Concerted and stepwise dissociative electron transfers. Oxidability of the leaving group and strength of the breaking bond as mechanism and reactivity governing factors illustrated by the electrochemical reduction of α-substituted acetophenones

Article abstract of DOI:10.1021/ja963674b

The cyclic voltammetric investigation of a series of α-substituted acetophenones allowed the identification of the concerted and stepwise character of the dissociative electron transfer reaction, and, in the stepwise cases, the determination of the cleava

Full text of DOI:10.1021/ja963674b

2420
J. Am. Chem. Soc. 1997, 119, 2420-2429
Concerted and Stepwise Dissociative Electron Transfers.
Oxidability of the Leaving Group and Strength of the Breaking
Bond as Mechanism and Reactivity Governing Factors
Illustrated by the Electrochemical Reduction of R-Substituted
Acetophenones
Claude P. Andrieux,1a Jean-Michel Save´ant,*,1a Andre´ Tallec,1b
Robert Tardivel,1b and Caroline Tardy1a
Contribution from the Laboratoire d’Electrochimie Mole´culaire de l’UniVersite´ Denis Diderot,
Unite´ Associe´e au CNRS No 438, 2 place Jussieu, 75251 Paris Cedex 05, France, and the
Laboratoire d’Electrochimie de l’UniVersite´ de Rennes, Unite´ Associe´e au CNRS No 439,
Campus de Beaulieu, 35042, Rennes Cedex, France
ReceiVed October 23, 1996X
Abstract: The cyclic voltammetric investigation of a series of R-substituted acetophenones allowed the identification
of the concerted and stepwise character of the dissociative electron transfer reaction, and, in the stepwise cases, the
determination of the cleavage rate constants and the standard potentials for the formation of the anion radical. Analysis
of the data, using thermodynamical parameters derived from experiment and from literature points to three mechanism
governing factors, the oxidability of the leaving group, the bond dissociation energy of the bond being broken, and
the LUMO energy. The first of these factors appears to be largely predominant in many cases in the control of the
concerted vs stepwise dichotomy. The fluoro substituent provides a reverse example where the bond strength
overcomes the unfavorable effect of the leaving group oxidability. It is also an exception, in terms of anion radical
cleavage reactivity, where the strength of the C-F bond significantly contributes to slow down the cleavage as
opposed to the other substituents where solvent reorganization appears as largely predominant. In the concerted
cases, the estimated lifetime of the anion radical is clearly larger than the time of a vibration. The concerted character
of the reaction thus results from an energetic advantage rather than from the “nonexistence” of the anion radical
intermediate.
The dichotomy between concerted and stepwise mechanisms
in dissociative electron transfer
The competition between the two reaction pathways depends
both upon driving force and intrinsic barrier factors. The first
of these factors results from the comparison between the
standard potentials, E01 ) -D1 + EX0 /X + T∆S for the
concerted: RX + e- h R• + X-
(I)
•
-
concerted reaction, and E02 ) ER0 X/RX•- for the stepwise reaction
(D1 is the bond dissociation energy in the ground state, ∆S is
the corresponding entropy change, and E0X /X is the standard
stepwise:RX + e- h RX•- (II) RX•- h R• + X- (III)
•
-
is an important issue in the understanding and prediction of the
type of chemistry that can be triggered upon injecting one
electron into a molecule by reaction with homogenous electron
donors2 as well as by electrochemical,2 photochemical,3 or pulse
radiolytic means.4
potential for the oxidation of the leaving group). The difference
between these two standard potentials represents the standard
free energy for the cleavage of the anion radical:
∆G03 ) -E01 + E02 ) D1 - EX0 •/X- + ER0 X/RX•- - T∆S
X Abstract published in AdVance ACS Abstracts, February 15, 1997.
(1) (a) Universite´ Denis Diderot. (b) Universite´ de Rennes.
(1)
(2) (a) Save´ant, J.-M. Single Electron Transfer and Nucleophilic Substitu-
tion in AdVances in Physical Organic Chemistry; Bethel, D., Ed.; Academic
Press: New York, 1990; Vol. 26, pp 1-130. (b) Save´ant, J.-M. Acc. Chem.
Res. 1993, 26, 455. (c) Save´ant, J.-M. Dissociative Electron Transfer. In
AdVances in Electron Transfer Chemistry; Mariano, P. S., Ed.; JAI Press:
New York, 1994; Vol. 4. pp 53-116. (d) Save´ant, J.-M. Tetrahedron 1994,
50, 10117.
(3) (a) Saeva, F. D Topics Current Chem. 1990, 156, 61. (b) Saeva, F.
D. Intramolecular Photochemical Electron Transfer (PET)-Induced Bond
Cleavage Reactions in some Sulfonium Salts Derivatives in AdVances in
Electron Transfer Chemistry; Mariano, P. S., Ed.; JAI Press: New York,
1994; Vol. 4, pp 1-25. (c) Arnold, B. R.; Scaino, J. C.; McGimpsey, W.
G. J. Am. Chem. Soc. 1992, 114, 9978. (d) Chen, L.; Farahat, M. S.; Gan,
H.; Farid, S.; Whitten, D. G. J. Am. Chem. Soc. 1995, 117, 6399.
(4) (a) Neta, P.; Behar, D. J. Am. Chem. Soc. 1980, 102, 4798. (b) Behar,
D.; Neta, P. J. Phys. Chem. 1981, 85, 690. (c) Behar, D.; Neta, P. J. Am.
Chem. Soc. 1981, 103, 103. (d) Behar, D.; Neta, P. J. Am. Chem. Soc. 1981,
103, 2280. (e) Bays, J. P.; Blumer, S. T.; Baral-Tosh, S.; Behar, D.; Neta,
P. J. Am. Chem. Soc. 1983, 105, 320. (f) Norris, R. K.; Barker, S. D.; Neta,
P. J. Am. Chem. Soc. 1984, 106, 3140. (g) Meot-Ner, M.; Neta, P.; Norris,
R. K.; Wilson, K. J. Phys. Chem. 1986, 90, 168.
Thus, if the concerted pathway has a stronger driving force
than the first step of the stepwise pathway, the cleavage of the
anion radical is thermodynamically favorable. It is sometimes
considered that stepwise mechanisms are followed when the
intermediate (here the anion radical) has a lifetime that is longer
than the time for a bond vibration (≈10-13 s), while concerted
mechanisms would occur under the opposite conditions, i.e.,
when the intermediate “does not exist”.5 However, another point
of view has been developed according to which the concerted
mechanism is considered to take place when its activation energy
is less than that of the stepwise pathway, regardless of the
“existence” of the anion radical.2 Of course, the nonexistence
of the anion radical is a sufficient condition for the concerted
mechanism to take place but it is not a necessary condition. In
(5) Eldin, S.; Jencks, W. P. J. Am. Chem. Soc. 1995, 117, 9415.
