Irreversibility of single electron transfer occurring from trivalent phosphorus compounds to iron(III) complexes in the presence of ethanol

Article abstract of DOI:10.1246/bcsj.75.1311

Various types of trivalent phosphorus compounds (1; Ph_(3-n)P(OR)_n) underwent single electron transfer (SET) to unsubstituted (2H) and 5-chloro-substituted tris(1, 10-phenanthroline)iron(III) complexes (2Cl) in the presence of ethanol in acetonitrile, resulting in the reduction of 2 to the corresponding iron(II) complexes. The rate constants (k_p) for the overall SET process were determined spectrophotometrically to show that within each series of 1 with an identical alkoxy group OR, log k_p correlates linearly with the difference in the half-wave potentials (ΔE_1/2) between 1 and 2. The slope of each correlation line gave an α-value for each series of 1. The α-values were significantly smaller than unity, indicating that the SET step is irreversible, even though this step is endothermic. The trivalent phosphorus radical cation 1^.+ generated in the SET step undergoes a rapid ionic reaction with ethanol, which is certainly the origin of the irreversibility. Upon examining the α-values more closely, it was found that the transition state of the SET step becomes earlier with increasing bulkiness of the substituent OR. It is concluded that 1 and 2 form a tight encounter complex to undergo SET from the former to the latter.

Full text of DOI:10.1246/bcsj.75.1311

c. Jpn.
© 2002 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 75, 1311–1318 (2002) 1311
e Chemical
ciety of Ja-
n
e Chemical
ciety of Ja-
n
Irreversibility of Single Electron Transfer Occurring from Trivalent
Phosphorus Compounds to Iron(Ⅲ) Complexes in the Presence of
Ethanol
SET from Trivalent Phosphorus Compounds
S. Yasui et al.
02
11
Shinro Yasui,* Kenji Itoh,† Munekazu Tsujimoto, and Atsuyoshi Ohno†, #
18
Laboratory of Biology and Chemistry, Tezukayama College, Gakuen-Minami, Nara 631-8585
†Institute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011
SJA8
09-2673
353
(Received October 1, 2001)
01
02
Various types of trivalent phosphorus compounds (1; Ph(3-n)P(OR)n) underwent single electron transfer (SET) to un-
substituted (2H) and 5-chloro-substituted tris(1, 10-phenanthroline)iron(Ⅲ) complexes (2Cl) in the presence of ethanol
in acetonitrile, resulting in the reduction of 2 to the corresponding iron(Ⅱ) complexes. The rate constants (kp) for the
overall SET process were determined spectrophotometrically to show that within each series of 1 with an identical
alkoxy group OR, log kp correlates linearly with the difference in the half-wave potentials (∆E1/2) between 1 and 2. The
slope of each correlation line gave an α-value for each series of 1. The α-values were significantly smaller than unity, in-
dicating that the SET step is irreversible, even though this step is endothermic. The trivalent phosphorus radical cation
1·+ generated in the SET step undergoes a rapid ionic reaction with ethanol, which is certainly the origin of the irrevers-
ibility. Upon examining the α-values more closely, it was found that the transition state of the SET step becomes earlier
with increasing bulkiness of the substituent OR. It is concluded that 1 and 2 form a tight encounter complex to undergo
SET from the former to the latter.
