Remote substituent effects on the reactivity of 9-aryl- and 9,10-diarylanthracene radical cations with anions and amines

Article abstract of DOI:10.1021/jp982094t

Radical cations of 9-aryl- and 9,10-diarylanthracenes with substituents on the 4 position of the aryl rings (PA-X^?+ and DPA-X^?+, respectively) have been generated by photoionization in acetonitrile. Their reactivity with n-butylamine (n-BuNH²) and 1,4-diazabicyclo[2.2.2]octane (DABCO) and a number of anions (CH₃CO₂^-, Br^-, CN^-, N_3-) has been studied using nanosecond laser flash photolysis. Reactions proceed by electron transfer and/or nucleophilic addition. Using PA-X and DPA-X as chemical probes, simple criteria are established that allow one mechanistic pathway to be distinguished from another. When electron transfer is thermodynamically feasible, this pathway dominates (e.g., DABCO and azide). For endothermic electron transfer, addition is not necessarily the preferred ultimate reaction pathway and an inner sphere process (addition/ homolysis) can compete. In these cases other, criteria including steric factors and the strength of the incipient bond become important. Simple kinetic criteria and an approach to estimate the thermochemistry of the addition process are developed. It is clear from these studies that reactivity trends in the radical cation chemistry cannot be generalized as easily as those in carbocation chemistry. This has some implications concerning the development and utility of "clock" reactions in radical cation chemistry.

Full text of DOI:10.1021/jp982094t

J. Phys. Chem. A 1998, 102, 6503-6512
6503
Remote Substituent Effects on the Reactivity of 9-Aryl- and 9,10-Diarylanthracene Radical
Cations with Anions and Amines1
Mark S. Workentin*,†
The Department of Chemistry, The UniVersity of Western Ontario, London, Ontario, Canada N6A 5B7
Vernon D. Parker¶
Department of Chemistry and Biochemistry, Utah State UniVersity, Logan, Utah 84322-030
Tracy L. Morkin‡,§ and Danial D. M. Wayner*,‡
Steacie Institute for Molecular Sciences, National Research Council of Canada, 100 Sussex DriVe Ottawa,
Ontario, Canada K1A 0R6
ReceiVed: May 4, 1998
Radical cations of 9-aryl- and 9,10-diarylanthracenes with substituents on the 4 position of the aryl rings
(PA-X•+ and DPA-X•+, respectively) have been generated by photoionization in acetonitrile. Their reactivity
-
with n-butylamine (n-BuNH2) and 1,4-diazabicyclo[2.2.2]octane (DABCO) and a number of anions (CH3CO2 ,
-
Br-, CN-, N3 ) has been studied using nanosecond laser flash photolysis. Reactions proceed by electron
transfer and/or nucleophilic addition. Using PA-X and DPA-X as chemical probes, simple criteria are
established that allow one mechanistic pathway to be distinguished from another. When electron transfer is
thermodynamically feasible, this pathway dominates (e.g., DABCO and azide). For endothermic electron
transfer, addition is not necessarily the preferred ultimate reaction pathway and an inner sphere process (addition/
homolysis) can compete. In these cases other, criteria including steric factors and the strength of the incipient
bond become important. Simple kinetic criteria and an approach to estimate the thermochemistry of the
addition process are developed. It is clear from these studies that reactivity trends in the radical cation chemistry
cannot be generalized as easily as those in carbocation chemistry. This has some implications concerning
the development and utility of “clock” reactions in radical cation chemistry.
Introduction
have found that 9-phenyl- and 9,10-diphenylanthracene deriva-
tives are useful radical cation precursors to study the dynamics
and mechanistic aspects of the initial interaction between an
aromatic radical cation and a “nucleophilic” species by laser
flash photolysis techniques.5,6 In these systems proton transfer,
when possible, has been shown to occur with rates of 103 M-1
s-1 or lower.19g Thus, these reactions are not commonly
observed because the same base will generally also react by
addition or electron transfer. These latter reactions can occur
with rate constants approaching the diffusion controlled limit.6,19
Delineating the competition between electron transfer and
addition is complicated because the initial products of each
process can be chemically linked or involve a common transition
state. For example, an apparent addition process may occur in
a single step or by a stepwise electron transfer followed by
combination of the oxidized nucleophile with the neutral
aromatic. Similarly, an apparent electron transfer reaction could
result from an initial addition followed by dissociation of the
σ-bonded radical product; a formal inner sphere ET (Scheme
1). Thus, product studies do not necessarily provide unequivocal
evidence for any of the possible mechanisms. Furthermore,
without a clear indication of the mechanism of the reaction,
product studies cannot be used to assign rate constants deter-
mined by time resolved techniques to a particular reaction
process or be used in competitive quenching experiments where
product ratios are used to obtain absolute rate constants using
The key role of radical cations as intermediates in electron
transfer initiated processes and the increased interest in their
use in synthetic strategies2,3 has prompted a number of time-
resolved studies of the dynamics of their reaction processes.2a,4-16
An area of particular interest has been the investigation of the
kinetic and mechanistic aspects of the bimolecular reactivity
of radical cations with electron rich species,4-6,13,14,16-20 as well
as experimental and theoretical21 comparisons of their reactivity
to the reaction of carbonium ions with the same reactants. A
reaction between a radical cation and an electron rich “nucleo-
philic” species, such as an anion or amine, can proceed by a
number of possible competing processes including, nucleophilic
addition, proton transfer, and electron transfer. Due to a general
lack of information on the thermodynamic and kinetic factors
that control the chemistry of radical cations, it is often difficult
to predict a priori which of the reaction pathways, if any, will
be favored for a particular radical cation-“nucleophile” com-
bination.
Our interest is in developing molecular probes that will allow
systematic studies of the dynamics of radical cation reactivity
in order to better understand these factors. In this regard, we
† The University of Western Ontario. Email: mworkent@julian.wwo.ca.
Fax: (519) 661-3022.
‡ National Research Council (NRC).
¶ Utah State University.
§ NRC WES summer student 1994-95.
S1089-5639(98)02094-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/21/1998

6504 J. Phys. Chem. A, Vol. 102, No. 32, 1998
Workentin et al.
SCHEME 1
ethanol (TFE) the reaction of the same radical cations with azide
is by nucleophilic addition as a result of a +500 mV shift (12
-
kcal mol-1) in the oxidation potential of N3
.
As an extension of our two previous studies we now report
absolute kinetic and spectroscopic data for the reactions of a
number of anions (CH3CO2-, CN-, Br-, and N3-) and amines
(n-butylamine and 1,4-diazabicyclo[2.2.2]octane) with the radi-
cal cations of para-substituted 9-arylanthracenes (PA-X) and
bis-para-substituted 9,10-diarylanthracenes (DPA-X) determined
using nanosecond laser flash photolysis techniques. By varying
the substituent at the 4-position of the phenyl group at the 9-
and 9,10-positions of PA and DPA, we are able to modulate
the redox properties of the radical cations in order to evaluate
the remote electronic effects on their reactivity. The results of
this study allow us to differentiate between nucleophilic addition
or electron transfer in the initial interaction between these
“nucleophiles” and the aromatic radical cations and provide
important insights into the factors that control the reactivity and
mechanisms of aromatic radical cations. This study comple-
ments other studies of the reactions of these and similar aromatic
radical cations and provide important details for those utilizing
radical cations in preparative reactions. In particular, we caution
against using absolute rate constants obtained for the reaction
of radical cations with nucleophilic species as the basis of radical
cation clock reactions for competitive quenching experiments
analogous to that used in carbocation chemistry.