S0002-7863(96)03674-8 CCC: $14.00 © 1997 American Chemical Society

R-Substituted Acetophenones
J. Am. Chem. Soc., Vol. 119, No. 10, 1997 2421
other words, concerted pathways may be followed under
conditions where RX•- has a finite lifetime, i.e., lives longer
than a vibration. As shown earlier, the kinetics of the concerted
and stepwise reactions may be modeled so as to obey the
following activation/ driving force relationships and intrinsic
barrier expressions.2 For each of the three reactions, the
activation free energy, ∆Gq, is a quadratic function of the
standard free energy of the reaction, ∆G0
2
∆G0
∆Gq ) ∆Gq0 1 +
(2)
q
(
)
4∆G0
where ∆Gq0 is the intrinsic barrier free energy (activation free
energy at zero driving force). The reaction standard free
energies are E - E0RX/R +X and E - ER0 X/RX for reactions I
•
-
•-
and II, respectively (E is the electrode potential, the standard
potentials being those defined earlier) and ∆GR0 X
for
•-
•
/R +X-
reaction III. For reactions I and III, the intrinsic barrier free
energies are related to the bond dissociation energy, D, and to
the solvent reorganization factor, λ0, according to
D1 or 3 + λ01
or3
∆Gq1,0
)
(3)
Figure 1. Variation of the cyclic voltammetric peak potential with
the scan rate for a concerted (full line) and stepwise (dotted line)
reaction. k3 ) 6 × 1010 s-1, kT/h ) 6 × 1012 M-1 s-1, λ0 ) 0.7, λ0
or3
4
1
2
) 0.5, λ03 ) 0.2 eV. Diffusion coefficient: 10-5 cm2 s-1, electrochemi-
cal frequency factor: 4 × 103 cm s-1. The numbers on each section of
the curves denotes the rare determining step.
while for reaction II
∆Gq2,0
(D11/2 - D31/2)2 + λ02
)
(4)
on experimental grounds in the reduction of arylmethyl halides,9
4
aromatic N-halosultams10 and sulfonium cations.6b The role of
In view of the large number of parameters involved, it is not
possible to describe the passage from the concerted to the
stepwise mechanism by a single equation. We may however
find typical examples where the anion radical has a finite
lifetime and where, nevertheless, the electrochemical reaction
(as investigated for example by mean of cyclic voltammetry)
follows a concerted mechanism, at least in the lower part of
the range of scan rates.6 Some such examples where k3 ) 6 ×
1010 and 1011 s-1, are represented in Figure 1 as peak potentials
vs scan rate plots. The procedure for constructing these
diagrams are detailed in the Supporting Information based on
the analyses in refs 7 and 8. Upon increasing the scan rate, the
peak potential shifts toward negative values and the mechanism
passes from concerted (noted 1 in Figure 1) to stepwise with,
first, reaction III and then reaction II as the rate determining
step.
DRXfR +X• was also demonstrated in the latter case. So far,
•
the role of the oxidation potential of the leaving group,
E0X /X has received much less attention. In the case of nitro-
•
-
substituted N-halosultams10
As a general trend, the concerted reaction prevails over the
stepwise reaction less easily than predicted on mere thermo-
dynamic grounds. An increase of the driving force of the former
at the expense of the latter will nevertheless tend to make the
system pass from the stepwise to the concerted mechanism
It follows that weak bonds, poor oxidability of the leaving
group, and low energies of the orbital where the incoming
electron may be accommodated (usually a π* orbital) will favor
the concerted pathway over the stepwise pathway and Vice Versa.
the stepwise character of the reaction with F, as opposed to Br
and Cl, was attributed to a larger bond dissociation energy. At
the same time however, the oxidability of the leaving halide
ion changes in a way that would conversely favor the concerted
pathway (E0X /X ) 2.62, 1.79, and 1.44 V vs SCE for F, Cl,
•
-
and Br respectively11). It may thus be concluded that, in this
series, the effect of bond strength overcompensates the effect
of leaving group oxidability.
The discriminating role of E0RX/RX has been clearly identified
•-
(6) (a) Passage from a concerted to a stepwise mechanism has been
observed in the electrochemical reduction of sulfonium cations.6b Andrieux,
C. P.; Robert, M.; Saeva, F. D.; Save´ant, J.-M. J. Am. Chem. Soc. 1994,
116, 7864.
In order to investigate the combined effect of these two factors
in the concerted/stepwise dichotomy and to examine whether
(7) (a) Nadjo, L.; Save´ant, J.-M. J. Electroanal. Chem. 1973, 48, 113.
(b) Andrieux, C. P.; Save´ant, J.-M. In Electrochemical Reactions in
InVestigation of Rates and Mechanisms of Reactions, Techniques of
Chemistry; Bernasconi, C. F., Ed.; Wiley: New York, 1986; Vol. VI/4E,
Part 2, pp 305-390.
(9) Andrieux, C. P.; Le Gorande, A.; Save´ant, J.-M. J. Am. Chem. Soc.
1992, 114, 6892.
(10) Andrieux, C. P.; Differding, E.; Robert, M.; Save´ant, J.-M. J. Am.
Chem. Soc. 1993, 115, 6592.
(8) Save´ant, J.-M. J. Phys. Chem. 1994, 98, 3716.
(11) See refs 9 and 10.

2422 J. Am. Chem. Soc., Vol. 119, No. 10, 1997
Andrieux et al.
Chart 1
Figure 2. Cyclic voltammetry of 4b (1.9 mM) in DMF + 0.1 M n-Bu4-
NBF4 at a scan rate of 0.1 V/s and variation of the difference between
anodic and cathodic peak potentials with the scan rate: temperature,
20 °C; 1 mm-diameter GC electrode.
of electron transfer is the strong localization of the negative
charge on the carbonyl oxygen in the anion radical. This
interpretation is confirmed by the fact that charge transfer to
acetophenone is also relatively slow. A value of 0.14 cm s-1
was indeed found from the variations of the cathodic and anodic
peak potentials with the scan rate.