.1
3.1
Single electron transfer (SET) is a fundamental process of
chemical reactions. The energetics of the SET processes oc-
curring in many types of redox pairs has been studied on both
theoretical and experimental bases.1,2 Nevertheless, concern-
ing to energetics of SET from trivalent phosphorus compounds
Z3P, very few are known in spite of the fact that Z3P is a good
one-electron donor to several types of electron-deficient com-
pounds.3–10
Recently, we found that Z3P undergoes SET to a dye,
rhodamine, in the photoexcited state, Rho+*.11 By analyzing
this process kinetically, we showed that the SET step in the
process remains irreversible, even when the SET step becomes
endothermic. This finding is in sharp contrast to the fact that
SET from amine counterparts usually takes place in a com-
pletely reversible way in the endothermic region.1,12 Endother-
mic irreversible SET has been acknowledged to occur when a
SET step is followed by a rapid chemical reaction.13–23 In oth-
er words, in competition between the back SET and the “fol-
low-up” reaction, the SET becomes irreversible if the latter
overcomes the former. This idea was applied to our reaction of
Z3P with Rho+*. Thus, we have concluded that in the SET oc-
curring from Z3P to Rho+*, the trivalent phosphorus radical
cation Z3P·+, generated in the SET step, undergoes a rapid ion-
ic reaction with water in the solvent; the ionic reaction com-
petes effectively with the back SET to render the SET step irre-
versible.11 Meanwhile, a kinetic model for endothermic irre-
versible SET presents an α-value, a parameter that is defined
by the slope of a correlation line between the logarithm of the
rate constant of the overall SET process and a free-energy
change in the SET step.13,15 This parameter is a measure of the
position of the transition state in the SET step. Therefore, we
can obtain more detailed information about a SET step, itself,
by examining the α-values while varying the structure of the
Z3P, reaction conditions, and so on.
A family of iron(Ⅲ) complexes is a typical one-electron ac-
ceptor, which has been used in a kinetic study of the SET pro-
cesses.24–26 These complexes have reduction potentials of
around 0.8 V vs Ag/Ag+, a value that is slightly lower than the
oxidation potentials of most of trivalent phosphorus com-
pounds, Z3P. Therefore, SET from Z3P to an iron(Ⅲ) complex,
if it takes place, would give an opportunity to examine the en-
dothermic SET. In a preliminary report,27 we showed that the
endothermic SET takes place from various types of Z3P to an
iron(Ⅲ) complex in the presence of ethanol. The SET step was
irreversible despite its endothermicity, showing that the radical
cation, Z3P·+, generated in the SET step, undergoes a rapid
ionic reaction with ethanol. More importantly, the α-value ob-
tained in the kinetic analysis was dependent on the bulkiness
of ligands Z on the phosphorus atom in Z3P; as the ligands on
the phosphorus become bulkier, the α-value becomes smaller,
showing that the steric bulk of Z3P determines the position of
the transition state of SET from Z3P.
In the present study, we elucidated the dependence of the α-
value in SET from Z3P to an iron(Ⅲ) complex on steric bulk of
Z3P. For this purpose, we took trivalent phosphorus com-
pounds 1a-n, in which ligands Z on the phosphorus were
changed systematically in the term of bulkiness, and allowed 1
# Present address: Fukui University of Technology, Fukui 910-
8505

1312 Bull. Chem. Soc. Jpn., 75, No. 6 (2002)
SET from Trivalent Phosphorus Compounds
Scheme 1.
to react with unsubstituted (2H) and 5-chloro-substituted
tris(1, 10-phenanthroline)iron(Ⅲ) complexes (2Cl) in the pres-
ence of ethanol. The reaction affords the corresponding redox
products, oxo-pentavalent compound 3 and iron(Ⅱ) complex 4,
providing the stoichiometry shown in Scheme 1. A conclusion
can be made that the SET takes place from 1 to 2 within a tight
encounter complex.
(= d[4]/dt) were plotted against the concentration of 1. Al-
though the linearity of the plot was not so excellent (correla-
tion coefficient r > 0.99), the slight scattering on the plot is
highly likely attributable to uncertainties with respect to the
concentration of 1 prepared in reservoir cells of a stopped-flow
apparatus because the scattering was not systematic, but ran-
dom. Thus, it is clear that the reaction obeys second-order ki-
netics with first-order with respect to the concentrations of 1
and 2, respectively. Taking into account the observed stoichi-
ometry, we obtain
Results
Reaction of Trivalent Phosphorus Compounds (1) with
Tris(1, 10-phenanthroline)iron(Ⅲ) Complex (2). When 1
−
was reacted with 2 (PF6 salt) in acetonitrile in the presence of
kobs = 2kp[1][2].
(1)
a large excess of ethanol under an argon atmosphere in the
dark at 25 °C, the blue color from 2 changed rapidly to dark
red. The resulting solution gave an UV-visible spectrum on a
spectrophotometer that is the same as that of the authentic
sample of tris(1, 10-phenanthroline)iron(Ⅱ) complex (4) (λmax
= 508 nm), indicating that the color change results from the
reduction of 2 to 4.24 Meanwhile, the reaction mixture of
triphenylphosphine 1a with 2H was analyzed on a gas chro-
matograph (GC) and a gas chromatograph-mass spectrometer
(GCMS) to show the formation of triphenylphosphine oxide
3a in a theoretical amount (Scheme 1).