“clock” reactions as has been done extensively in cation-
nucleophile reactions.22
For a number of years, one of us has been involved in studies
examining the reaction of anions and amines with the radical
cations of anthracene derivatives by electrochemical tech-
niques.17-19 The electrochemical method is limited to studies
involving those nucleophiles whose redox behavior does not
interfere with the generation or observation of the radical cation
of the anthracene species. Additionally, an evaluation of the
kinetics of the reaction involved in the initial interaction between
the nucleophile and the radical ion is difficult to evaluate because
in the analysis one must consider that the intermediates formed
are also electroactive and reactive species. In the majority of
these studies, the mechanism for the bimolecular reactions of
arylanthracene radical cations with nucleophiles and bases has
been interpreted as proceeding through an initial π-complex
followed by deprotonation or rearrangement to a σ-bonded
complex (eqs 1 and 2). However, no direct evidence for the
π-complex is available using the electrochemical methods.
ArH+• + Nuc- h π-complex
π-complex f [ArHNuc]• σ-complex
(1)
(2)
In an attempt to provide evidence for the π-complex and to
extend the electrochemical studies, we began a time-resolved
spectroscopic evaluation of the kinetics of these systems using
nanosecond laser flash photolysis techniques. Our studies have
shown that the radical cations of 9-phenylanthracene (PA-H)
and 9,10-diphenylanthracene (DPA-H) are easily generated by
photoionization methods allowing their reactivity with many
electron rich species to be studied relatively simply.5,6 In one
of these earlier studies we examined the reactivity of these two
radical cations with a number of primary, secondary, and tertiary
amines.6 Primary amines, which are more difficult to oxidize
than PA-H and DPA-H, react with both radical cations by
nucleophilic addition to the 10-position, forming the corre-
sponding dihydroanthracenyl radicals with rates that depend on
the basicity of the amine and steric factors. Tertiary amines,
which were in general easier to oxidize than PA-H and DPA-
H, react by electron transfer. Secondary amines, whose
oxidation potentials were similar to the arylanthracenes, react
by both nucleophilic addition and electron transfer with
competitive rate constants. In the cases studied, we suggested
that the partitioning between the reaction pathways was
governed by the relative oxidation potential of the amines and
PA-H and DPA-H and by the factors that affect the rates of
nucleophilic addition. In none of the cases was direct evidence
for a π-complex observed.
Results and Discussion
Generation of Radical Cations. Previously we demon-
strated that irradiation of acetonitrile solutions of PA-H and
DPA-H using 355 nm laser pulses of a nanosecond laser flash
photolysis system generates the corresponding radical cations
PA-H•+ and DPA-H•+ by photoionization (eq 3).5,6 In the
present study, the radical cations PA-X•+ and DPA-X•+ were
generated from their corresponding aryl substituted anthracenes
(PA-X or DPA-X) in the same way. This method is advanta-
geous to study the spectroscopy and kinetics of the reactions
of the radical cations. Because a sensitizing electron acceptor
is not required, there is no interference from strongly absorbing
radical anions of those acceptors commonly used as sensitizers
(although the solvated electron is eventually captured to form
the anthracene anion radical). Only DPA-NO2 did not photo-
ionize under these conditions (or with 266 nm excitation);
however, DPA-NO2•+ could be generated using chloranil as an
electron-transfer sensitizer.
Using relative oxidation potentials to predict the reactivity
of the same radical cations with anions is complicated by the
lack of reliable standard potentials of anions in aqueous and
organic solution, which, unlike the delocalized aromatic radical
ions, are highly sensitive to solvent effects. In a preliminary
communication we found that azide,5 which is a nonselective
nucleophile used as a clock in carbonium ion chemistry,22 is a
powerful reducing agent in its reactivity with PA-H•+ and DPA-
H•+ (and indeed other radical cations) in acetonitrile and
proceeds by electron transfer. In contrast, in 2,2,2-trifluoro-
The transient absorption spectra of the radical cations
measured using oxygen saturated acetonitrile (MeCN) solutions
exhibit similar features: a relatively intense absorption band
centered at 315-320 nm, a weak absorption centered near 420
nm, and a broad, structured absorption band in the 500-750
nm range. Representative spectra, obtained from 1 × 10-4
M

Arylanthracene Radical Cations
J. Phys. Chem. A, Vol. 102, No. 32, 1998 6505
TABLE 1: Absorption Maxima of the Radical Cations of Para-Substituted 9-Aryl- (PA-X•+) and Bis-Para-Substituted
9,10-Diarylanthracene (DPA-X•+)a and the Oxidation Potentialsb of the Corresponding Substituted Arylanthracenes (PA-X and
DPA-X) and the Anions/Amines Used in This Study
compound
E°/Vc vs SCE
FE° (kcal mol-1
)
λ
max/nmd
nucleophile
E°/V vs SCE
Ep > 1.6g
PA-OCH3
PA-CH3
PA-H
PA-CF3
PA-CN
DPA-OCH3
DPA-CH3
DPA-H
1.236
1.268
1.300
1.352
1.386
1.183
1.237
1.270
1.394
f
28.5
29.2
30.0
31.2
32.0
27.2
28.5
29.3
32.1
315, 550, 600, 655
315, 590, 645, 710
315, 580, 650, 710
320, 540, 575, 650, 715
320, 540, 580, 655, 715
320, 570, 635, 670
320, 560, 600, 660,700
320, 550, 600, 650, 720
320, 550, 590, 655, 720
310, 600, 735
n-butylamine
DABCO
0.80g
acetate, AcO-
cyanide, CN-
1.78h
1.65i
-
azide, N3
0.78j
1.48i
DPA-CN
DPA-NO2
bromide, Br-
e
a Radical cations generated by 355 nm laser irradiation of ca. 1 × 10-4 M solution of PA-X or DPA-X in acetonitrile at 23 °C. b Measured in
dry acetonitrile/Et4NClO4 (0.1 M) at 25 °C. c Oxidation potentials measured vs. ferrocene oxidation (E°(Fc+/Fc) ) 0.448V vs SCE) Error is (0.005mV.
d Error is (5 nm. Values in bold face are the wavelengths of the absorptions with the highest extinction coefficients at the short and long wavelength
bands and are the wavelengths where the kinetics where monitored. e Radical cation generated using chloranil as a sensitizer. f Oxidation potential
not measured. g Reference 6. h Value from Eberson, L. Acta Chem. Scand. 1984, B38, 439 with appropriate correction. See: Eberson, L. In Electron
Transfer in Organic Chemistry; Springer-Verlag: New York, 1989. i Ref 20. Some of the values are reported in DMF and these are assumed to be
•
-
similar in acetonitrile. In ref 20b, the value for E°CN /CN is incorrectly listed; the value above is that determined by correcting the standard potential
in water using the appropriate values given in the paper. j Reference 5.
solutions of PA-OCH3, PA-CN, DPA-OCH3, and DPA-CN in
MeCN, are shown in Figure 1. The region between 350 and
400 nm is not shown because it is an area of considerable
bleaching (i.e., negative ∆OD) of the arylanthracene absorptions.
The wavelengths of the absorption maxima of each of the bands
observed for PA-X•+ and DPA-X•+ are summarized in Table
1. In nitrogen-saturated solution, the transient absorption spectra
are essentially identical except for an increase in absorption
intensity around 420-430 nm, which is attributable to a
contribution from the triplet states of PA-X and DPA-X.23 The
absorptions at 420-430 nm are the only observable transients
in nonpolar solvents, such as hexane (i.e., no photoionization
occurs), and in all cases these absorptions are quenched by the
addition of oxygen. The contribution to the absorption spectra
from the triplet is substrate dependent, typically larger for PA-X
than for DPA-X, consistent with the larger values for Φisc in
the former.24 To minimize interference from the triplet absorp-
tion, the kinetic studies described below were performed using
air or oxygen saturated solutions, which has no observable effect
Figure 1. Transient absorption spectra recorded 200-300 ns after 355
nm excitation of 1 × 10-4 M solutions of (a) PA-OCH3, (b) PA-CN,
(c) DPA-OCH3, and (d) DPA-CN in oxygen saturated acetonitrile
solutions at 23 °C.
on the kinetics or spectroscopy of the radical cations themselves.