All other compounds in the 1, 2, 3, 4, 5, and 7 series exhibit
two successive waves at low scan rates (below 1 V/s). The
first is irreversible, and its height corresponds to the exchange
of one electron per molecule. The electron stoichiometry of
the first wave was checked by comparison with the one-electron
reversible wave of either benzophenone or 4-cyanoacetophe-
none, which have similar diffusion coefficients. In this deter-
mination, the influence of the irreversibility of the first wave
of the substituted acetophenone on the peak height was taken
into account through the value of the apparent transfer coef-
ficient, Rap,7 (ip ) 0.496FSC0D1/2(RapFV/RT1/2) determined from
the peak-width (see below). The second is quasi-reversible and
corresponds to the reduction of acetophenone (or of the 4-nitro-
or 4-cyano-, 4-methoxy- or 4-methyl acetophenones with 1b,
1c, 1d, 4c, and 5b, respectively) as verified with authentic
samples.16 The electron stoichiometry of the second wave is
half that of the first wave, thus corresponding formally to half-
an-electron. The situation is thus quite different from what was
found with the 6 and 8 series where there is a single irreversible
one-electron wave along which the anion radical dimerizes rather
than cleaves.12 These observations suggest the reduction
mechanism depicted in Scheme 1 where the one-electron and
half-an-electron stoichiometries at the first and second waves
derive from the reaction of the acetophenone enolate formed at
the first wave along a two-electron process with one molecule
of starting material.
the effect of E0X /X may be predominant, we investigated the
•
-
electrochemistry in N,N′-dimethylformamide (DMF) at glassy
carbon electrodes of the following series of acetophenones
bearing a leaving anionic group in the R-position, and, for some
of them, a substituent on the phenyl ring (see Chart 1).
For comparison purposes, earlier results concerning the
reduction of R-alkoxy-acetophenones (6a,b) and R-amino-
acetophenones (8a-f) showing that their anion radicals do not
cleave but rather dimerize yielding the corresponding pinacol12
will also be used in the following discussion.13
Results and Discussion
Main Characteristics of the Cyclic Voltammetric Re-
sponses. 4b is the only compound in the whole series that
exhibits a single one-electron reversible wave (Figure 2) down
to the lowest scan rate (0.1 V/s). From the variation of the
peak potential with the scan rate and the value of the standard
potential (E02 ) -0.79 V vs SCE),14 or from the separation
between cathodic and anodic peak potentials (Figure 1),12c an
apparent standard rate constant of electron transfer (2) of kS2 )
0.22 cm s-1 is found, a value substantially smaller than for usual
aromatic compounds.15 The reason for this relative slowness
(12) Andrieux, C. P.; Save´ant, J.-M. Bull. Soc. Chim. Fr. 1972, 3281.
(13) (a) Cleavage of several substituted acetophenone ketyls generated
by homogeneous electron donors has also been investigated (see ref 13b
and references cited therein) as well as for electrogenerated arylcyclopropyl
acetophenone ketyls (see reference 13c and references cited therein). More
recently, the cleavage rate constants of anion radicals of acetophenones
bearing various phenoxy R-substituents have been reported.13d (b) Tanner,
D. D.; Chen, J. J.; Chen, L.; Luelo, C. J. Am. Chem. Soc. 1991, 113, 8074.
(c) Tanko; J. M.; Drumright, R. E.; Sulemen, N.; Brammer, L. E. J. Am.
Chem. Soc. 1994, 116, 1585. (d) Andersen, M. L.; Mathivanan, N.; Wayner,
D. D. M. J. Am. Chem. Soc. 1996, 118, 4871.
In line with this reaction mechanism, the first irreversible
peak doubles in height when a strong enough acid is added to
the solution. Figure 3 gives an example of this behavior in the
case of 4a with addition of phenol. The second wave represents
the reduction of the acetophenone generated at the first wave.
In the absence of acid, its height corresponds to 0.5 electron in
agreement with the above reaction scheme. Upon addition of
the acid it grows up to two electrons which falls also in line
with Scheme 1, taking into account the fact that acetophenone
(14) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; Wiley: New
York, 1980.
(15) Kojima, H.; Bard, A. J. J. Am. Chem. Soc. 1975, 97, 6317.

R-Substituted Acetophenones
J. Am. Chem. Soc., Vol. 119, No. 10, 1997 2423
Scheme 1
undergoes a 2e- + 2H+ reduction in acidic medium.16b Table
1 summarizes several examples where the same behavior was
found.
Assignment of the Cleavage Mechanism. In the investiga-
tion of the concerted vs stepwise dichotomy, the simplest
characterization is that of the stepwise mechanism. Evidence
for the intermediacy of the anion radical can be obtained in
two ways. When its lifetime is not too short, direct evidence
can be gained if the first wave becomes reversible upon raising
the scan rate, while, simultaneously, the acetophenone wave
should disappear. This is what is observed with 7b at 20 °C,
where reversibility is reached at a rather low scan rate, 10 V/s
(Figure 4). The fact that the acetophenone wave appears at a
low scan rate beyond the reduction wave of 7b is a clear
indication that its anion radical does cleave rather than dimerizes
as occurs for compounds 6 and 8.12 At 20 °C, none of the other
compounds exhibits any reversibility up to scan rates as high
as 30 000 V/s using a gold ultramicroelectrode.17 However,
with 5a, upon lowering the temperature down to -16 °C,
The reason that we have included in Scheme 1 the formation
of the enol of acetophenone (reactions 6 and 6′) concurrently
with the formation of acetophenone itself (reactions 5 and 5′)
from the common enolate/carbanion intermediate derives from
high scan rate cyclic voltammetric experiments where the
kinetics of these reactions manifest themselves upon raising the
scan rate. At lower scan rates the thermodynamically favorable
formation of acetophenone predominates and therefore the enol
does not appear in the stoichiometric balances of Scheme 1.
This point is not central to our discussion of the concerted vs
stepwise dichotomy. Its detailed analysis will be published
elsewhere.
(16) (a) The acetophenone wave is not entirely reversible because of
dimerization of the anion radicals leading to the pinacolate.16b (b) Nadjo,
L.; Save´ant, J.-M. J. Electroanal. Chem. 1971, 3, 419.
(17) Andrieux, C. P.; Hapiot, P.; Save´ant, J.-M. Chem. ReV. 1990, 90,
723.

