Hydrogen atom abstraction by radical cations. The reactions of 9-substituted acridine radical cations with cyclohexa-1,4-diene

Article abstract of DOI:10.1039/b102285g

9-Substituted acridine radical cations undergo facile hydrogen atom abstraction reactions with cyclohexa-1,4-diene in dichloromethane-Bu₄NPF₆ (0.2 M). The kinetics of the reaction were studied by derivative cyclic voltammetry and obs

Full text of DOI:10.1039/b102285g

View Article Online / Journal Homepage / Table of Contents for this issue
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Hydrogen atom abstraction by radical cations. The reactions of
9-substituted acridine radical cations with cyclohexa-1,4-diene†
Kishan L. Handoo, Jin-Pei Cheng‡ and Vernon D. Parker*
2
Department of Chemistry and Biochemistry, Utah State University, Logan, UT 84322-0300,
USA
Received (in Cambridge, UK) 12th March 2001, Accepted 29th May 2001
First published as an Advance Article on the web 6th July 2001
9-Substituted acridine radical cations undergo facile hydrogen atom abstraction reactions with cyclohexa-1,4-diene
in dichloromethane–Bu4NPF6 (0.2 M). The kinetics of the reaction were studied by derivative cyclic voltammetry
and observed to conform to the EC mechanism (charge transfer followed by irreversible first-order reaction). The
products of the reactions are the protonated acridine substrate and dimer–substrate adducts of cyclohexa-1,4-diene.
A comparison is made between the reactions of 9-phenylacridine and 9-phenylanthracene radical cations. The former
undergoes rapid hydrogen atom abstractions, while the latter slowly dimerizes in the absence of nucleophiles. It is
suggested that localization of the odd electron in a p-orbital on nitrogen imparts the hydrogen abstraction activity
of the radical cation. Since there are no free p-orbitals on the carbon atoms of the aromatic rings in arene radical
cations the hydrogen atom abstraction reaction pathway is not observed.
Introduction
Results
It has recently been pointed out1 that although radical cations
have dual functionality, their polar reactions have received very
much more attention than reactions that depend upon the
unpaired electron. Of the three most characteristic reactions
of free radicals (dimerization, atom abstraction and reaction
with dioxygen) only the former is frequently encountered
with radical cations. Radical cations of a number of olefins,2
including some tetraalkyl olefins3 related to biadamantylidene,
react with dioxygen to form dioxetanes, but the radical cations
derived from most other compounds are inert to dioxygen.
We have previously observed that some substituted acridine
radical cations abstract hydrogen atoms from the solvent upon
generation in acetonitrile.4
Radical cation substrates and the hydrogen atom donor
Although the radical cation of the parent compound in the
series, acridine, will most likely take part in hydrogen atom
transfer reactions, we omit this substrate due to practical
difficulties in obtaining kinetic data. We have used two series of
substituted acridines, 1 and 2, in our studies, along with the
hydrogen atom donor cyclohexa-1,4-diene (CHD), 3.
Dimerization, or coupling with substrate, reactions more
commonly exhibit the radical behaviour of radical cations. But
here again dimerization can only be expected when there are
specific structural features, in addition to the odd electron,
which promote free radical-like behavior. For example, dimeriz-
ation is a common pathway of methoxy-substituted benzene
radical cations, as long as there are unsubstituted positions
ortho or para to the methoxy group. It has previously1 been
proposed that the most important structural feature which
promotes dimerization of radical cations is the presence of
substituents which effectively decouple the charge and spin by
localization of the charge onto the substituent, for example on
the methoxy group.
Reversible oxidation potentials and hydrogen atom transfer
product yields
The reaction between an acridine radical cation and CHD is
illustrated by eqn. (1).
In this paper we present kinetic data for hydrogen atom
transfer reactions of the only series of radical cations, sub-
stituted acridine radical cations, which have been observed
to undergo the reaction. The data reported here support the
hypothesis that this reaction series, like those which undergo
radical cation dimerization, possess a structural feature that
promotes decoupling of charge and spin, resulting in the pro-
pensity to abstract hydrogen atoms from suitable donors.
ϩ
1ϩ ϩ CHD → 1-H ϩ 4
ؒ
(1)
The reaction is conveniently followed using cyclic voltam-
metry. Before addition of CHD to the solution containing the
substrate, the peak potential and peak current for the oxidation
of the substrate are recorded. After exhaustive electrolysis in
the presence of CHD the voltammogram is recorded before and
after neutralization by the addition of a suitable base. The yield
of the conjugate acid of the substrate can then be determined
by the peak current for oxidation of substrate measured at the
end of the procedure described above.