PA-X (DPA-X)
9
hν8 PA-X•+ (DPA-X•+) + e-
(3)
solvated
The para-substituents on the 9- and 10-aryl rings of PA-X
and DPA-X affect the relative redox properties. The reversible
oxidation potentials (E°) of PA-X and DPA-X measured in dried
MeCN/Et4NClO4 (0.1M) solution measured versus an internal
ferrocene (Fc/Fc+) couple at 298 K are summarized in Table
1; when available they agree with those previously reported.19f
As expected, the E° values correlate with the Hammett σ+
parameter. Plots of (F/2.303RT)E° vs σ+ (for substituents on
the 9- and 10-aryl rings) give F+ values of 1.65 (r ) 0.981)
and 1.24 (r ) 0.989) for PA-X and DPA-X, respectively,
suggesting that there may be some saturation of the substituent
effects in the DPA-X system.
The kinetics for decay of the radical cations are most easily
monitored at wavelengths where the absorptions are most intense
(highlighted in boldface in Table 1). In the absence of any
added reactant, radical cations decay with approximately second-
order kinetics having half-lives of several microseconds. In the
less nucleophilic solvent, 2,2,2-trifluorethanol, the lifetimes of
the radical cations are increased by a factor of 10 or more. Upon
addition of the anions or amines to the solution, the lifetime of
the long wavelength absorptions are shortened and the decays
become first order. The lifetime of the short wavelength band
(320 nm) is also decreased under these conditions (as expected
for an absorption due to the same species); however, in some
cases the decay kinetics do not appear first order due to the
concomitant growth and subsequent decay of product absorp-
tions that overlap with the decay of the radical cation in this
region (vide infra).
Bimolecular rate constants for the reaction of PA-X•+ and
DPA-X•+ with a number of anions, which include bromide
(Br-), acetate (CH3CO2-), cyanide (CN-), and azide (N3-) and
the amines n-butylamine (n-BuNH2) and 1,4-diazabicyclo[2.2.2]-
octane (DABCO) were obtained by measuring the pseudo-first-
order rate constants for decay (kobs) of the radical cation
(typically monitoring the decay of the long wavelength band)
as a function of the concentration of added reactant. The
observed rate constants for decay increased linearly according
to eq 4, where k0 is the sum of the rate constants for decay of
the radical cation in the absence of quencher. The slopes of
the plots yield bimolecular rate constants (kR) for reactions of
PA-X•+ and DPA-X•+ (summarized in Table 2). Rate constants
for reaction of DPA-NO2 were not determined since many of

6506 J. Phys. Chem. A, Vol. 102, No. 32, 1998
Workentin et al.
TABLE 2: Second-Order Rate Constants for the Reaction
of PA-X•+ and DPA-X•+ with a Variety of Anions and
Aminesa
TABLE 3: Ratio of the Bimolecular Rate Constants,
kPA/kDPA, for the Reactions of PA-X•+ and DPA-X•+ with the
Various Anions and Aminesa
-
-
kR/109 M-1 s-1
kPA/kDPA n-butylamine DABCO CH3CO2
CN- Br- N3
n-butylamine DABCO CH3CO2 CN- Br- N3
X ) CH3O
X ) CH3
X ) H
110
77
62
1.3
1.1
1.1
2.0
363
11
7
213
7.5 14
35
1.2
1.1
1.1
-
-
PA-OCH3
PA-CH3
PA-H
PA-CF3
PA-CN
DPA-OCH3
DPA-CH3
DPA-H
0.66
1.0
1.8
1.7
2.2
0.0060
0.013
0.029
0.093
14
13
15
14
14
11
12
14
16
18
20
25
25
0.044
1.7
2.8
8.5 0.035 34
7
7
12
22
25
24
0.06
31
X ) CN
24
1.6
1.4 0.7 2.2
0.075 28
0.11b
0.12b
32
38
a Calculated from data taken from Table 2.
0.04 0.001 28
1.6 0.0044 29
3.1 0.011 25
the presence of 0.02 M of n-BuNH2 are shown in Figure 2. At
this concentration and on this time scale, the radical cation
(monitored at 655 nm) is almost completely quenched; the
increase in absorption at 315 nm due to the long-lived transient
is assigned to the corresponding dihydroanthracenyl radical
product (PA(Nuc)-X• or DPA(Nuc)-X•) that results from addition
of the amine (Nuc ) CH3(CH2)3NH2) to the 10-position of the
radical cation (eq 6). These species are characterized by an
absorption spectrum with a maximum near 330 nm (this
absorption maximum varies by (5 nm as a function of the
remote substituent). A transient absorption spectrum measured
on the same solution at a time after the signal due to the radical
cation is completely decayed in also shown in Figure 2. The
position of the absorption maximum and its reactive insensitivity
to oxygen is consistent with the known spectra and reactivity
of other triphenylmethyl type radicals.25 Similar spectral
features are observed for the reaction of DPA-X•+ with n-
BuNH2, although the rate constants are found to be 20-110
times slower depending on the substituent (see Table 2). We
have previously assigned this rate constant difference between
PA-X•+ and DPA-X•+ to the increased steric interaction in DPA
due to the extra substituent at the 10-position where addition is
expected to occur.
DPA-CN
7.0
16
17
0.17
17
a Measured in acetonitrile at 25 °C by monitoring the decay of the
radical cation. Errors in the rate constants are ca. 10-15%. b From
growth of the absorption of Br2•-. The overlapping growth of the
absorption of Br2 at long wavelengths made measuring the decay
kinetics of the radical cation difficult over a broad concentration range.
•-
the anions and the amines used in this study reacted readily in
a thermal reaction with the chloranil used as a photoinduced
electron transfer (PET) sensitizer. Rate constant ratios (kPA
/
k
DPA), comparing the reaction of PA-X•+ and DPA-X•+ with a
particular reactant are given in Table 3.
kobs ) k0 + kR[Nuc]
(4)
Reaction with Amines. In the presence of DABCO, the
kinetics of decay of both the long and short wavelength
absorptions of PA-X•+ and DPA-X•+ are equivalent. Further-
more, the bleach of the absorption of PA-X and DPA-X between
360 and 400 nm, which occurs as a result of the loss of its
absorption on photoionization and formation of the radical
cation, recovers completely with kinetics identical to that
measured for the decay of the radical cation. This indicates
that the reaction of these radical cations with DABCO does not
involve any net consumption of the aromatic precursor. These
kinetic and absorption spectral features are similar to those
described and illustrated in our previous study where the
reactivity of PA-H and DPA-H with other amines were
examined6 and are indicative of an electron-transfer reaction
from DABCO to the radical cation (eq 5). This was confirmed
in all cases by the concurrent growth of a transient absorption,
with a maximum at 480 nm, that can be assigned to DABCO•+.6
The rate constants of the reactions of DABCO with PA-X•+
and DPA-X•+ are all diffusion controlled, consistent with the
observation that electron transfer is calculated to be exergonic
by at least 8 kcal mol-1 for all systems. Furthermore, the rate
constant ratios (kPA/kDPA) are equal to 1, indicative of a process
insensitive to steric effects. The lack of variation in rate
constants can be used as a probe to differentiate between ET
and addition in these systems (vide infra).
A plot of log k versus FE° for the reactions of DPA-X•+ and
PA-X•+ with n-BuNH2 and DABCO is shown in Figure 3.
Examination of this figure and Table 2 show that the reaction
of n-BuNH2 with PA-X•+ is only slightly sensitive to the
electronic nature of the remote substituent; electron donating
substituents having slightly lower rate constants for reaction.