2424 J. Am. Chem. Soc., Vol. 119, No. 10, 1997
Andrieux et al.
peak potential, Ep, with the scan rate and of the values of the
half peak width, Ep/2 - Ep. In order to keep ohmic drop
negligible, two GC disk electrodes were used successively: a
3 mm-diameter electrode between 0.1 and 10 V/s (open squares
in Figures 5 and 6) and a 1 mm-diameter electrode between 10
and 100 V/s (open circles in Figures 5 and 6). In order to free
the data from the kinetic influence of the follow-up protonation
steps they should be gathered either with no acid added, in which
case the electron stoichiometry is 1, or with addition of acid in
sufficient concentration for the electron stoichiometry to reach
2. Some typical results are shown in Figure 5.
If, in the stepwise mechanism (reactions II and III), the rate
determining step is reaction III while reaction II remains at
equilibrium the peak potential varies linearly with logv with a
slope ∂Ep/∂ log V ) 29 mV at 20 °C, while the peak width Ep/2
- Ep ) 47 mV at the same temperature.7 Conversely, if reaction
II is the rate determining step, ∂Ep/∂ log V ) (29/R2) while Ep/2
- Ep ) (47/R2) mV. Increasing the scan rate makes the kinetics
pass from the former situation to the latter.7 We may thus derive
from the peak width data an apparent transfer coefficient Rap
) 47/(Ep/2 - Ep) that is expected to vary from 1 to R2 (which
should be close to 0.5) when the kinetic control passes from
reaction III to reaction II. We have already noted that electron
transfer to acetophenone is not very rapid. It follows that mixed
kinetic control by reactions II and III is likely to be encountered
in the present series of compounds. It is thus possible to use
as diagnostic criterion of the stepwise mechanism, the observa-
tion that Rap is larger than 0.5 in a significant portion of the
range of scan rates explored. For more precision in the
mechanism diagnosis, it is desirable to jointly use the Ep -log
v data and Ep/2 - Ep data. This was done by testing the
consistency of the Ep -log v and Ep/2 - Ep data according to
a procedure described in details in the Supporting Information.
That this test of consistency leads to satisfactory results can be
seen in Figure 5 by the fact that the solid line in the upper
diagrams representing the theoretical reconstruction of the Ep
-log v variation from the Ep/2 - Ep data passes through the Ep
-log v points within experimental uncertainty. For 5a, the
occurrence of the stepwise mechanism at -16 °C was demon-
strated by the reversibility observed upon raising the scan rate
(Figure 4). The same conclusion holds at 20 °C as results from
the Ep -log v and Ep/2 - Ep data. This is also true for 1c and
3, where reversibility could not be reached even at low
temperature. Figures similar to Figure 5 for all the other
compounds of the same category can be found in the Supporting
Information. The mechanistic conclusions are summarized in
Table 2. The above test of consistency of the Ep -log V and
Ep/2 - Ep data also provides the value of the two parameters
Figure 3. Cyclic voltammetry of 4a (1 mM) in DMF + 0.1 M n-Bu4-
NBF4 at a scan rate of 0.2 V/s in the absence of acid (a) and in the
presence of 5 mM phenol: temperature, 20 °C; 3mm-diameter GC
electrode.
Table 1. Addition of Acids
compd (concn in mM)
acid (concn in mM)
1a (2)
1d (2)
2 (2)
3 (1.1)
4a (1)
7a (2)
7b (1.3)
CF2HCO2H (2)
CF2HCO2H (2)
CF2HCO2H (2)
CH3CO2H (4)
PhOH (5)
PhOH (10)
PhOH (3)
C1 ) log(Fk3D2/2RTkS4,2
)
(5)
and
C2 ) E02 + (RT ln10/F) log(2k3D/k2S,2
)
(6)
Figure 4. Cyclic voltammetry (in DMF) of 7b (1.3 mM), 0.1 M n-Bu4-
NBF4, at a GC disk electrode, temperature, 20 °C, and of 5a (2.9 mM),
0.4 M Et4NBF4, at a gold disk electrode, diameter, 25 µm (a), 10 µm
(b, c), temperature, -16 °C.
(see Supporting Information), from the values of which k3 and
E0 may be derived provided the value of kS,2 can be estimated.
kS,2 was taken as equal to 0.14 cm/s, the value for acetophenone,
for 3, 4a, 4c, 5a, and 5b and to 0.22 cm s-1, the value for the
nitro-derivative 4b, for 1b and 1c. Because the cyclic volta-
mmogram of 7b ceases to be irreversible upon raising the scan
rate, the foregoing treatment is not applicable to this particular
reversibility is reached at 8000 V/s (Figure 4). Similar
observations were made with 5b, where full reversibility at -20
°C is reached at 22 855 V/s (see Supporting Information). In
these three cases, the reduction thus follows a stepwise
mechanism.
compound. In this case, the values of k3 and E0 were derived
2
In cases where the reduction wave is irreversible over the
whole range of accessible scan rates, another approach to the
mechanism consists in the observation of the variation of the
by direct simulation of the voltammograms. The simulation
also provides the value of kS,2, 0.16 cm s-1, which lies, as
expected, in between the values found for acetophenone, and

R-Substituted Acetophenones
J. Am. Chem. Soc., Vol. 119, No. 10, 1997 2425
Figure 5. Cyclic voltammetric peak characteristics of 1c (1 mM), 3 (1 mM) and 5a (1 mM,) in DMF + 0.1 M n-Bu4NBF4, at a 3 mm (0) and 1
mm (O) diameter glassy carbon electrode: temperature, 20 °C. The lower diagrams represent the values of Rap derived from the peak width. The
solid lines in the top diagrams represent the variations of Ep derived from the peak width values according to the procedure described in the text
and in the Supporting Information.
Table 2. Reductive Cleavage Mechanism
cient:
compd
mechanism
1.857 RT
F(Ep/2 - Ep)
R )
1a
1b
1c
1d
2
concerted
stepwise
stepwise
concerted
concerted
stepwise
stepwise
stepwise
stepwise
stepwise
stepwise
stepwise
stepwise
ks
RT RT
+
Ep ) E0 - 0.78
ln
(
)
RF RF
x
RFVD/RT
3
4a
4b
4c
5a
5b
7a
7b
Inspection of Figure 6 shows that the consistency between the
two sets of data is again satisfactory. The fact that R is
significantly smaller than 0.5 in all three cases is a first
indication that the reduction of these compounds follows a
concerted mechanism. Indeed, according to the theory of
dissociative electron transfers recalled in the introduction, the
intrinsic barriers are usually large because they are mostly
governed by the dissociation energy of the bond being broken.