To determine the stoichiometry of the reaction, 2 at a con-
stant concentration (1.00 × 10−2 M) (1 M = 1 mol dm−3) was
reacted with 1 at different concentrations. The yield of 4 from
each reaction was determined based on the absorbance at 508
nm, which gave the stoichiometry [1]0:[2]0 = 1:2. As typical
examples, the results obtained in the reactions of 1a and 1b
with 2H are shown in Fig. 1.
Kinetics. Kinetics was carried out under pseudo-first-or-
der conditions where the concentrations of 1, 2, and ethanol
were 2.00 × 10−3 M, 1.00 × 10−1 M, and 1.00 × 10−2 M, re-
spectively. Rapid growth of the absorbance at 508 nm from 4
was monitored on a stopped-flow spectrophotometer to follow
the progress of the reaction. For each reaction with 1a-n, first-
order kinetics was maintained at least up to 90% formation of
4. Experiments were performed with 1 at different concentra-
tions, and the observed pseudo-first-order rate constants kobs
Table 1 summarizes the second-order rate constants (kp) deter-
mined based on Eq. 1, the errors in kp resulting from the scat-
tering in the kobs − [1] plots (vide supra). The rate constant
(kp) was kept constant by increasing the concentration of etha-
nol up to 1.70 × 10−1 M. In addition, when ethanol was re-
placed by methanol, propanol, 1-butanol, 2-propanol, or t-bu-
tyl alcohol, kp was identical within 5%, which was well within
the range of the experimental errors.
Kinetics was carried out under identical conditions for the
reactions of dibenzylamine 5a and t-butylamine 5b with 2.
The second-order rate constants (kp) for these reactions are
also listed in Table 1.
Peak Potentials. The peak oxidation potentials (Epox) of
1a-n28 and 5 and the peak reduction potentials (Epred) of 2 were
determined by cyclic voltammetry (CV). These values were
independent of the scan rate over the range 50 – 400 mV s−1
within an instrumental error, indicating that the anodic oxida-
tion of 1 and 5 and the cathodic reduction of 2 were nearly re-
versible. The reproducibility of each measurement was satis-
factory. From these Epox and Epred values, the half-wave poten-
tials (E1/2) of these compounds were estimated.29 The E1/2 val-
ues of 1 and 5 are listed in Table 1. The E1/2 values of 2H and
2Cl are 0.76 V and 0.86 V vs Ag/Ag+, respectively.
Discussion
Mechanism of the Reaction of 1 with 2.
The iron(Ⅲ)

S. Yasui et al.
Bull. Chem. Soc. Jpn., 75, No. 6 (2002) 1313
moles of 2 to 4; that is, the stoichiometry of the reaction should
be [1]0:[2]0 = 1:2, which has in fact been observed (Fig. 1).
In further support of this mechanism, when the reaction of 1a
was carried out in the presence of benzyl alcohol in place of
ethanol, benzyl ether was found on GC, although the yield was
low.32
Endothermic SET. Let us assume a SET that takes place
according to Scheme 3. Thus, an electron donor (D) and an ac-
ceptor (A) initially form an encounter complex at a diffusion-
limited rate (k12); then, SET takes place within the complex to
give a pair of D·+ and A·− (k23), which is followed by a follow-
up reaction (k30). A steady-state approximation with respect to
the encounter complex and the radical pair provides a rate ex-
pression,
k12k23k30
kp =
.
(2)
(
)
k32 + k30 )(k21 + k23 −k32k23
When the SET step is endothermic, i. e., ∆G23 > 0,33 the rate
expression depends on whether the SET step is reversible or ir-
reversible.11,13 If the SET step is reversible (k32 >> k30), the
rate is expressed by
kp = K12k30exp (−∆G23/RT)
(3)
with k12/k21 = K12 and k23/k32 = K23 = exp (−∆G23/RT). On
the other hand, if the SET step is irreversible (k32 << k30),
with the aid of the Horiuchi–Polanyi equation,34 we obtain
kp = K12kiexp (−α∆G23/RT),
(4)
where ki is the rate constant for a reaction when ∆G23 = 0, and
α is a proportionality constant that takes a value from 0 to 1.