This limited electronic effect can be rationalized by the apparent
leveling off of the rate constants at approximately 2 × 109 M-1
s-1 as the substituent is made more electron withdrawing. Such
a leveling of the rate constants a full order of magnitude below
the diffusion-limited rate constant in this solvent (2 × 1010 M-1
s-1) has been previously observed for the reaction of amines of
varying basicity with PA-H•+,6 the reactions of substituted
styrene radical cations with similar amines4c,d and for the
reactions of diarylcarbocations with amines.26 If one considers
a reaction scheme in which the reactants form an intermediate
complex prior to formation of product, then this leveling may
result from the competition between dissociation of the complex
and product formation (a preequilibrium). Alternatively, we
kET
8
PA-X•+ (DPA-X•+) + DABCO
PA-X (DPA-X) + DABCO•+ (5)
In contrast, the reactivity of PA-X•+ and DPA-X•+ with
n-BuNH2 results in permanent bleaching (on the laser flash time
scale) of the absorptions due to PA-X and DPA-X. While the
kinetics for decay of the long wavelength absorptions are first
order in the presence of amine, the short wavelength absorptions
appear more complex due to overlapping growth of a long-
lived transient product that also absorbs in this region. Kinetic
traces for PA-OCH3•+, measured at 315, 385, and 655 nm in

Arylanthracene Radical Cations
J. Phys. Chem. A, Vol. 102, No. 32, 1998 6507
Figure 4. Transient absorption spectrum recorded measured 2 µs after
the laser pulse of a sample containing DPA-CN + 1 mM AcO-. Inset
shows kinetic traces measured on the same sample at 320 and 385 nm.
the growth of a product transient with spectral features
characteristic of the addition product dihydroanthracenyl radical,
(iii) kPA/kDPA values close to 1, and (iv) the spectral observation
of the oxidized form of the amine. Criteria used as an indicator
for a nucleophilic addition process include: (i) complete and
permanent bleaching of the radical cation precursor, (ii)
formation of a new long-lived transient dihydroanthracenyl
radical with maximum near 330 nm, (iii) kPA/kDPA values
significantly greater than 1, and (iv) reaction rate constants
significantly faster than those expected for ET based on Marcus
theory.
Figure 2. (a) Kinetic traces for PA-OCH3 + 0.02 M n-butylamine
measured at 320, 385, and 650 nm. (b) Transient absorption spectrum
measured on the same sample 2 µs after the laser pulse.
Reaction with Anionic Species. Acetate (AcO-) and
Cyanide (CN-). The kinetic and spectral characteristics of the
reaction of acetate with PA-X•+ and DPA-X•+ are similar to
those observed for the reaction with n-BuNH2. Addition of
acetate results in a product whose absorption spectrum (Figure
4) is consistent with addition (i.e., triarylmethyl-type radicals
corresponding to PA(OC(O)CH3)-X• or DPA(OC(O)CH3)-X• ).
Furthermore, there is complete and permanent bleaching of the
absorption of the radical cation precursors (inset in Figure 4).
The rate constants for PA-X•+ with acetate are insensitive to
electronic factors (all within a factor of less than 2 of each other)
and all near the diffusion limit, >1010 M-1 s-1. The reduction
in rate constants expected for DPA-X•+ due to steric factors is
observed, except for the most reactive DPA-CN•+, which reacts
close to the diffusion controlled rate. In all cases, the rate
constants are several orders of magnitude faster than those
expected for outer sphere ET. For easy comparison, a plot of
log k vs FE° for the reactions of DPA•+ with all anions is shown
in Figure 5.
Figure 3. Plot of log(k) versus FE° for the reactions of PA-X•+ (solid
symbols) and DPA-X•+ (hollow symbols) with n-butylamine (circles)
and DABCO (triangles). Data from Table 2. Lines drawn through data
for DABCO are at the diffusion limit. The lines drawn for PA-X with
n-butylamine are arbitrary, but meant to show the leveling of the rate
constants observed at 2 × 109 M-1 s-1
.
previously suggested that in these reactions of hindered radical
cations with an amine that the reaction may be entropically
controlled and not every diffusional encounter would result in
reaction.6 Because both mechanistic possibilities would exhibit
similar activation parameters to those determined for these
reactions (small apparent activation energies and low preexpo-
nential factors) and because there has been no direct spectral
evidence for the formation of an intermediate complex, we are
unable to differentiate between these two possibilities. The
reaction of DPA-X•+ is more sensitive to the electronic effects
of the remote substituent as expected for the lower overall
reactivity observed in these systems with n-BuNH2 due to
increased steric interaction at the 10-position. If we account
for leveling of the rate constants with n-BuNH2 with PA-X•+
then, for a given E° value, DPA-X•+ reacts approximately 2
orders of magnitude more slowly than PA-X•+.
In principle, main features of the reactions of the radical
cations with n-BuNH2 and DABCO can be used to differentiate
between addition and ET in the initial interaction of the same
radical cations with anionic species. Criteria used to identify
ET include: (i) complete recovery of the initial bleach of the
radical cation precursors with rate constants identical to the rate
constant for decay of the radical cation, (ii) no observation of
The spectral features observed for the reaction of the radical
cations in the presence of cyanide also appear to be those
expected for addition. For example, concomitant with the decay
of the radical cation is the formation of a new long-lived
transient whose absorption spectrum is similar to that observed
with acetate or n-butylamine as the reactant (Figure 6) and the
k
PA/kDPA values show the same steric effect as is observed for
-
the reactions with CH3CO2 (i.e., approaching a value of 1.0
for the most electron deficient aromatics where the reaction
becomes diffusion controlled in both systems). Furthermore,
the bleach does not recover with kinetics identical to the decay
of the radical cation and all of the rate constants are, again,
several orders of magnitude larger than those expected for a
simple outer sphere ET.
Interestingly, for PA-X the absorption of the initial product,
which we assign to the 10-cyano-9,10-dihydroanthracen-9-yl

6508 J. Phys. Chem. A, Vol. 102, No. 32, 1998
Workentin et al.
X•) is a weakly bound complex that is in equilibrium with PA-X
(DPA-X) and CN• (eqs 7 and 8, Nuc ) CN); the latter then can
•-
combine with excess CN- in solution to form (CN)2 (eq 9,
Nuc ) CN), presumably a diffusion controlled reaction.30
PA-X•+ + Nuc- h PA(Nuc)-X•
PA(Nuc)-X• h PA-X + Nuc•
(7)
(8)
(9)
Nuc• + Nuc- f (Nuc)2
•-
The second alternative is a reaction in which the initial
addition (k7) is reversible (k-7) and competes with a slow,
endergonic outer sphere electron-transfer reaction (eq 10); the
CN• formed in this slow reaction then reacts with excess CN-
at the diffusion controlled rate (eq 9).
Figure 5. Plot of log(k) versus E° F for the reactions of DPA-X•+
-
-
with Br- (b), CN-(2), CH3CO2 (O), and N3 (4). Data from Table
2.
PA-X•+ + Nuc- f PA-X + Nuc•
(10)
Both of these alternatives require that the PA(CN)-X• adduct
radical be only weakly bound so that eqs 7 and 8 are reversible
on the experiments time scale. The C-CN bond energy for
PA(CN)-X• can be roughly estimated as the difference between
the C-CN bond energy in 10-cyano-9-phenyl-9,10-dihydroan-
thracene (ca. 103 kcal mol-1, based on group additivities) and
the triplet energy of 9-phenylanthracene (42 kcal mol-1). The
value of 61 kcal mol-1, even given substantial error in the
approximation, is too strong for the addition to be expected to
be reversible (eq 7) or for homolysis (inner sphere ET, eq 8) to
occur. This simple approximation gives the correct value for
the C-H bond of the 9,10-dihydroanthracen-9-yl radical (ca.
40 kcal mol-1) and what appears to be reasonable values for
Figure 6. Transient absorption spectra measured for the reaction of
•+
PA-OCH3 with 0.015 M CN-: (b) 100 ns following laser pulse,
showing spectrum assigned to PA-OCH3(CN)•; (O) 4 µs after laser
pulse, showing spectrum assigned to (CN)2•-. Inset shows the kinetic
traces measured on the same sample 1 µs after the laser pulse.