It follows that the reduction takes place at a potential much
more negative than the dissociative electron transfer standard
Table 3. Standard Potentials and Cleavages Rates for the Stepwise
Reactions
compd C1 (log(s/V)) C2 (V vs SCE) log k3 (s-1
)
-E0 (V vs SCE)
potential, E0RX/R +X-, and therefore that the transfer coefficient
1b
1c
3
4a
4c
5a
5b
7a
7b
0.50
1.83
0.79
1.15
2.33
0.10
0.10
1.00
-1.35
-0.79
-1.57
-1.61
-1.70
-1.67
-1.74
-1.58
6.6
7.8
6.1
6.4
7.6
5.4
5.4
6.1
1.2a
0.83
1.04
1.75
1.81
1.97
1.81
1.88
1.77
1.87a
•
(symmetry factor)
E - E0RX/R•+X-
R ) 0.5 1 +
(7)
4∆Gq0
(
)
should be significantly smaller than 0.5. This conclusion may
be confirmed as follows. Because of inductive effects, the
standard potential for the formation of the anion radical of 1a,
1d, and 2 is expected to be more negative than that of 3. Thus
if the reduction of 1a, 1d, and 2 were to follow a stepwise
mechanism, kinetically controlled by reaction II, their peak
potentials given by
a From the CV reversibility data in Figure 4.
4b. The results obtained for the whole set of compounds
undergoing a stepwise reaction are listed in Table 3.
Compounds 1a, 1d, and 2 exhibit a different behavior (Figure
6) indicating that the reduction is controlled by an electron
transfer reaction, either I or II, with a value of Rap clearly below
ks,2
RT RT
+
Ep,2 ) E02 - 0.78
ln
0.5. In these cases, the test of consistency of the Ep and Ep/2
-
(
)
RF RF
Ep data (see Supporting Information) simply consists in the
application of the two equations relating each of these to the
transfer coefficient, R, which is now a true transfer coeffi-
x
RFVD/RT
should be more negative than

2426 J. Am. Chem. Soc., Vol. 119, No. 10, 1997
Andrieux et al.
Figure 6. Cyclic voltammetric peak characteristics of 1a (2 mM,), 1d (2 mM), and 2 (2 mM) in DMF + 0.1 M n-Bu4NBF4 + 2 mM CF2HCO2H,
at a 3 mm (0) and 1 mm (O) diameter glassy carbon electrode: temperature, 20 °C. The lower diagrams represent the values of R derived from
the peak width. The solid lines in the top diagrams represent the variations of Ep derived from the peak width values according to the procedure
described in the text. The solid lines in the lower diagrams represent the predicted values of Rap according to eq 7.
ks,2
RT RT
+
E02(3) - 0.78
ln
(
)
RF RF
x
RFVD/RT
The solid line in Figure 7 represents this maximized variation
using the data in Table 3, taking D ) 10-5 cm2 s-1, R ) 0.5,
and kS,2 ) 0.14 cm s-1 (the value for acetophenone). The actual
Ep values are thus in average at least 430, 400, and 250 mV
more positive than predicted for a stepwise mechanism.
Equivalently, the stepwise mechanism would imply that the
heterogeneous standard rate constant of reaction II would be
larger than 1028, 382, and 20 cm/s for 1a, 1d, and 2,
respectively. Obviously, these values are absurdly large; the
heterogeneous standard rate constant for the reduction of
anthracene in DMF, one of the fastest electron transfer reaction,
is only 3 cm/s.18 We can thus safely conclude that the reduction
of these three compounds follows a concerted mechanism as
reported in Table 2.
Figure 7. Predicted maximized variation of the peak potential variation
with scan rate for 1a, 1d, and 2 if they were to follow a stepwise
mechanism (solid line) as compared to the experimental variations (open
squares and circles).
Relationships between Molecular Structure, Mechanism,
and Reactivity. In order to establish these relationships we
need to extract a series of parameters from the experimental
data and gather others from the literature.
eV, respectively9,21 pointing to a greater stability of the phenacyl
radical as compared to the benzyl radical as expected from the
following resonant forms
For the compounds following a concerted mechanism, the
bond dissociation energy (BDE) of the starting molecule, RX,
may be derived from the value of the peak potential and the
value of the standard potential for the oxidation of the leaving
group, E0X /X (Table 4). For example, at 0.1 V/s9,10
- 19b
(19) (a) A reliable value, 1.06 V vs SCE, is available for CH3CO2
,
but not for PhCO2-. The value given in ref 19b, 0.36 V vs SCE, is smaller
than for CH3CO2-, which seems counterintuitive in view of the fact that
-
-
CH3CO2 is a stronger base than PhCO2 (the pKa’s in acetonitrile are
22.3 and 20.7, respectively19c,d). Starting from the E0 value of CH3CO2
,
-
•
-
-
we estimated the value for PhCO2 from the following equation
EP0hCO2- - EC0 H3CO2- (V) DPhCO H - DCH CO
+
2
3
-
-
H
D1 ) (E0X•/X- - Ep) + 0.3
(8)
2
3
2
0.06 (pKa,PhCO H - pKa,CH CO H
)
2
3
2
assuming that the entropic terms, ∆SRX/R•+Η•, are ca. the same in both cases.
(b) Eberson, L. Acta Chem. Scand B 1984, 439. (c) Kolthoff, I. M.; Chantoni,
M. K.; Bhownik, S. J. Am. Chem. Soc. 1968, 90, 23. (d) Coetzee, J. F.
Prog. Phys. Org. Chem. 1967, 4, 76.
The BDE of 1a and 2 thus obtained are significantly smaller
than those of the corresponding benzyl halides, 2.52 and 3.12
(20) (a) Hapiot, P.; Pinson, J.; Yousfi, N. New J. Chem. 1992, 16, 877.
(b) Andrieux, C. P.; Hapiot, P.; Pinson, J.; Save´ant, J.-M. J. Am. Chem.
Soc. 1993, 115, 7783.
(18) Andrieux, C. P.; Garreau, D.; Hapiot, P.; Save´ant, J.-M. J.
Electroanal. Chem. 1988, 248, 448.