† This paper is dedicated to the memory of Professor Lennart Eberson
whose many contributions had an enormous influence on the chemistry
of radical cations and on the work of one of the authors (V. D. P.).
‡ Present address: Department of Chemistry, Nankai University,
Tianjin 300071, China.
Peak potentials for the oxidation of the various 9-substituted
acridines, along with the apparent % yields of protonated
1476
J. Chem. Soc., Perkin Trans. 2, 2001, 1476–1480
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b102285g

View Article Online
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Table 1 Reversible oxidation potentials of 9-substituted acridines and
yields of protonated substrates obtained during exhausive electrolytic
oxidation in acetonitrile.a
Table 3 Effect of cyclohexa-1,4-diene concentration on apparent rate
constants for the reactions with 9-methoxyacridine 1b and 9-phen-
oxyacridine 1c radical cations in dichloromethane–Bu4NPF6 (0.2 M)
at 298 K
b
Compound
Eox
Yield (%)c
1b
1c
9-Dimethylaminoacridine 1a
9-Phenoxyacridine 1b
0.478
1.210
1.067
1.322
1.243
0.496
—
83
80
86
94
90
[Diene]/
mM
10Ϫ3k2/
M
10Ϫ3 k1/
10Ϫ4 k2/
M
Ϫ1 sϪ1b
k1/sϪ1a
Ϫ1 sϪ1b
sϪ1a
9-Methoxyacridine 1c
9-Phenylacridine 2a
9-(4-Methoxyphenyl)acridine 2b
9-(4-Dimethylaminophenyl)acridine 2c
a Solvent containing Bu4NPF6 (0.1 M) at 298 K. b Reversible electrode
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
60.0
70.0
80.0
90.0
100
—
—
—
—
13.3
13.8
11.8
10.4
10.8
1.136
1.508
1.635
1.803
2.53
2.88
2.97
3.71
3.78
—
—
—
—
—
11.4
10.1
8.18
7.21
8.43
8.21
7.42
8.24
7.55
265
346
353
365
432
441
462
491
578
691
790
868
potential measured vs. ferricinium/ferrocene by derivative cyclic vol-
tammetry at 100 V sϪ1 c Yield of conjugate acid of the substrate, see
.
Experimental section for details.
9.80
9.24
8.18
8.26
8.64
8.78
8.68
Table 2 Substrate reaction order studies for the reactions of 9-
substituted acridine radical cations with cyclohexa-1,4-diene in
dichloromethane–Bu4NPF6 (0.2 M) at 298 K
—
—
—
—
—
Substrate
[Substrate]/mM
[Diene]/mM
νc /V sϪ1a
a First-order rate constants for the EC mechanism. b Apparent second-
order rate constants for the reactions of the radical cations with
cyclohexa-1,4-diene.
2a
0.50
1.00
1.50
2.00
0.50
1.00
1.50
2.00
0.50
1.00
1.50
2.00
10.0
10.0
10.0
10.0
20.0
20.0
20.0
20.0
100
579
565
569
561
158.0
158.4
155.3
156.6
81.8
81.2
83.9
82.7
2b
2c
Table 4 Effect of cyclohexa-1,4-diene concentration on apparent
rate constants for the reactions with 9-phenylacridine 2a and 9-(4-
methoxyphenyl)acridine 2b in dichloromethane–Bu4NPF6 (0.2 M) at
298 K
100
100
100
2a
2b
[Diene]/
mM
10Ϫ3 k1/
10Ϫ5 k2/
Ϫ1 sϪ1b
10Ϫ2 k1/
10Ϫ4 k2/
M
a Voltage sweep rate necessary to achieve a given RЈI value; RЈI = 0.4, 0.5
and 0.6 for 9-phenylacridine 2a, 9-(4-methoxyphenyl)acridine 2b, and
9-(4-dimethylamino)acridine 2c, respectively.
sϪ1a
M
sϪ1a
Ϫ1 sϪ1b
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100
2.40
3.87
4.57
5.45
6.52
7.79
11.3
11.1
17.8
16.6
2.40
1.93
1.52
1.36
1.30
1.30
1.61
1.39
1.97
1.66
—
—
5.42
8.60
10.2
11.0
12.9
13.4
16.4
18.0
20.9
2.71
2.87
2.54
2.20
2.15
1.91
2.05
2.00
2.09
products after exhaustive electrolysis in dichloromethane–
Bu4NPF6 (0.1 M) in the presence of CHD (0.1 M) at 298 K, are
summarized in Table 1. In the case of 9-dimethylaminoacridine
1a oxidation, the conjugate acid of the substrate was not the
exclusive product and the results were not reproducible. In all
other cases the yield was observed to be 80% or greater.
a,b See Table 3.