-
the PA(Y)-X• radicals for Y ) CH3CO2 (25 kcal mol-1) and
Y) Br (13 kcal mol-1). The preceding thermochemical
assessment was based on the assumption that the addition
resulted in the formation of a C-CN bond. In principle,
formation of a C-NC bond cannot be ruled out. In fact, the
C-NC bond is significantly weaker than the C-CN bond (by
ca. 23 kcal mol-1). However, that would still leave a bond
radical, is quenched by CN-. This is most noticeable at high
CN- concentrations (higher than is necessary for complete
quenching of the radical cation). The bimolecular rate constant
for quenching of PACN-X• by CN- is 1 × 108 M-1 s-1 and is
relatively insensitive to the substituent on PA-X. What is even
more perplexing is that the decay of PACN-X•, monitored at
320 nm, is accompanied by the return of the initial bleach of
the PA-X and growth of a new product with a relatively weak
absorption and a maximum around 420 nm (Figure 6). This
with a dissociation energy on the order of 40 kcal mol-1
.
The large bond energies appear to rule out an inner sphere
(i.e., addition/homolysis) or reversible process. Thus the only
remaining option is a bimolecular reaction between CN- and
PA(CN)-X•. While the conclusion is inescapable, the reaction
itself seems to be unprecedented. Two possibilities consistent
with the apparent steric effect are direct reaction of CN- with
the σ-bonded nitrile or isonitrile (PA(CN)-X• or PA(NC)-X•).
While the reaction with an isonitrile seems more plausible, we
have no evidence that favors one intermediate over the other.
This SN2-like mechanism is consistent with the apparent lower
reactivity of the DPA(CN)-X• radical which would be entropi-
cally inhibited since substitution would be accompanied by
flattening of an sp3 carbon which requires rotation of the aryl
ring. Furthermore, as pointed out above, these results also are
consistent with the apparent DPA mediated electrochemical
oxidation of CN-.28,29
new absorption is similar to the reported spectrum of the
•- 27
pseudohalogen dimer, (CN)2
.
In contrast, even at high CN-
concentrations, the initial addition adduct observed for the
reaction with DPA(CN)-X•+ is not quenched by CN- under our
reaction conditions. However, earlier studies by Evans and
Blount28 and Papouchado, Adams, and Feldberg29 have shown
that the product of the reaction of electrochemically generated
DPA•+ with CN- is consistent with reduction of the radical
cation (greater than 95% DPA recovered) and not cyanide
addition. Thus, we expect that, on time scales longer than were
investigated here, the recovery of DPA and overall reduction
of CN- should be observed. Unfortunately, we are unable to
observe the recovery with the techniques available in our
laboratory.
This observation raises an important question regarding how
electron-transfer products result from a bimolecular reaction
with a putative σ-bonded radical intermediate (i.e., C-Nuc).
In addition, the mechanism must account for the difference in
apparent reactivity between PA(CN)-X• and DPA(CN)-X•.
There are at least three possibilities to rationalize the results.
The first is a formal inner sphere ET in which the initially
proposed σ-bonded radical product (PA(CN)-X• or DPA(CN)-
Why does the same reaction not occur after addition of
CH3CO2-? In the case of acetate the initially formed radical
intermediate (PA(OC(O)CH3)-X•) appears to be more weakly
bonded than the nitrile analogue or isonitrile analogue (the C-O
bond energy is estimated to be 28 kcal mol-1, see above).
Estimated bond dissociation energies of Et-OC(O)CH3 (85 kcal
mol-1), Et-CN (120 kcal mol-1), and Et-NC (98 kcal mol-1 31
)
also are in agreement with this order. As was the case for CN-,
the addition is expected to be completely irreversible so the
more simple inner sphere mechanisms (eqs 7-9, Nu )

Arylanthracene Radical Cations
J. Phys. Chem. A, Vol. 102, No. 32, 1998 6509
2 of the experimental values. Thus, for DPA-X, the variation
in rate constant as a function of substituent simply reflects the
change in the driving force. This distinguishes the reactions
of bromide from the reactions of n-BuNH2, cyanide, and acetate
which react many orders of magnitude faster than the predicted
outer sphere electron-transfer rate.
The variation in rate constant for reaction of bromide with
PA-X•+ is smaller than predicted, based on a simple outer sphere
mechanism. This decreased sensitivity to substituent and the
source of the apparent steric effect may point to some inner
sphere character in these reactions. In an attempt to observe
the formation of the Br•/PA-X (DPA-X) π-complex, Br atoms
were generated in the presence of PA-X and DPA-X by
photolysis of R-bromoacetophenone (eq 11). No evidence for
π-complex formation was obtained. What was observed was
the growth of the absorption due to formation of the radical
cation of the aromatics, formed by an exothermic ET (k-8, eq
8, Nuc ) Br). The rate constant k-8 was determined to be 3 ×
1010 M-1 s-1 (i.e., diffusion controlled) by monitoring the
growth of the radical cation as a function of [PA-X] or [DPA-
X].
•-
Figure 7. Transient absorption spectrum of Br2 measured from
photolysis of an air-saturated solution of 1 × 10-4 M PA-CN in the
presence of 0.02 M Br- and recorded 2 µs after the laser pulse. Inset
shows kinetic traces showing decay of PA-CN•+ (measured at 715 nm)
and the growth of Br2•- (measured at 385 nm) measured in the presence
of 0.02M Br-.
CH3CO2-) can be ruled out. Acetate is more difficult to oxidize
than CN- so the overall energetics for the electron transfer,
including a competing outer sphere mechanism (eq 10) are less
favorable than for CN-. Furthermore, the formal dimer (X2•-),
which stabilizes the halogens and pseudohalogens, lies on a
completely dissociative surface in the case of acetate (i.e., the
formal dissociative reduction product of acetyl peroxide). Thus,
the dimerization that provides an additional driving force for
cyanide is not available to acetate. In addition, the resulting
carbonyloxyl radical is known to decarboxylate on the nano-
second time scale, so methylated products should be isolated
preparatively if any of these processes should occur. Parker
has shown that acetate reacts with PA-H•+ to yield formal acetate
addition products.18b
hν
•
C6H5COCH2Br
9
8 C6H5COCH2 + Br•
(11)
These experiments together with the ones summarized in
Table 2 provide both the forward and reverse rate constants for
reaction 10 allowing ∆G10 values to be calculated (∆G10 ) -RT
ln K10 ) -F∆E°). These values agree well with those calculated
from ∆E° values (E°(Br-) - E°(PA-X) ) ∆G10/23.06) obtained
from electrochemical measurements. For example, the ∆G10
value for PA-H obtained from the K10 value is 3.6 kcal mol-1
compared to the value of ∆E° from cyclic voltammetry of 4.1
kcal mol-1. Similarly, for DPA-H the calculated and experi-
mental values are 4.7 and 4.8 kcal mol-1, respectively.
Considering the close agreement between the experimental and
estimated rate constants (within a factor of 2 for DPA-X and a
factor of 10 for PA-X from Marcus theory) and the ∆E° values
(within (50 mV for all but one; PA-OCH3 which is within 70
mV), there is no need to invoke the intermediacy of a π-complex
in this reaction.
The results of electron-transfer reaction of these aromatic
radical cations with Br- is in sharp contrast to that of the
nucleophilic substitution reported by Schepp and Johnston for
the reaction of substituted styrene radical cations with Br- in
the same solvent, even though ET is thermodynamically more
favorable in those systems than in the system described here
(although still mainly slightly endothermic, depending on
substituents).4a This difference is likely explained by the
differences in the C-Br bond strength in the two different type
of radical products; in their case, addition generates a â-bro-
Bromide. In the presence of Br- the decay of the radical
cations is accompanied by a complete return of the bleach and
a growth of a new long-lived transient in the same region. In
all cases, the growth kinetics measured at 385 nm matched the
decay of the radical cations at all concentrations of Br-.
Representative kinetic traces and an absorption spectrum
obtained in the presence of sufficient Br- to quench the radical
cation are shown in Figure 7. The spectrum, characterized by
a λmax at 365 nm, is identical to the spectrum observed by
Scaiano et al on irradiation of vicinal dibromides in the presence
of Br- and is readily assigned to the well-known radical anion
•- 32
dimer, Br2
.