R-Substituted Acetophenones
Table 4. Reactivity Dataa
J. Am. Chem. Soc., Vol. 119, No. 10, 1997 2427
compd
D1
E0X /X
-E0Rx/Rx
log k3 (s-1
)
-∆G03
-Epb concerted
-Epb stepwise
•
-
•-
concerted
1a
2
1d
stepwise
1b
1c
3
4a
4b
4c
5a
5b
6a
6b
7a
c
e
predicted
1.76-2.01
1.76-2.01
1.76-2.15
expal
0.83
1.04
1.75
1.81
0.79
1.97
predicted
expal
1.19
1.36
1.20
predicted
0.89
predicted
1.46-1.71
1.46-1.71
1.46-1.85
expal
0.64
0.88
1.61
1.67
0.82
1.82
1.69
1.76
1.8
1.8
1.64
1.84
1.83
2.05
2.40
2.07
d
1.44
1.79
1.44
10.1-11.0
10.1-11.0
10.0-11.4
expal
6.6
7.8
1.32-1.57
1.32-1.57
1.30-1.69
1.85
1.90
3.12-3.46
2.53
2.53
2.53
1.78
1.78
2.56
2.57
1.77
2.14
2.67
1.44e
1.44e
2.62e
0.59
0.75
1.20-1.54
0.70
-0.15
0.86
0.44
0.96
6.1
1.24f,g
1.24f,g
1.24f,g
0.24f,h
≈0.24
0.06f,i
≈0.06
0.10f,j
0.06f,i
≈-1.0k
6.4
2.10
2.10
2.10
1.98
2.06
3.33
3.35
2.11
2.70
4.76
(-0.9)m
7.6
1.81
5.4
5.4
1.88
0.51
1.80l
(-2.8)m
(-2.9)m
6.1
-0.53
-0.54
0.27
-0.04
-1.87
1.80l
1.77
7b
8d
1.87
1.2
1.83l
(-18.8)m
a Energies in eV, potentials in V vs SCE. b At 0.1 V/s. c Estimated as described in the text. d From the experimental Ep data. e See footnote 11.
f Assumed to be the same as in acetonitrile. g See footnote 11. h From ref 20a. i From ref 19b. j From ref 20b. k For i-Pr2N- in THF, from ref 19b.
l From ref 12. m Estimated (see text).
It also appears that the difference in BDE is larger for the
chloro (0.72 eV) than for the bromo (0.47 eV) derivative. This
observation may be interpreted in terms of repulsive dipole-
dipole interactions in the starting molecule (δ+CdOδ- against
δ′+C-Xδ′-) which are expected to be larger with Cl than with
Br. Assuming that these interactions are weak for Br, we may
estimate the BDE s of the other R-substituted acetophenones,
4a, 5a, 6a, 6b, 7a, and 7b, from the following equation
Figure 8. Compounds reduced according to a stepwise mechanism
(O, 0, and ]). Experimental variation of the cleavage rate constant
with the standard free energy of the reaction. Estimation of the anion
radical cleavage rate constants for the compounds reduced according
to a concerted mechanism.
D
) D
- D
+ D
PhCOCH2X
PhCH2X
PhCH2Br PhCOCH2Br
sponds to the formation of two particles out of one. It is about
constant in the series and equal in average to 0.59 meV K-1
.
using for the R-substituted toluenes the values gathered from
ref 21. For 1b and 1c, the value for 1a has been corrected for
the weakening of the bond by electron-withdrawing para
substituents as for the corresponding benzyl bromides.9,22 As
seen with 1d and with the corresponding substituted benzyl
bromides, the effect of electron-donating para substituents, such
as CH3 and CH3O, is weak. The same value of D1 as for 4a
and 5a was therefore taken for 4c and 5b, respectively. The
BDE of 3 cannot be derived from that of 1a according to the
above procedure since the dipole-dipole interactions in 3 are
expected to be strong, even stronger than in 2. The only piece
As seen in Figure 8, there is a good correlation, obeying eq 2,
between the cleavage rate constant and the driving force of the
reaction with the exception of 3 and 7a which we will discuss
later on. The very existence of the correlation and the fact that
it is endowed with a large intrinsic barrier has been discussed
elsewhere23 in terms of solvent reorganization attending the
displacement of the negative charge from the carbonyl oxygen
to the leaving group as the cleavage proceeds.
We may use this correlation to estimate what would have
been the values of k3 for compounds 1a, 1d, and 2 (Table 4) if
their reduction had followed a stepwise rather than a concerted
mechanism. This estimation requires the value of ∆G03 for
each of these three compounds and thus of E02 ) ER0 X/RX•-. For
1a and 2, the latter parameter can be bracketed between the
values for 3 and for acetophenone, and, for 1d, between the
values for 3 and for 4-methoxyacetophenone (-2.15 V vs SCE).
The values of ∆G03 listed in Table 4 ensue. k3 is then
of knowledge we have for the moment is that DPhCOCH F
<
2
DPhCH F - DPhCH Cl + DPhCOCH Cl, i.e., the BDE of 3 is smaller
2
2
2
than 3.46 eV. For the compounds that follow a stepwise
mechanism, the free energy of reaction III, ∆G03, can be
derived (Table 4) from the values of D1, E0X /X-, E0
)
•
2
E0RX/Rx and ∆S according to eq 1. The latter term corre-
•-
(21) (a) Benson, S. N. Thermochemical Kinetics; Wiley: New York,
1976. (b) Handbook of Chemistry and Physics, 72nd ed.; CRC: Cleveland,
OH, 1991-1992, pp 9-121.
(23) Andrieux, C. P.; Save´ant, J.-M.; Tallec, A.; Tardivel, R.; Tardy, C.
J. Am. Chem. Soc. 1996, 118, 9788.
(22) Clark, K. B.; Wayner, D. D. M. J. Am. Chem. Soc. 1991, 113, 9363.

2428 J. Am. Chem. Soc., Vol. 119, No. 10, 1997
Andrieux et al.
Table 5. Factors Governing the Concerted/Stepwise Dichotomy in
bracketed by means of the extrapolation procedure depicted in
Figure 8. In all three cases the predicted values of k3, albeit
large, are clearly below 1013 s-1, pointing to the conclusion that
the reduction of these three compounds follows a concerted
mechanism in spite of the fact that their anion radicals “exist”
thus illustrating the discussion of this point given in the
Introduction.
the Acetophenone Seriesa
comparison with 2
comparison with 1a
0
0
b
-c
d
-e
compd
-1.5∆D1
∆EX /X
-1.5∆D1
∆EX /X
•
•
3
-1.08/-1.59
-0.20
0.93
-0.24
-0.26
0.94
0.83
-0.55
-1.55
-1.73
-1.73
-1.69
-1.73
-2.99
-1.61/-2.11
-0.72
0.40
-0.77
-0.78
0.41
1.18
-0.20
-1.20
-1.38
-1.38
-1.34
-1.38
-2.64
4a
5a
6a
6b
7a
7b
8d
We may also, from these predicted values of k3, bracket the
peak potential (at 0.1 V/s) that would have been observed if
the reduction of compounds 1a, 1d, and 2 had followed a
stepwise mechanism, assuming, as seems reasonable, that kS,2
is close to that of acetophenone. To achieve this estimation,
we used a procedure that is the reverse of the procedure that
0.39
-0.41
-0.14
-0.93
a Energies in eV, potentials in V vs SCE. b ∆D ) DPhCOCH X
-
2
c ∆E0X ) E0X /X- - E0
. .
d ∆D ) DPhCOCH X - DPhCOCH Br
-
2
Cl /Cl
was used earlier to estimate k3 and E0 from the peak potential
2
2
DPhCOCH Cl
.