Here, ∆G23 is related to the difference in the half-wave poten-
tials between 1 and 2 (∆E1/2) by Eq. 5, where F is the Faraday
constant.
Fig. 1. Plots of the yield of iron(Ⅱ) complex 4H against the
ratio of initial concentrations of the reactants in the reac-
tions of Ph3P (1a) (top) and Ph2P(OMe) (1b) (bottom), re-
spectively, with iron(Ⅲ) complex 2H. The reactions were
carried out in acetonitrile under argon atmosphere at 25 °C
in the dark. [2H]0 = 1.00 × 10−2 M and [EtOH]0 = 1.00
× 10−1 M.
∆G23 = ∆G0 − (wr + wp)
ꢀ F∆E1/2 − (wr + wp),
(5)
where ∆G0 is the free-energy change for the overall SET pro-
cess, and wr and wp are work terms that stand for the work re-
quired to bring two components at certain distances in the en-
counter complex and the resulting pair, respectively. A reason-
able assumption has been made that ∆E1/2 is nearly equal to the
difference in the standard redox potentials (∆E0). Finally, Eqs.
3 and 4, respectively, are rewritten as
complex 2 was reduced to the iron(Ⅱ) complex 4 upon a reac-
tion with a trivalent phosphorus compound 1. Certainly, the
reduction results from single electron transfer (SET) from 1 to
2. Evidence for the SET occurring in the present reaction is a
concomitant conversion of 1 to the corresponding oxo-pen-
tavalent compound 3; the SET generates the trivalent phospho-
rus radical cation 1·+, which collapses to 3 in the presence of
alcohol according to the sequence shown in Scheme 2. In this
sequence, a nucleophilic attack upon 1·+ by ethanol can easily
take place; several types of trivalent phosphorus radical cations
easily undergo a nucleophilic attack by a nucleophile, such as
alcohol,4 water,4,10 thiol,4 or pyridine.5 Oxidation of the result-
ing phosphoranyl radical 1·-OR by a second molecule of 2 to
the corresponding phosphonium cation 1+-OR is energetically
highly favorable, because 1·-OR has a much lower oxidation
potential than the parent trivalent phosphorus compound 1.30,31
This mechanism predicts that each mole of 1 reduces two
kp = K12k30exp [−F{∆E1/2 − (wr + wp)}/RT]
(6)
and
kp = K12kiexp [−αF{∆E1/2 − (wr + wp)}/RT].
(7)
Under the postulate that K12, k30, and ki are constant irrespec-
tive of ∆G23, Eq. 6 predicts that when the SET step is revers-
ible, log kp correlates linearly with ∆E1/2, with the slope being
−F/(2.3RT) (= −16.8 at 25 °C). Meanwhile, Eq. 7 states that
a linear correlation between log kp and ∆E1/2 also holds when

1314 Bull. Chem. Soc. Jpn., 75, No. 6 (2002)
SET from Trivalent Phosphorus Compounds
Ⅲ
Table 1. Reaction of Trivalent Phosphorus Compound 1 with Fe Complex 2 in the Presence
of EtOHa)
Z3P(1)
Ph3P (1a)
Ph2P(OMe) (1b)
PhP(OMe)2 (1c)
P (OMe)3 (1d)
2b)
kp/M−1 s−1
E1/2/Vc)
∆E1/2/Vd)
2H
2H
2H
2H
2Cl
2H
2H
2Cl
2H
2Cl
2H
2H
2H
2Cl
2H
2H
2H
2H
2H
(2.0 0.15) × 105
(7.4 0.45) × 103
(9.4 0.49) × 10
(2.5 0.00) × 10−1
1.8 0.16
0.91
1.21
1.49
1.81
0.15
0.45
0.73
1.05
0.95
0.41
0.71
0.61
1.11
1.01
0.40
0.60
1.07
0.97
0.38
0.62
1.09
0.37
0.44
Ph2P(OEt) (1e)
PhP(OEt)2 (1f)
(1.6 0.11) × 104
(4.6 0.16) × 102
(1.2 0.028) × 104
1.3 0.14
(1.4 0.09) × 10
(2.8 0.08) × 104
(2.5 0.19) × 103
7.4 0.05
1.17
1.47
P(OEt)3 (1g)
1.87
Ph2P(OPri) (1h)