The latter is formed by the reaction of Br• with
excess Br- in solution, which is known to occur with a rate
constant of 1.6 × 1010 M-1 s-1 in MeCN with an equilibrium
constant of 105 M-1 32
.
Based on the relative redox potentials the reaction with
bromide ion might be predicted to be by nucleophilic addition
to yield the σ-bonded dihydroanthracenyl radical type product
[PA(Br)-X]• or [DPA(Br)-X]•. However, this reaction is iso-
electronic with the reaction of Br• with the neutral PA-X or
DPA-X which is known to favor the formation of a π-complex.33
Thus if the ET mechanism is to be an inner sphere process, as
discussed for CN- (eq 7-9, Nuc ) Br) and discussed previously
for addition of Br- to 1,3-dioxole radical cations,15a it must
proceed via the π-complex. Our spectroscopic evidence is
consistent with an ET process with rate determining endothermic
electron transfer followed by Br atom reaction with excess Br-
as shown in eqs 9 and 10 (Nuc ) Br).
mophenethyl radical with an equilibrium constant (kadd/k-add
)
of 5000 M in MeCN.4a
Azide. Reaction of N3- with our radical cations also exhibit
the spectral and kinetic features expected for an electron-transfer
reaction: complete recovery of bleach with identical kinetics
to the decay of the radical cation (Figure 8) and no observation
of a long-lived radical addition product at 320 nm. Azide had
often been assumed to react with radical cations by nucleophilic
addition by analogy to the reactivity of unstable carbonium ions.
This assumption may have been due in part to some inconsis-
tencies in the reported values for the oxidation potential of azide
in solution. However, our recent work5 and that of Wendt34
has made it clear that in acetonitrile azide has a relatively low
oxidation potential (0.775 V vs SCE) and will act as an efficient
Interestingly, the steric effects (suggested by the kPA/kDPA
values) are usually indicative of an addition mechanism.
However in this case the rate constants for outer sphere electron
transfer to DPA-X•+ from Marcus theory are within a factor of

6510 J. Phys. Chem. A, Vol. 102, No. 32, 1998
Workentin et al.
cenyl radical with maximum near 330 nm, and (iii) reaction
rate constants significantly faster than those expected for ET
based on Marcus theory. In some cases the rate constant ratio
(kPA/kDPA) can also be diagnostic. For example, (i) if kPA/kDPA
)
1 and the rate constants are less than diffusion controlled then
the reaction must be ET or (ii) if kPA/kDPA * 1 and one of the
values is diffusion controlled, then the reaction must be an
addition. Alternatively, if kPA/kDPA ) 1 and both rate constants
are diffusion controlled or if kPA/kDPA > 1 and neither rate
constant is diffusion controlled, then other criteria must be used
to decide the mechanism. Examples illustrating these criteria
are found in Tables 2 and 3.
•-
Figure 8. Transient absorption spectrum of N6 measured from
Only n-BuNH2, acetate, and cyanide gave direct spectroscopic
evidence for the initial formation of the addition to the radical
cation at the 10 position forming the dihydroanthracenyl radical.
In these cases, the trends in the rate constants for the addition
exhibited the usual factors expected for addition reactions: steric
effects illustrated by the kPA/kDPA and the effect of the remote
substituent altering the electrophilicity of the cationic center of
the radical cation, especially in DPA-X, where the rate constants
were lower than the diffusional limit due to the steric effect.
Remote substituents tend to modulate the electrophilicity and,
thus, the nucleophilic reactivity of the radical cation (i.e., the
reactivity increases as the electron demand of the substituent
increases). This modulation of reactivity by the substituent is
also apparent for the endergonic ET reactions with Br- where
the substituent acts to alter the redox property of the PA-X
(DPA-X) and thus the driving force for ET. In no case was
there direct spectroscopic evidence for the formation of a
π-complex prior to the formation of the C-X bond. However,
it is important to note that this does not exclude complex
formation as the first step in the reaction.
photolysis of an air-saturated solution of 1 × 10-4 M DPA-OCH3 in
the presence of 6 mM azide and recorded 1 µs after the laser pulse.
•+
Inset shows kinetic traces showing decay of DPA-OCH3 (measured
at 650 nm) and the complete recovery of the initial bleach of the
absorption of DPA-OCH3 (measured at 385 nm) measured in the
presence of 0.5 mM azide.
reducing agent with all radical cations with an E° greater than
ca.0.8 V vs SCE.5 In all cases presented here the driving force
for electron transfer (E°(N3-) - E° (PA-X•+ or DPA-X•+) is at
least 9 kcal mol-1 exergonic. In addition to the kinetic evidence
we also obtain spectroscopic evidence for the formation of N3 .
The absorption of the azidyl radicals produced by the ET is not
•
•
observed directly, however N3 reacts with excess azide at the
diffusion controlled rate to form the unusual pseudo-halogen
radical anion dimer N6•- by a reaction strictly analogous to the
case of bromide.5,35,36 Unlike the formation of Br2•-, the growth
of N6•- is not observed because the equilibrium constant for its
formation is about 3 orders of magnitude lower than that for
Br2•- and is not a strong enough absorption to observe at room
-
temperature at the low concentrations of N3 required for
For endergonic electron-transfer processes the stability of the
dihydroanthracenyl radical adduct can be estimated from the
complete quenching of the radical cation.35 However, at higher
concentrations a strong absorption is observed at 700 nm and
difference between the triplet energy of PA-H (42 kcal mol-1
)
•-
has unequivocally been assigned to N6 (Figure 8).35
and the C-Nuc bond energy in the 10-Nuc-9-phenyl-9,10-
dihydroanthracene. This bond energy can often be estimated
using group additivities. For example, in the case of acetate,
the bond energy is estimated to be ≈25 kcal mol-1 (eq 8). Thus
the adduct is stabilized by about 15 kcal mol-1 with respect to
the ion pair (eq 7, ∆H7 ≈ ∆H8 - F∆E°).
Summary
Our results indicate that the mechanism for the initial
interaction for the radical cation/nucleophile combination can
be described by the general mechanism illustrated in Scheme
2. The initial partitioning between addition and ET can often
be predicted by the relative redox potentials of the radical cation
and the anion or amine. When these values are known they
provide an estimate for the driving force for ET. When the
driving force for ET is exergonic, this pathway will dominate.
This was illustrated by the reactions of PA-X•+ and DPA-X•+
with DABCO and N3-. The reverse is not true; in other words,
if the driving force for ET is endergonic, addition is not
necessarily the preferred reaction pathway. In this case other
criteria including steric factors and the strength of the incipient
bond become important. Thus, using PA-X and DPA-X as
chemical probes, we have been able to establish simple criteria
to delineate one mechanism from another. For electron-transfer
one should observe (i) complete recovery of the initial bleach
of the radical cation precursors with rate constants identical to
the rate constant for decay of the radical cation, (ii) absence of
the growth of a product transient with spectral features
characteristic of the addition product dihydroanthracenyl radical,
and (iii) the spectral observation of the oxidized form of the
anion/amine (when a clear spectroscopic signature is available).
For nucleophilic addition one should observe the following: (i)
complete and permanent bleaching of the radical cation precur-
sor, (ii) formation of a new long-lived transient dihydroanthra-
Photoinduced electron transfer and electrochemical methods
have been used to successfully prepare substitution products
from addition of CN-,37 acetate,18b amines,38 and other electron
rich species to aromatic radical cations.39 The efficiency of the
reaction to yield addition products varies dramatically as a
function of the aromatic precursor and the reaction conditions.
This is generally explained by the redox potentials of the
reaction partners involved and by the relative nucleophilicities.
However, the present results (especially those with CN-) caution
against these generalizations. Ultimately, the products observed
in these radical cation/nucleophile reactions will not only depend
on the mechanism of the initial interaction, but also on the fate
of the initial formed intermediates, which can either be the
radical adduct formed on addition or the aromatic radical cation
precursor and the oxidized form of the anion/amine formed by
ET.