•
•
2
e ∆E0X ) E0X /X - EB0 r /Br
.
-
•
-
•
data for the real stepwise reactions. Namely, the value of log
V + C1 is derived from the values of k3 and kS,2 according to eq
5. Then, the value of the Ep - C2 is obtained from the
theoretical relationship between this parameter and log V + C1
(Figure S3a in the Supporting Information section). Deriving
potential becomes more negative, with a 1.5 coefficient, thus
favoring the stepwise mechanism over the concerted mechanism.
A decrease of E0X /X has the same effect with a unity coef-
•
-
ficient. Table 5 summarizes the contributions of these two
factors which make the various compounds of the acetophenone
series follow a stepwise mechanism by comparison with the
two compounds in the same series, 2 and 1a, which undergo a
concerted mechanism. Inspection of Table 5 reveals that, with
the exception of 3, the stepwise mechanism (or a noncleaving
mechanism) is followed predominantly because of the poor
oxidability of the leaving group rather than because of a
strengthening of the cleaving bond. There are even cases, such
as 5a, 7a (compared with both 2 and 1a), and 7b (compared
with 2), where the BDE factors plays against the stepwise
mechanism but is overcompensated by a decrease in the leaving
group oxidability.
3 is a counter example to these rules; in spite of the large
uncertainty on its determination, it is seen that the BDE is the
dominant factor is the increase of this factor, while the leaving
group oxidability factor plays against the stepwise mechanism,
as was already observed with N-halonitrosultams (see Introduc-
tion).
The fact that 1b and 1c follow a stepwise mechanism, whereas
1a and 1d undergo a concerted mechanism, is a reflection of
the decrease of the standard potential for the formation of the
anion radical (which is a measure of the LUMO energy) which
overcomes a small decrease of the BDE. The situation is very
similar to what has been observed previously with benzyl
halides9 and N-halosultams.10
As seen in Figure 8, the points representing 3 and 7a fall
clearly out of the correlation between the kinetics and thermo-
dynamics of the anion radical cleavage. The fact that the data
point for 3 falls (even with a large uncertainty in the value of
the driving force) much below the correlation line may be
explained within the interpretation already given for the other
anion radicals.23 It was shown that the intrinsic barrier
C2 from the values of k3, kS,2, and E0 finally provides the value
2
of Ep. As expected with such fast follow-up reactions, the peak
potential is entirely governed by the forward electron transfer
step, i.e., by the values of kS,2 and E0. For all three compounds
2
(Table 4), the peak potential values thus predicted are much
more negative than the experimental values thus confirming that
their reduction does follow a stepwise mechanism. We also
see, in the lower diagrams of Figure 7, that the values of R
derived from eq 7 in the framework of a concerted mechanism
(using eq 8 and the numbers in Table 4) are in satisfactory
agreement with the experimental values.
Conversely, for the compounds that follow a stepwise
mechanism, the value of the peak potential (at 0.1 V/s) that
would have been observed if their reduction had followed a
concerted mechanism may be estimated, with the exception of
3, by application of eq 8 (Table 4). In all cases, the “concerted
peak potentials” are much more negative than the experimental
peak potentials thus providing a further confirmation of the
reduction mechanism. For the compounds that give rise to an
anion radical which is stable in the whole range of scan rates,
such as 4b•-, or that dimerizes rather than cleaves, such as 6a•-,
6b•-, 8d•-, the driving force for cleavage may be estimated by
application of eq 1, and, from it, the predicted values of k3,
using the logk3 - ∆G0 correlation established earlier (Figure
3
8). These are very approximate estimations since the values
of E0X /X are themselves very approximate (Table 4). Never-
•
-
theless, the very low values of k3 thus predicted allows one to
understand why these compounds do not undergo any significant
cleavage. The fact that 4b•- does not dimerize, whereas 6a•-,
6b•-, and 8d•- do, may be explained by the presence of the
nitro group in para which leaves little unpaired electron density
on the carbonyl carbon.
The case of 3 is worth some additional comments. An upper
limit of the BDE has already been obtained. A lower limit may
be derived from the very fact that its reduction follows a
stepwise rather than a concerted mechanism. The value of D1
that would give rise, for a concerted mechanism, to the same
peak potential value as observed experimentally is 3.12 eV. The
BDE of 3 should thus lie between 3.12 and 3.46 eV.
D
RX•- + λ0
∆Gq0 )
4
consists in a intramolecular cleavage contribution, DRX•-/4, and
a solvent reorganization contribution, λ0/4. From the data in
Figure 8, the total intrinsic barrier may be bracketed between
0.89 and 1.02 eV. The intramolecular cleavage contribution8,23
We may now examine what are the molecular factors that
govern the occurrence of the concerted vs the stepwise mech-
anism. Among the compounds deriving from acetophenone with
no ring substituents, there is no large variation of the standard
D
RX•-/4 ≈ (DRX + E0RX/RX•- - ER0 •/R-)/4
potential E0RX/RX
. The main factors that govern the con-
lies between 0.34 and 0.43 eV, and thus the solvent reorganiza-
tion contribution is in between 0.55 and 0.59 eV, i.e., 61-58%
of the total. Both contributions are larger than for the other
•-
certed/stepwise dichotomy are therefore D1 and E0 -. Ac-
•
X /X
cording to eq 8, an increase of D1 makes the predicted peak

R-Substituted Acetophenones
J. Am. Chem. Soc., Vol. 119, No. 10, 1997 2429
anion radicals of the acetophenone series.23 Concerning the
intramolecular cleavage contribution, this is essentially a
consequence of the fact than the C-F bond in the parent
molecule is the strongest of the whole series. The increased
solvent reorganization energy is the result of the small size of
the fluoride ion on which the charge, initially located on the
carbonyl oxygen, is transferred upon cleavage.