Kinetic studies of the reactions between 9-substituted acridine
radical cations with cyclohexa-1,4-diene
The kinetics of the reactions were studied by derivative
cyclic voltammetry5 (DCV), which has been used in numerous
radical ion kinetic studies in our laboratory. Ordinary cyclic
voltammetry is not suitable for detailed kinetic studies due to
the difficulty in establishing a baseline for measurement of
the reverse peak current. This difficulty is overcome with
DCV, since the baselines for both forward and reverse scans are
close to zero. Kinetic studies were carried out in dichloro-
methane–Bu4NPF6 (0.2 M) at 298 K. This solvent system has
the advantage that it does not provide as good a source of
hydrogen atoms as the acetonitrile–Bu4NPF6 (0.1 M) system,
so the radical cations are longer lived in the absence of added
hydrogen atom donor. This latter is advantageous when
measuring second-order rate constants between radical cations
and hydrogen atom donors, since the effect of the background
reaction is eliminated at lower hydrogen atom donor
concentrations.
tions second-order in radical cation. The values of νC are,
within experimental error, independent of CA in all three cases
shown in Table 2. The latter provides convincing evidence that
the reactions of the radical cations with CHD are cleanly first-
order in radical cation.
Since the radical cations are not long-lived in the absence of
added hydrogen atom donor it is necessary to insure that the
background reaction is insignificant under the conditions that
second-order rate constants for hydrogen atom abstraction are
determined. A convenient way to do this is to evaluate apparent
second-order rate constants over a range of hydrogen atom
donor concentrations, [CHD] in this work. The data in Tables 3,
4 and 5 illustrate the determination of second-order rate con-
stants for the radical cation hydrogen atom abstractions. The
first-order rate constants for the reactions of the radical cations
were assigned by comparison of experimental to theoretical
data for the EC mechanism, which is an irreversible first-order
reaction following charge transfer. If the background reaction
is not negligible, apparent second-order rate constants will be
larger than the value for the reaction between radical cation and
CHD. The data in all three tables show that at the low end
of [CHD] the apparent second-order rate constants are large
and that they decrease to a constant (within experimental
The data in Table 2 show the effect of substrate concen-
tration (CA) on the voltage sweep rate (νC) necessary for a par-
ticular value of RЈI, the ratio of reverse to forward derivative
peak heights. If the reaction order in radical cation is unity,
νC is predicted by the reaction order approach5 to be constant
independent of CA, while νC varies directly with log CA for reac-
J. Chem. Soc., Perkin Trans. 2, 2001, 1476–1480
1477

View Article Online
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Table 5 Effect of cyclohexa-1,4-diene concentration on apparent rate
constants for the reactions with 9-(4-dimethylaminophenyl)acridine 2c
in dichloromethane–Bu4NPF6 (0.2 M) at 298 K
[Diene]/mM
RЈI
k1/sϪ1a
k2/MϪ1 sϪ1b
0.104
0.156
0.209
0.260
0.312
0.365
0.418
0.520
0.728
0.718
0.676
0.639
0.608
0.571
0.525
0.503
0.457
0.409
131.5
155.9
180.5
203.4
234.3
280.3
306.2
371.2
459.0
1255
999
864
782
751
768
733
714
631
a,b See Table 3.
error) value as [CHD] is increased. The magnitudes of [CHD]
necessary for second-order rate constant determination depend
upon the particular radical cation reaction being studied.
The second-order rate constants for the reactions between
9-substituted acridine radical cations and cyclohexa-1,4-diene
are: 8.63 (0.38) × 103 (1b); 7.89 (0.48) × 104 (1c); 1.56 (0.26) ×
105 (2a); 4.39 (0.24) × 104 (2b); and 7.41 (0.69) × 102 MϪ1 sϪ1
(2c).
Discussion
Hydrogen atom abstraction is a rare reaction of radical ions.