The primary purpose of this study was to understand the
factors that control the competition between addition and
electron transfer in the initial interaction between an electron
rich species with an aromatic radical cation. Our studies have
shown that knowledge of these processes and their rate constants
do not necessarily allow one to design appropriate conditions
to maximize yields in preparative reactions. Clearly, any

Arylanthracene Radical Cations
J. Phys. Chem. A, Vol. 102, No. 32, 1998 6511
SCHEME 2
synthetic strategy requires kinetic and mechanistic information
for the reaction pathways of the intermediates formed in the
initial reaction. Furthermore, the results illustrate the complexi-
ties of comparing of the reactivity of olefin radical cations to
aromatic radical cations; the reactivity does not only depend
on relative redox potentials. Mechanistic studies continue to
provide insights into the reactivity of radical cations with
electron rich species. It is clear from our studies much work
needs to be done before generalizations, such as those used in
carbocation chemistry, can be made with radical cations. This
has important ramifications particularly in that the development
of “clock” reactions in radical cation chemistry.
the sample by microliter syringe as aliquots of an appropriate
solution of the quencher in MeCN or as the neat compound.
Cyclic Voltammetry. Electrochemical measurements were
made using a standard three electrode single cell arrangement
and an E. G. & G. model 173 Potentiostat interfaced to a
Tektronix digital oscilloscope. In our arrangement the pseudo
reference electrode was a Ag wire in electrolyte solution, the
working electrode was a glassy carbon disk (0.5 mm), and a
platinum wire was used as the counter electrode. Samples were
purged with argon while stirring prior to data acquisition.
Measurements were made on 1-2 mM samples in MeCN
containing 0.1 M tetraethylammonium perchlorate as the sup-
porting electrolyte. Peak potentials were referenced internally
to the ferrocene (Fc+/Fc) couple and converted to SCE
(E°(Fc+/Fc) ) 0.448 V vs SCE).
Experimental Section
The compounds 9-phenyl- and 9,10-diphenylanthracene were
commercially available from Aldrich. The remaining 9-aryl-
and 9,10-diarylanthracenes were prepared by methods reported
separately.24 Each were purified by column chromatography
and repeated recrystallizations and were verified to be of high
purity using standard analytical methods. In addition, all
compounds were of sufficient purity that time-resolved fluo-
rescence measurements exhibit a single-exponential decay.24
The tetrabutylammonium salts of cyanide, acetate, bromide,
(all from Aldrich), chloride and azide (TCI) were used as
received. DABCO and n-butylamine (Aldrich) were purified
prior to use. Acetonitrile (BDH, OmniSolv) and 2,2,2-trifluo-
roethanol were used as received for the NLFP experiments
unless otherwise specified. For the electrochemical measure-
ments MeCN was fractionally distilled over calcium hydride
prior to use. Tetraethylammonium perchlorate was recrystal-
lized 3 times from ethyl acetate:hexane (9:1) and dried prior to
use.
Acknowledgment. V.D.P. thanks the National Science
Foundation (Grant CHE-9708935) for financial support and
M.S.W. thanks the Natural Sciences and Engineering Research
Council of Canada for financial support. The authors also thank
Dr. L. J. Johnston (NRC) for use of the LFP facility.
References and Notes
(1) Issued as NRCC publication #40881.
(2) (a) Schmittel, M.; Burghart, A. Angew. Chem., Int. Ed. Engl. 1997,
36, 2550-2589. (b) Fox, M. A., Chanon, M., Eds. Photoinduced Electron
Transfer, Elsevier: Amsterdam, 1988; Parts A-D. (c) Roth, H. D. Top.
Curr. Chem. 1992, 163, 133-245.
(3) (a) Yoon, U.-C.; Givens, R. S. In AdVances in Electron-Transfer
Chemistry; Mariano, P. S., Ed.; JAI Press, Inc.: New York, 1994; Vol. 4.
(b) Bauld, N. L. In AdVances in Electron-Transfer Chemistry; Mariano, P.
S., Ed.; JAI Press Inc.: New York, 1992; Vol. 2. (c) Lewis, F. D. In
AdVances in Electron-Transfer Chemistry; Mariano, P. S., Ed.; JAI Press,
Inc.: New York, 1996; New York, 1996; Vol. 5. (d) Cossy, J.; Pete, J.-P.
In AdVances in Electron-Transfer Chemistry; Mariano, P. S. Ed.; JAI Press,
Inc: New York, 1996; Vol. 5.
Laser Flash Photolysis. Nanosecond laser flash photolysis
experiments employed the pulses from a Lumonics HY750
Nd3+: YAG laser (355 nm, 20 mJ/pulse, 20 ns pulse width)
and a computer controlled detection system that has been
described elsewhere.40 Solutions of PA-X and DPA-X were
prepared at concentrations such that the absorbance at the
excitation wavelength was 0.4-0.5. Samples were contained
in rectangular 7 × 7 mm2 Suprasil quartz cells. Transient
absorption spectra were measured employing a flow system
which ensured a fresh volume of sample was irradiated by each
laser pulse. Experiments using nitrogen or oxygen saturated
solutions were prepared by bubbling a dry stream of the
appropriate gas through the solution for at least 10 min.
Quenching rate constants were measured on static samples, at
ambient temperature (23-25 °C) on air saturated solutions
unless specified otherwise. Quenchers were added directly to
(4) (a) Johnston, L. J.; Schepp, N. P. J. Am. Chem. Soc. 1993, 115,
6564-6571. (b) Schepp, N. P.; Johnston, L. J. J. Am. Chem. Soc. 1994,
116, 6895-6903 (c) Johnston, L. J.; Schepp, N. P. Pure Appl. Chem. 1995,
67, 71-78. (d) Johnston, L. J.; Schepp, N. P. In AdVances in Electron-
Transfer Chemistry; Mariano, P. S., Ed.; JAI Press, Inc.: New York, 1996;
Vol. 5, pp 41-102. (e) Schepp, N. P.; Johnston, L. J. J. Am. Chem. Soc.
1994, 116, 10330-10331. (f) Schepp, N. P.; Johnston, L. J. J. Am. Chem.
Soc. 1996, 118, 2872-2881.
(5) Workentin, M. S.; Schepp, N. P.; Johnston, L. J.; Wayner, D. D.
M. J. Am. Chem. Soc. 1994, 116, 1141-1142.
(6) Workentin, M. S.; Johnston, L. J.; Wayner, D. D. M.; Parker, V.
D. J. Am. Chem. Soc. 1994, 116, 8279-8287.
(7) (a) Dinnocenzo, J. P.; Farid, S.; Goodman, J. L.; Gould, I. R.; Todd,
W. P.; Mattes, S. L. J. Am. Chem. Soc. 1989, 111, 8973-8975. (b)
Dinnocenzo, J. P.; Todd, W. P.; Simpson, T.; Gould, I. J. Am. Chem. Soc.
1990, 112, 2462-2464. (c) Todd, W. P.; Dinnocenzo, J. P.; Farid, S.;
Goodman, J. L.; Gould, I. R. Tetrahedron Lett. 1993, 34, 2863.
(8) (a) Maslak, P.; Vallombroso, T. M.; Chapman, W. H., Jr.; Narvaez,
J. N. Angew. Chem., Int. Ed. Engl. 1994, 33, 73-75. (b) Maslak, P. Top.
Curr. Chem. 1993, 168, 1-46 and references therein.

6512 J. Phys. Chem. A, Vol. 102, No. 32, 1998
Workentin et al.
(9) Sankararaman, S.; Perrier, S.; Kochi, J. K. J. Am. Chem. Soc. 1989,
111, 6448-6449.
(10) (a) Masnovi, J. M.; Sankararaman, S.; Kochi, J. K. J. Am. Chem.
Soc. 1989, 111, 2263-2276. (b) Bobrowski, K.; Das, P. K. J. Phys. Chem.
1985, 89, 5079-5085.