It is also worth noting, that in the three cases where a
concerted mechanism was identified, the lifetime of the anion
radical is clearly larger than the time of a vibration. The
concerted character of the reaction thus results from an energetic
advantage rather than from the “nonexistence” of the anion
radical intermediate.
In the case of 7a•-, the intramolecular reorganization is
negligible and the solvent reorganization contribution is 0.57
eV, significantly smaller than with the other compounds (0.7
eV). This observation may be explained by the fact that the
charge is less localized than with the other acetophenones when
it is transferred from the carbonyl oxygen to the leaving group
where delocalization between the sulfur atom and the phenyl
ring may take place. It is presumably for the same reason that
the data point for 5a is somewhat above the correlation line
due to some delocalization over the phenoxy moiety, albeit to
a lesser extend than in the PhS case. With 5b, such an effect
is expected to be counteracted by some more charge localization
on the carbonyl oxygen in the initial state under the influence
of the electron donating 4-methyl group.
Experimental Section
Chemicals. N,N′-Dimethylformamide (Fluka puriss absolute) and
the supporting electrolyte n-Bu4NBF4 (Fluka puriss) were used as
received. Phenacyl chloride and bromide were used as supplied by
Aldrich, Lancaster, or Maybridge Chemical Co. 3 was prepared as
described in ref 24. Phenacyl benzoates (4a, 4b, and 4c) were obtained
according to literature procedure25 from the corresponding phenacyl
bromides and benzoic acid in the presence of triethylamine. Phenoxy
derivatives (5a, 5b) were synthesized, under phase-transfer-catalyzed
conditions,26a from the reaction between a phenacyl bromide and
m-cresol in presence of sodium hydroxide.26b Phenacyl sulfides (7a,
7b) were prepared by reacting phenacyl bromide with the sodium salt
of the appropriate thiol. 7b was obtained under the same catalytic
conditions26a as the phenoxy derivatives.
r-(Benzoyloxy)acetophenone 4a: mp 118 °C; IR (CHCl3) 1725,
1705 cm-1; NMR (CDCl3) δ 5.45 (s, 2H), 7.20-8.15 (10H).
Conclusions
r-(Benzoyloxy)-p-nitroacetophenone 4b: mp 140 °C; IR 1725,
1710 cm-1; NMR δ 5.50 (s, 2H), 7.20-8.35 (9H).
r-(m-Methylphenoxy)acetophenone 5a: mp 70 °C; IR 1690 cm-1
NMR δ 5.15 (s, 2H), 6.75-8.00 (9H).
Both concerted and stepwise mechanisms were identified by
means of cyclic voltammetry in the investigated series of
R-substituted acetophenones. In the stepwise cases, the cleavage
rates and the standard potentials for the formation of the anion
radical were derived from peak potential and peak width data.
With the help of thermodynamical parameters obtained from
the present study and from literature data, it was possible to
identify and estimate the parameters that govern the concerted-
stepwise dichotomy.
With acetophenones that do not bear electron withdrawing
ring substituents, the two main governing factors are the bond
dissociation energy and the oxidability of the leaving group.
With only one exception, the latter parameter is largely
predominant. It may even reverse an opposing effect of the
bond dissociation energy. The same is true for compounds
where the cleavage of the anion radical is so slow that it is
overrun by dimerization. R-Substituted acetophenones thus
provide several striking examples of the role of the oxidability
of the leaving group as a mechanism controlling parameter.
The fluoro-substituent provides a converse example where
the strength of the bond is so large that it overcompensates the
unfavorable effect of the oxidability of F-.
;
r-(m-Methylphenoxy)-p-methylacetophenone 5b: mp 68 °C; IR
1685 cm-1; NMR δ 5.15 (s, 2H), 6.75-7.90 (8H).
Cyclic Voltammetry. The electrodes were carefully polished and
ultrasonically rinsed with ethanol before each voltammogram. The
ultramicroelectrodes were built from a gold wire (10 and 25 µm
diameter) by using a reported procedure.27 The counter-electrode was
a platinum wire and the reference electrode an aqueous SCE electrode.
The potentiostat, equipped with a positive feedback compensation and
current measurer, used from 0.1 V/s until 500 V/s was the same as
previously described.28a The instrument used with ultramicroelectrodes
at high scan rates has been described elsewhere.28b The cyclic
voltammetry experiments were carried out at 20 °C using a cell
equipped with a double-wall jacket allowing circulation of water. At
low temperature, the cell was thermostated by an isopropyl alcohol
circulation and the reference electrode was maintained at 20 °C (the
bridge containing the reference electrode was equipped with a double-
wall jacket allowing circulation of water).
Supporting Information Available: Equations representing
the construction of the diagrams of Figure 1 and figures of
reversible cyclic voltammograms, cyclic voltammetric peak
characteristics, and theoretical variations of Ep and Ep/2 - Ep
with log V (5 pages). See any current masthead page for
ordering and Internet access instructions.
With electron withdrawing para-substituents such as, NO2 and
CN, the passage to a stepwise mechanism or the formation of
a noncleaving anion radical is the result of a decrease in the
LUMO energy.
With compounds following the stepwise mechanism, solvent
reorganization appears in most cases as the major contributor
to the intrinsic barrier of the cleavage reaction. It arises from
the transfer of the negative charge initially located on the
carbonyl oxygen to the leaving group. In this case too the fluoro
compound exhibits a particular behavior. Not only is the solvent
reorganization energy bigger, owing to the small size of the
fluoride ion, but also the strength of the C-F is such that
intramolecular reorganization is also much larger than with the
other compounds.
JA963674B
(24) Elkik, E.; Assadi-Far, H. Bull. Soc. Chim. Fr. 1970, 991.
(25) Sheehan, J-C.; Umezawa, K. J. Org. Chem. 1973, 38, 3771.
(26) (a) Chin-Hsien, W.; Xiang-Te, L.; Xaio-Hun, C. Synthesis 1982,
858. (b) The authors are grateful to A. Benchettara, U.S.T.H.B. Alger, for
his active contribution in the synthesis of phenacyl derivatives.
(27) Andrieux, C. P.; Garreau, D.; Hapiot, P.; Pinson, J.; Save´ant, J.-M.
J. Electroanal. Chem. 1988, 243, 321.
(28) (a) Garreau, D.; Save´ant, J.-M. J. Electroanal. Chem. 1972, 35, 309.
(b) Garreau, D.; Hapiot, P.; Save´ant, J.-M. J. Electroanal. Chem. 1989,
272,1.