PhP(OPri)2 (1i)
P(OPri)3 (1j)
1.16
1.36
1.83
(9.0 0.36) × 10
(2.1 0.12) × 104
(7.4 0.36) × 102
1.4 0.07
Ph2P(OBu) (1k)
PhP(OBu)2 (1m)
P(OBu)3 (1n)
1.14
1.38
1.85
1.13
1.20d)
(PhCH2)2NH (5a)
(7.7 0.30) × 103
ButNH2 (5b)
(8.2 0.22) × 102
a) In MeCN under an argon atmosphere in the dark at 25 °C. [1]0 = (2.00–8.00) × 10−3
M, [2]0 = 1.00 × 10−4 M, [EtOH]0 = 1.00 × 10−2 M. b) 2H; Fe (phen)3 PF6−, 2Cl;
Ⅲ
Ⅲ
Fe (5-Cl-phen)3 PF6−. c) vs Ag/Ag+. Esitmated from the peak oxidation potential by
using an equation, E1/2 = Epox −0.03. d) Difference in half-wave potentials between 1
and 2, E1/2(1)−E1/2(2). d) Ref. 12.
Scheme 2.
Scheme 3.
the SET step is irreversible, but the slope is −αF/(2.3RT) in
radicals, followed by an ionic reaction of 1·+ with ethanol
(Scheme 4). Besides, the half-wave potentials (E1/2) of triva-
lent phosphorus compounds 1a-n as well as those of amines
this case. Since the α-value is less than unity, the slope is less
negative in the latter case than in the former. In analogy with
the Brönsted coefficient, the α-value is an index used to mea-
sure the position of the transition state of the SET under dis-
cussion; α becomes smaller with the transition state of the SET
step going to the reactant side. Equations equivalent to Eq. 7
have been presented by us11 and by other groups13,15 to discuss
the position of the transition state of the SET step.
Energetics of the SET from 1 to 2. The SET from 1 to 2
takes place according to Scheme 3. In this case, a follow-up
reaction (k30-step) is the separation of the ion pair into free-
Scheme 4.

S. Yasui et al.
Bull. Chem. Soc. Jpn., 75, No. 6 (2002) 1315
5a-b are higher than those of 2H and 2Cl, thus showing the
in Fig. 2). Although the kp and ∆E1/2 values include experi-
mental errors, these uncertainties are no larger than the range
of the symbols in Fig. 2 (see Results section). Thus, these
three lines are clearly distinguishable. From the slopes of
these lines, the α-values for the reactions of 1 are calculated
based on Eq. 7. As summarized in Table 2, the α-value be-
comes smaller in the order of R = Me > Et ꢀ Bu > Pri in 1
(Ph(3-n)P(OR)n). That is, as the substituent OR on the phospho-
rus becomes bulkier, the transition state of the SET step be-
comes earlier.
This observation is interpreted in the term of SET occurring
within the encounter complex [1 ≥ 2] to give the complex [1·+
≥ 4]. Complexation is tighter in the former complex, “reac-
tant” of the SET step, than in the latter, “product” of the SET
step, because in the latter complex both components are posi-
tively charged. As a result, destabilization caused by a bulky
ligand on the phosphorus is more significant in the former
complex than in the latter. Thus, if we suppose two extreme
cases here, namely, the reaction of a compound with a bulky
ligand and that for a small ligand, then we can make up energy
diagrams as illustrated in Fig. 3, which represent how the com-
pound with a bulky ligand (solid curves) and the compound
SET step (k23-step) examined here to be endothermic (∆G23
>
0). That is, the SET from 1 as well as 5 to 2 is interpreted in
terms of the endothermic SET formulated above.