(11) (a) Anne, A.; Hapiot, P.; Moiroux, J.; Neta, P.; Save´ant, J.-M. J.
Phys. Chem. 1991, 95, 2370-2377. (b) Anne, A.; Hapiot, P.; Moiroux, J.;
Neta, P.; Save´ant, J.-M. J. Am. Chem. Soc. 1992, 114, 4, 4694-4701.
(12) Baciocchi, E.; Del Giacco, T.; Elisei, F. J. Am. Chem. Soc. 1993,
115, 12290-12295.
(13) Neiminen, K.; Niiranen, J.; Lemmetyinen, H.; Sychtchikova, I. J.
Photochem. Photobiol. A 1991, 61, 235-243.
(14) Brede, O.; David, F.; Steenken, S. J. Chem. Soc., Perkin Trans. 2
1995, 23-32.
(15) (a) Trampe, G.; Mattay, J.; Steenken, S. J. Phys. Chem. 1989, 93,
7157-7160. (b) Steenken, S.; Warren, C. J.; Gilbert, B. C. J. Chem. Soc.,
Perkin Trans. 2, 1990, 335-342.
(16) Koike, K.; Thomas, J. K. J. Chem. Soc., Faraday Trans. 1992, 88,
195-200.
Kanagasabapathy, V. M.; Banait, N. S.; Steenken, S. J. Am. Chem. Soc.
1991, 113, 1009-1104 and references cited.
(23) Carmichael, I.; Helman, W. P.; Hug, G. L. J. Phys. Chem. Ref.
Data 1987, 6, 239.
(24) Morkin. T. L.; Magri, D.; Workentin, M. S. Manuscript in
preparation.
(25) Chatgilialogula, C. In Handbook of Photochemistry; Scaiano, J. C.,
Ed.; CRC Press: Boca Raton, FL, 1989; Vol. II, Chapter 1.
(26) McClelland, R. A.; Kanagasabapathy, V. M.; Banait, N. S.;
Steenken, S. J. Am. Chem. Soc. 1992, 114, 1816-1823.
(27) (a) Draganic, I. G.; Draganic, Z. D.; Holroyd, R. A. J. Phys. Chem.
1971, 75, 608-612. (b) Martins, L. J. A. J. Chem. Soc., Faraday Trans 1
1982, 78, 533-543.
(28) Evans, J. F.; Blount, H. N. J. Am. Chem. Soc. 1978, 100, 4191-
4196.
(29) Papouchado, L.; Adams, R. N.; Feldberg, S. W. J. Electroanal.
Chem. 1969, 21, 408.
(30) Nagarajan, V.; Fessenden, R. W. J. Phys. Chem. 1985, 89, 2330-
2335.
(31) Griller, D.; Kanabus-Kaminska, J. M.; Maccoll, A. J. Mol. Struct.
(THEOCHEM) 1988, 163, 125.
(32) Scaiano, J. C.; Barra, M.; Krzywinski, M.; Sinta, R.; Calabrese, G.
J. Am. Chem. Soc. 1993, 115, 8340-8344.
(33) Raner, K. D.; Lusztyk, J. Ingold, K. J. Phys. Chem. 1989, 93, 564-
570.
(34) Plazk, V.; Wendt, H. Ber. Bunsen-Ges. Phys. Chem. 1979, 83, 481-
495.
(17) (a) Parker, V. D. Acc. Chem. Res. 1984, 17, 243-250. (b)
Hammerich, O.; Parker, V. D. AdV. Phys. Org. Chem. 1984, 20, 55-189.
(18) (a) Parker, V. D.; Tilset, M. J. Am. Chem. Soc. 1987, 109, 2521-
2523. (b) Reitsto¨en, B.; Norrsell, F.; Parker, V. D. J. Am. Chem. Soc. 1989,
111, 1, 8463-8465. (c) Parker, V. D.; Reitsto¨en, B.; Tilset, M. J. Phys.
Org. Chem 1989, 2, 580-584. (d) Wang, H.; Zheng. G.; Parker, V. D.
Acta Chem. Scand. 1995, 49, 311-312.
(19) (a) Reitsto¨en, B.; Parker, V. D. J. Am. Chem. Soc. 1991, 113, 6954-
6958. (b) Norrsell, F.; Handoo, K. L.; Parker, V. D. J. Org. Chem. 1993,
58, 4929-4932. (c) Reitsto¨en, B.; Parker, V. D. Acta Chem. Scand. 1992,
46, 464-468. (d) Parker, V. D.; Chao, Y.; Reitsto¨en, B. J. Am. Chem. Soc.
1991, 113, 2336-2338. (e) Reitsto¨en, B.; Parker, V. D. J. Am. Chem. Soc.
1990, 112, 4968-4969. (f) Parker, V. D.; Pedersen, M.; Reitsto¨en, B. Acta
Chem. Scand. 1993, 47, 560-564. (g) Xue, J.; Parker, V. D. J. Org. Chem.
1994, 59, 6564-6566. (h) Parker, V. D.; Chao, Y. T.; Zheng, G J. Am.
Chem. Soc. 1997, 119, 11390-11394.
(20) (a) Wayner, D. D. M. In Handbook of Photochemistry; Scaiano, J.
C., Ed.; CRC Press: Boca Raton, 1989; Vol. II, Chapter 18. (b) Save´ant,
J.-M. J. Phys. Chem. 1994, 98, 3716-3724.
(21) (a) Shaik, S.; Pross, A. J. Am. Chem. Soc. 1989, 111, 4306. (b)
Shaik, S. Prog. Phys. Org. Chem. 1985, 15, 197.
(35) Workentin, M. S.; Wagner, B. D.; Negri, F.; Zgierski, M.; Lusztyk,
J.; Siebrand, W.; Wayner, D. D. M. J. Phys. Chem. 1995, 99, 94-101.
(36) Workentin, M. S.; Wagner, B. D.; Lusztyk, J.; Wayner, D. D. M.
J. Am. Chem. Soc. 1995, 117, 119-126.
(37) (a) Yasuda, M.; Pac, C.; Sakurai, H. J. Chem. Soc., Perkin Trans.
1 1981, 746-750. (b) Andreades, S.; Zahnow, E. W. J. Am. Chem. Soc.
1969, 91, 4181-4190. (c) Cauquis, G.; Serve, D. Bull. Soc. Chim. France
1979, 2, 145.
(38) (a) Yasuda, M.; Matsuzaki, Y.; Shima, K.; Pac, C. J. Chem. Soc.,
Perkin Trans. 2 1988, 745-750. (b) Yasuda, M.; Shiomori, K.; Hamasuna,
S.; Shima, K.; Yamashita, T. J. Chem. Soc., Perkin Trans. 2 1992, 305-
310. (c) Yasuda, M.; Yamashita, T.; Matsumoto, T.; Shima, K.; Pac, C. J.
Org. Chem. 1985, 50, 3667. (d) Yasuda, M.; Yamashita, T.; Shima, K.;
Pac, C. J. Org. Chem. 1987, 52, 753. (e) Yamashita, T.; Shiomori, K.;
Yasuda, M.; Shima, K. Bull. Chem. Soc. Jpn. 1991, 64, 366-374.
(39) Yasuda, M.; Yamashita, T.; Shima, K. Bull. Chem. Soc. Jpn. 1990,
63, 938-946.
(22) For examples, see: (a) Amyes, T. L.; Jencks, W. P. J. Am. Chem.
Soc. 1989, 111, 7888-7900. (b) Richard, J. P. J. Am. Chem. Soc. 1989,
111, 1455-1465 and references therein (c) Young, P. R.; Jencks, W. P. J.
Am. Chem. Soc. 1977, 99, 8238-8248. (d) Fishbein, J. C.; McClelland, R.
A. J. Am. Chem. Soc. 1987, 109, 2824-2825. (e) McClelland, R. A.;
(40) Kazanis, S.; Azarani, A.; Johnston, L. J. J. Phys. Chem. 1991, 95,
4430-4435.