 12/23/1999

5
74 J . Org. Chem., Vol. 65, No. 2, 2000
Wiberg and Freeman
F igu r e 2. Effect of hydroxyl ion concentration on the pseudo-
F igu r e 3. Effect of substituents on the rate constants for the
permanganate oxidation of benzaldehyde in 0.25 M sodium
hydroxide solutions. From left to right, the substituents are
first-order rate constants for the disappearance of permanga-
-
3
nate ion with a benzaldehyde concentration of 9.8 × 10 M.
The solid line gives the best fit to the data, and the dashed
p-MeO, p-Me, H, m-MeO, p-Cl, p-CN, m-NO
2
2
, and p-NO .
1 2
line indicates the contribution of just k and k .
monoanion with permanganate is 84 L mol- s-1, and
most of the substituent effect on the rate of reaction may
be attributed to the formation of the monoanion. The
rate-determining step has a small negative value of F.
1
Ta ble 2. Kin etic Isotop e Effect in th e Oxid a tion of
Ben za ld eh yd e
2
-1
2
-1
[
HO-]
kH × 10 s
kD × 10 s
k/kD
0
0
0
0
0
.01
.05
.25
.35
.50
2.52 ( 0.04
2.77 ( 0.02
6.29 ( 0.02
8.57 ( 0.02
12.63 ( 0.03
1.14 ( 0.08
1.21 ( 0.02
2.05 ( 0.02
2.44 ( 0.02
3.59 ( 0.02
2.2
2.2
3.1
3.5
3.5
a
The pK of the monoanion is not known, but because
it has a negative charge its acidity should be considerably
less than that of an alcohol. If it were estimated to be
-
6
20, K
B
would be on the order of 10 , and the rate
constant for the reaction of the dianion with permanga-
-
8
-1 -1
-
d[MnO -]/dt ) k [PhCHO][MnO ] +
nate would be on the order of 10 L mol
s . This is
4
1
4
k2[PhCHO][MnO4-][HO ] +
-
close to diffusion controlled.
There are four possible reaction paths for the anion-
permanganate reactions: (a) hydride donation from the
anion to permanganate; (b) hydrogen atom donation from
the anion to permanganate; (c) one-electron transfer from
the anion to permanganate; and (d) two-electron transfer
from the anion to permanganate with loss of the aldehyde
hydrogen as a proton.
-
- 2
k3[PhCHO][MnO ][HO ]
4
) 2.10 L mol s-1, k ) 15.2 L mol s-1, and
) 11.3 L mol
The kinetic isotope effect was examined using benzal-
dehyde-d, giving the data in Table 2. The value of k /k
-
1
2
-2
where k
1
2
3
-3 -1
k
3
s .
H
D
appears to increase somewhat on going from 0.01 to 0.25
M base and then levels off. The first two values may,
however, be somewhat in error because the rates were
so slow that formation of manganese dioxide began to
be a problem.
Finally, the effect of substituents was examined using
a set of eight aldehydes and 0.25 M sodium hydroxide.
The results are summarized in Figure 3. A fairly good
linear relationship was obtained with F ) +1.54.
The data suggest that benzaldehyde initially adds
hydroxyl ion and that the ion thus formed may be further
deprotonated in an equilibrium step to form the dianion:
All three species (benzaldehyde, the monoanion, and
the dianion) may then be attacked by permanganate in
rate-determining steps. It was of interest to estimate the
rate constants for the rate-determining steps. This
Paths b and c are one-electron transfers giving an
intermediate that would react with another permanga-
nate to give the product, benzoic acid. Paths a and d are
two-electron processes that give benzoic acid directly
along with Mn(V) that would react with another per-
requires information on the equilibrium constants K
. K has been measured in aqueous solution and was
found to be 0.18, and the F value for the addition was
A
and
K
B
A
4
+
2.24. Thus, the rate constant for the reaction of the
(4) Greenzaid, P. J . Org. Chem. 1973, 38, 3164.