In Fig. 2, are plotted the logarithms of the rate constants kp
(Table 1) against ∆E1/2. In this figure, the points for the report-
ed values of the rate constants in the SET from alkylbenzenes
to 2H are also given. The SET from alkylbenezens is known to
take place in a completely reversible way24 and, as expected,
these points make up a line whose slope is predicted by Eq. 6
(dotted line). Importantly, the points for SET from dibenzy-
lamine 5a and t-butylamine 5b to 2 also fall on this line, show-
ing that the SET from amines 5 to 2 is reversible. In contrast,
the points for the SET from 1 are largely deviated from the
dotted line. Apparently, although log kp for the SET from 1 to
2 depends negatively on ∆E1/2, the dependence is weaker than
predicted by Eq. 6. The kinetic behavior observed here is well
described by Eq. 7.35 Standing on a premise by which Eq. 7
was derived, the observation demonstrates that a follow-up
chemical reaction of trivalent phosphorus cation radicals 1·+
(k30) is rapid enough to overcome the back SET (k32). The fact
that kp is independent of either concentration or kinds of alco-
hols strongly suggests that this is the case. In other words, the
high reactivity of 1·+ toward ethanol results in an irreversible
SET in an endothermic region. The indispensability of a rapid
follow-up reaction for SET to occur is shown by the fact that
SET from 5 is reversible; amine radical cations 5·+ act as a
radical rather than a cation,36,37 and no free-radical which can
be coupled with the amine radical cations exists in the system.
Examining the points for 1 more closely, a linear correlation
is found to hold within each series of 1 (Ph(3-n)P(OR)n) with an
identical alkoxy group OR on the phosphorus atom (solid lines
Table 2. α-Values for the Reaction of 1 with 2a)
R in Ph(3−n)P(OR)n
α-Value
rb)
Me
Et
0.44 0.02
0.34 0.02
0.35 0.004
0.32 0.002
0.998
0.998
> 0.999
> 0.999
Bu
Pri
a) At 25 °C. b) Correlation coefficient of the correlation
between logkp and ∆E1/2 in Fig. 2.
Fig. 2. Dependence of log kp on ∆E1/2 in the reacrion of
Ph(3−n)P(OR)n (1) with iron(Ⅲ) complexes (2). Each solid
line is drawn for the reactions of a series of 1 with 2H; R =
Me (ꢀ), Et (ꢁ), Pri (ꢂ), and Bu (ꢃ). Symbols ꢄ, ꢅ, and
ꢆ denote the reactions of 1d (R = Me, n = 1), 1g (R = Et,
n = 1), and 1j (R = Pri, n = 1) with 2Cl, respectively. A
symbol denotes the reaction of 1a with 2H. Symbols ꢇ
Fig. 3. Schematic presentation of energy diagrams of the
SET reaction of 1 with 2. The reactions of the phosphorus
compounds with larger and smaller ligand(s) on the central
phosphorus proceed according to solid curves and dashed
curves, respectively.
and
denote the reactions of amines 5 and alkylben-
zenes, respectively, with iron(Ⅲ) complexes. A dashed
line is drawn based on Eq. 6.

1316 Bull. Chem. Soc. Jpn., 75, No. 6 (2002)
SET from Trivalent Phosphorus Compounds
with a small ligand (dashed curves) undergo reactions to afford
products. Clearly, the SET from a compound with a bulkier
ligand group has an earlier transition state.
SET” having appeared so far in the literature. In Table 3 are
collected examples of such reactions together with the α-val-
ues. There is no simple tendency in magnitude of α depending
on the rates of the follow-up reactions. For example, SET
from ascorbic acid to an iron(Ⅲ) complex (in the bottom row
in Table 3) is followed by the second SET that must be very
fast, but its α-value is the largest, that is, the transition state is
the latest, among the reactions listed in this table. Reorganiza-
tion of the reactants associated with SET could play an impor-
tant role in determining the position of the transition state of
the SET step.