Until now the strongest evidence for this type of reaction has
been our observation of a solvent deuterium kinetic isotope
effect of about 20 when the diazodiphenylmethane anion
radical reacted with the solvent, either CH3CN or CD3CN.6,7
The results could only be explained by a rate-determining
hydrogen atom abstraction by the anion radical from the
solvent. Diphenylmethane was observed to be a primary prod-
uct of the reaction that involves an initial hydrogen atom
abstraction, presumably followed by rapid protonation and loss
of dinitrogen (Scheme 1).
Scheme 2
The primary objective of this work was to see if conditions
could be found under which acridine radical cations act as the
acceptor in hydrogen atom transfer reactions. Since radical
cations are prone to react with nucleophiles,9,10 non-nucleo-
philic dichloromethane was selected as the solvent and
cyclohexa-1,4-diene was chosen as an unambiguous hydrogen
atom donor. The results presented in Table 2 verify that the
reactions are first-order in radical cation. First-order behaviour
in [CHD] is demonstrated in Tables 3–5. The radical cations
react slowly in dicloromethane–Bu4NPF6 (0.2 M) in the absence
of intentionally added reactants. We are not sure what the
background reaction is, but the data clearly show that as [CHD]
is increased the apparent second-order rate constants decrease
and approach constant values, equal to k2 (Scheme 3).
Scheme 1
In our earlier communication we reported hydrogen atom
abstraction reactions from solvent acetonitrile by 9-substituted
acridine radical cations.4 The mechanism of this reaction
apparently does not involve a simple hydrogen atom transfer
from the solvent, since we failed to observe significant rate
differences when the solvent was changed from CH3CN to
CD3CN. The reaction was observed to be first-order in radical
cation and in H2O and measurements in the presence of D2O
revealed a deuterium kinetic isotope effect of close to 1.5. It was
proposed that the rate-determining step in the reaction was
attack of the nucleophile (water) at the 9-position of the radical
cation, activating hydrogen atom abstraction by the nitrogen
center (Scheme 2). A similar mechanism may account for the
cyclic voltammetry data reported for substituted flavin radical
cations.8
Scheme 3
Attempts to find linear Hammett relationships for the hydro-
gen atom abstraction reactions of the 9-substituted acridine
radical cations are summarized in Table 6. The results for com-
pounds 2a–c, with remote subsituents, on the phenyl ring, are
shown in the upper half of Table 6. Although only three rate
constants are available, reasonably good correlations are found
with both σp and σpϩ, with the highest correlation coefficient (r2)
Our study4 did not provide evidence for the direct involve-
ment of radical cations in hydrogen atom abstraction reactions
analogous to the earlier report7 for the radical anion reaction.
1478
J. Chem. Soc., Perkin Trans. 2, 2001, 1476–1480

View Article Online
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Table 6 Hammett relationships for remote (Y, compounds 2) and
direct (X, compounds 1) 9-substituents on acridine radical cations in
reactions with Cyclohexa-1,4-diene
of the free radical behaviour due to localization of the odd
electron in a non-bonding orbital on nitrogen.
An alternative mechanism involving electron transfer
between the 9-substituted acridine radical cations and CHD to
give the substrate and CHD radical cation is less likely. This
reaction might be expected to be followed by proton transfer
from the CHD radical cation to the 9-substituted acridine.
Assuming the latter, the final products are expected to be
the same for the two mechanisms. On the other hand, the
voltammetry response is inconsistent with the electron
transfer mechanism, which is expected to show electrocatalytic
behaviour.
a
Compound
Y
log (k2/MϪ1 sϪ1
)
)
σpa
σϩ
p
2a
2b
2c
ρ
H
OMe
NMe2
5.19
4.64
2.87
0
0
0.100
Ϫ0.43
3.95
Ϫ0.80
Ϫ1.73
1.36
r2
0.92
0.97
a
Compound
X
log (k2/MϪ1 sϪ1
σpa
σϩ
p
We conclude that hydrogen atom abstraction by radical
cations is not generally a favorable reaction pathway. The
reaction can only be expected in cases where there is a structural
feature favoring localization of the odd electron in a non-
bonding orbital. The latter is expected to promote free radical-
like behavior.
1
Ph
OPh
OMe
5.19
3.94
4.90
0.039
0.063
0.100
N.A.b
Ϫ0.20
Ϫ0.53
Ϫ0.80
N.A.b
1b
1c
ρ
a Substitutent constants from reference 11. b Scatter, no correlation.
Experimental
observed with the latter (0.97). Remote resonance electron
donating substituents substantially decrease the rate of hydro-
gen atom abstraction from CHD. Linear Hammett correlations
are not observed for substitution directly on the acridine ring.