Permanganate Oxidation of Benzaldehyde
J . Org. Chem., Vol. 65, No. 2, 2000 575
Ta ble 3. Ca lcu la ted En er gies
Total Energies (H) and Zero-Point Energies (kcal/mol)
B3P86
compound
ꢀ ) 1
ꢀ ) 80
s2
MP2
s2
PMP2
ZPEa
PhCH(OH)2
PhCH(OH)O
PhCHO2
-423.30974
-422.7274
-421.94593
-422.64446
-422.11418
-421.41830
-423.31835
-422.81387
-422.25068
-422.65671
-422.19381
-421.66661
-420.69582
-420.09920
-419.27707
-420.04346
-419.49366
-418.76404
83.6
74.5
65.0
73.1
66.4
58.2
-
2
-
PhC(OH)2
PhC(OH)O
PhCO2
0.776
0.763
0.756
0.766
1.035
0.809
-420.04560
-419.51153
-418.76903
-
2
-
C-H Bond Dissociation Energies (kcal/mol)b
compound
PMP2
B3P86, ꢀ ) 1
B3P86, ꢀ ) 80
PhCH(OH)2
PhCH(OH)O
PhCHO2
85
48
0
93
63
11
91
68
46
-
-
2
a
HF/6-31G* zero-point energies scaled by 0.893. b The MP2/6-31G* energy of a hydrogen atom is 0.49823 H. DFT models lead to an
incorrect energy for a hydrogen atom because they attempt to introduce correction for electron correlation in this atom, for which there
cannot be any such energy. Thus, the correct value, 0.50000 H, was used.
manganate to give Mn(VI). Similar pathways may be
written for the dianion.
electrostatic energies by over 100 kcal/mol. The effect of
a polar solvent was examined using the SCIPCM reaction
field model, and the results are included in Table 3. In
8
Path c seems relatively unlikely because it does not
directly involve the aldehyde hydrogen and could only
give a secondary isotope effect. Thus, this process is not
likely to give a kinetic isotope effect of 3.5, as has been
observed for the reaction. It is not possible to distinguish
between the remaining paths on the basis of just the
available experimental data.
the case of anions, the electron density distribution is
fairly diffuse. As a result, a significant part of the total
3
electron density lies outside the 0.0004 e/au isodensity
surface that is usually taken to give the size and shape
of the solute. As a result, it was necessary to scale the
charge so that all would lie within the solvent cavity. It
can be seen that the calculated dissociation energies for
benzaldehyde hydrate and the monoanion are not much
affected by going from the gas phase to a polar medium.
In the case of the dianion, the calculated BDE increases
significantly on going to a polar medium, and as noted
below, this is due to the two negative charges being better
dispersed in the dianion radical than in the dianion. The
BDE is, nevertheless, significantly lower than that for
the monoanion. Further, only part of the charge redis-
tribution can occur in the transition state for C-H bond
cleavage in the hydrate dianion. Thus, the kinetic effect
of going from the hydrate monoanion to the dianion
should be between the gas phase and solution calculated
values.
Thus, we have carried a series of ab initio calculations
for the species that are involved. If the one-electron
transfer process c were involved, the conversion of the
monoanion to the dianion would have to result in a
marked reduction in the C-H bond dissociation energy
to account for the much more rapid reaction of the
5
dianion. Geometry optimizations were carried out at the
6
B3P86/6-311+G** level, which makes use of a relatively
flexible basis set including diffuse functions that have
been found to be important in describing anions and
includes correction for the effects of electron correlation.
In addition, density functional theory models have been
found to minimize spin contamination in free radicals.7
The results are summarized in Table 3.
Benzaldehyde hydrate has a calculated C-H bond
dissociation energy similar to that for toluene (88 kcal/
mol). The monoanion has a much smaller calculated
dissociation energy, and that for the dianion is remark-
ably small. It seemed possible that the density functional
method might give an anomalous energy for the dianion
radical. Thus, geometry optimizations were carried out
at the MP2/6-31G* level, and the calculated energies are
included in Table 3. The B3P86 and MP2 derived
dissociation energies differ by about 10 kcal/mol. How-
ever, the changes in BDE on going from benzaldehyde
hydrate to the monoanion and the dianion are the same
in both series, and both predict that the dianion will have
a low C-H dissociation energy.
The low dissociation energy of the dianion makes
process c a possible mechanism for the rate-determining
step. Process a is also possible because the dianion would
be a better hydride transfer reagent than the monoanion.
Process d is not excluded by any of the data. This brings
us back to the old question as to whether elements such
as manganese are able to accept two electrons in one step,
or if they are required to accept electrons one at a time.
Further calculations are being carried out to locate the
transition state for the reaction and to examine its
electron-transfer characteristics, taking the solvent into
account. These calculations will also allow the predication
of the kinetic hydrogen isotope effect, which may be
compared with the observed value.
2
-
Dianions are generally unstable in the gas phase and
eject an electron, forming a monoanion. They are, how-
ever, often stable in polar solvents that reduce their
The stability of the dianion radical, PhCO • , is
2
interesting in itself. To gain information on its bonding
and electron density distribution, its MP2 wave function
was examined using Bader’s theory of atoms in mol-
(
5) Gardner, K. A.; Mayer, J . M. Science 1995, 269, 1849. These
9
ecules. Both the AIM analysis and the Mulliken popu-
authors have shown that the rates of permanganate oxidation of C-H
bonds are related to the bond dissociation energies.
(
6) Becke, A. D. J . Chem. Phys. 1993, 98, 5648. Perdew, J . P. Phys.
Rev. 1986, B33, 5048.
7) Wiberg, K. B.; Cheeseman, J . R.; Ochterski, J .; Frisch, M. J . J .
Am. Chem. Soc. 1995, 117, 6535
(8) Foresman, J .; Keith, T. A.; Wiberg, K. B.; Snoonian, J .; Frisch,
M. J . J . Phys. Chem. 1996, 100, 16098.
(9) Bader, R. F. W. Atoms in Molecules: A Quantum Theory;
Clarendon Press: Oxford, 1990.
(

5
76 J . Org. Chem., Vol. 65, No. 2, 2000
Wiberg and Freeman
lationanalysis agreed that the CO group had a unit
2
movement of the first electron could be closely followed
by the second, and it would be difficult to observe the
separate steps. However, with benzaldehyde hydrate
dianion, one-electron transfer to manganese might be
immediately followed by transfer of the second electron
to the phenyl ring. If this were found, it would demon-
strate the operation of one-electron transfers.
negative charge (i.e., it is an ordinary carboxylate group),
and the phenyl ring had the odd electron and a negative
charge. Thus, the low dissociation energy derives at least
in part from the possibility for transferring the odd
electron of the radical as it is formed to the phenyl ring,
thus separating the two negative charges.
This observation may prove significant in the context
of determining via computations whether two electrons
may be simultaneously transferred to manganese. In
most cases, if one-electron transfers were operative, the
Ack n ow led gm en t . This investigation was sup-
ported by a grant from the National Science Foundation.
J O991540L