Both 1b and 1c show a decreased rate of hydrogen abstraction
relative to phenyl to a small degree.
All of the results clearly show that 9-substituted acridine
radical cations, in the absence of nucleophiles, undergo
facile hydrogen atom abstraction reactions. What is the
structural feature that promotes this unusual reaction? The
9-phenylanthracene radical cation might be considered to be
structurally related to the radical cation of 9-phenylacridine.
The former undergoes typical radical cation reactions including
combination with nucleophiles when present and dimerization
in their absence. We find no evidence for the dimerization of
9-phenylacridine radical cation and there is no evidence that the
9-phenylanthracene radical cation takes part in hydrogen atom
abstraction reactions. Obviously, the structural difference
between the two radical cations is the presence or absence of a
nitrogen atom at the 10-position.
Materials
Dichloromethane, after passing through active neutral alumina,
was used without further purification. Tetrabutylammonium
hexafluorophosphate (Aldrich) was recrystallized from di-
chloromethane–diethyl ether before use. Cyclohexa-1,4-diene
(Aldrich) was of the highest purity available and used as
received. 9-Substituted acridines were prepared using standard
literature methods.12
Instrumentation and data handling procedures
Cyclic voltammetry was performed with a Princeton Applied
Research (Princeton, NJ) Model 273 potentiostat/galvanostat
driven by a Hewlett Packard 3314A function generator. After
passing through a Stanford Research Systems, Inc. Model
SR640 dual channel low pass filter, the data were recorded on a
Nicolet Model 310 digital oscilloscope with 12-bit resolution.
The oscilloscope and function generator were controlled by a
personal computer via an IEEE interface. The current potential
curves were collected at selected trigger intervals to reduce
periodic noise,13 and 20 curves were averaged before being
treated with a frequency domain low pass digital filter and
numerical differentiation.
It has been proposed1 that much of radical ion reactivity can
be accounted for in terms of the degree of coupling between the
radical and ionic centers. For example neither the radical anion
nor the radical cation of anthracene undergo rapid dimeriz-
ation reactions. However, localization of charge on a substituent
of the radical ion promotes dimerization. 9-Cyanoanthracene
radical anion and 9-methoxyanthracene cation radical undergo
facile dimerization. A number of other examples of similar
phenomena have been given.
Removal of an electron from the π-system of a 9-substituted
acridine might be expected to produce a radical cation with the
charge and the odd electron delocalized over the π-system. If
this were the dominant structure of the radical cation, it seems
unlikely that the unusual hydrogen atom abstraction behavior
would be observed. However, if an electron from the unshared
pair on nitrogen is promoted to the π-system, the odd electron
becomes localized in a nonbonding orbital and is free to
take part in typical radical reactions such as hydrogen atom
abstraction. Obviously, there is no comparable structural
feature available to the 9-phenylanthracene radical cation and
it does not take part in hydrogen atom abstraction reactions.
The results from the Hammett correlations appear to be
consistent with the explanation presented in the previous
paragraph. The resonance donating substituents promote
delocalization of the charge and the odd electron into the π-
system and decrease the hydrogen abstraction activity of the
radical cations. The reasonably good Hammett correlations for
compounds 2a–c (Table 6) most likely reflect the delocalization
of the charge and the odd electron into the 9-aryl group; the
greater the extent of delocalization, the lesser the importance
Cyclic voltammetry measurements
A standard three-electrode one compartment cell was used for
all kinetic measurements. Positive feedback IR compensation
was used to minimize the effects of uncompensated solution
resistance. Reference electrodes were Ag/AgNO3 (0.01 M) in
acetonitrile constructed in the manner described by Moe.14 The
working electrodes, 0.4 mm Pt, were prepared by sealing wire in
glass and polishing to a planar surface as described previously.15
The cell was immersed in a water bath controlled at 25 0.2 ЊC.
Kinetic measurements
Rate constants were obtained by comparing derivative cyclic
voltammetry5 data to theoretical data obtained by digital
simulation.16 The reactions were studied by using solutions
containing substrates (0.5–2.0 mM) and cyclohexa-1,4-diene
(0.01–0.8 M).
Product studies
Yields of acridinium ion products were obtained by cyclic
voltammetry peak current measurements. The voltammogram
was recorded before exhaustive controlled potential coulo-
metric oxidation of substrate (1.0 mM) in the presence of a
large excess of cyclohexadiene. Immediately after the oxidation
J. Chem. Soc., Perkin Trans. 2, 2001, 1476–1480
1479