Finally, we should examine our kinetic data in terms of the
conventional theories. As we have seen before, the kinetics of
the SET from 1 does not follow the prediction by Eq. 6. Since
Eq. 6 (as well as Eq. 3) is an approximated version of the
Rehm–Weller equation for SET occurring in an endothermic
region, the observation shows that the SET from 1 is not in a
category of the Rehm–Weller’s model. On the other hand, it
seems that such a kinetic behavior could be analyzed based on
the Marcus equation.2 However, any set of variables in the
equation in fact failed to produce a quadratic curve that traces
the experimental points. The “classical” Marcus theory has
been evolving to afford several types of theories.39–42 For ex-
ample, the kinetics of nonadiabatic SET followed by dissocia-
tion of a covalent bond is well described in terms of an
“evolved” Marcus theory, which has been developed while tak-
ing into account quantum mechanics as well as modified ener-
getics brought about by a follow-up reaction.42 Although the
kinetic behavior observed in the present study may be explain-
able in terms of these theories, unfortunately our data set is too
limited to allow such a discussion. In any case, it should be
emphasized that a kinetic analysis according to Eq. 7 is not
against the concept of the Marcus theory. The present study
succeeded to abstract simplified aspects from intrinsically
complicated kinetics of SET.
The difference in the free-energy at the encounter complex
stage corresponds to the difference in the work term (wr) in Eq.
7. Such a situation, found in the present study, is in contrast to
that in the typical outer-sphere SET, for which the work term
(wr) is often negligible. In other words, for the SET from 1 to
2 to undergo, the distance between the reactants within the en-
counter complex is required to be rather short. In this connec-
tion, an experimental fact is worth noting: an “ate” complex
[IrCl6]2−, whose reduction potential is only about 0.1 V lower
than that of 2,26a did not give rise to the SET from 1. If the dif-
ference in the half-wave potentials (which is sometimes termed
“driving force”) is the only factor to determine the reaction
rate, the rate constant (kp) of the reaction of 1a with the iridate
[IrCl6]2− would be in the order of 105 M−1 s−1. The observed
inertness of the iridate to undergo SET is due to the intrinsical-
ly electron-rich character of the phosphorus compound 1,
which prevents the iridate from approaching 1 within an “ef-
fective” distance. It has been shown that SET from alkylmet-
als, such as R4Sn to the iron(Ⅲ) complex 2 and the iridate
[IrCl6]2−, is subject to steric effects.38 When the steric hin-
drance is significant between the reductant and the oxidant,
SET takes place according to an outer-sphere mechanism. On
the other hand, the alkylmetal undergoes an inner-sphere SET
to a less-hindered reductant, namely, the iridate, via a tight en-
counter complex. An electrostatic force has been proposed to
be a contributor to result in the tightness of the encounter com-
plex. The consideration here thus demonstrates an inner-
sphere character of the SET from 1 to 2, where a tight encoun-
ter complex is formed prior to the SET.
The reaction of 1 with the chloro-substituted iron(Ⅲ) com-
plex 2Cl, except for the reaction of methyl ester 1d, has ac-
quired acceleration in the rate compared with the reactions
with the unsubstituted complex 2H. As can be seen in Fig. 2,
the points for the reactions with 2Cl are deviated upward from
the lines made up by the points for the corresponding reactions
with 2H. Since 2Cl is bulkier than 2H, a release of the steric
congestion associated with the SET is more significant in the
reaction with 2Cl than in the reaction with 2H, resulting in a
rate acceleration for the reactions with 2Cl. Such a steric ef-
fect has vanished in the reaction of 1d. The methyl ester 1d is
too small to recognize the difference in size between 2H and
2Cl.
Summary
We studied kinetically SET from trivalent phosphorus com-
pounds (1; Ph(3-n)P(OR)n) to iron(Ⅲ) complexes 2 in acetoni-
trile containing ethanol. By examining the α-values obtained
in a log kp − ∆E1/2 plot, we found that (1) the SET step is irre-
versible, even though it is endothermic, and (2) the transition
state of the SET step becomes earlier with increasing bulkiness
of OR.
The latter finding leads to the conclusion that a tight en-
counter complex is formed between 1 and 2 prior to the SET.
It is interesting to survey the α-values for an “irreversible
Table 3. Irreversible SET Reactions and α-Values
Donor
Acceptor
Supposed follow-up reaction
α-Value
Ref.
2, 6-Di-t-butylphenol
Tetraalkyllead (R4Pb)
Arene radical anion
Fused aromatic compound
Dihydroflavin
Singlet oxygen
Formation of a C–O bond
Cleavage of a C–Pb bond
Cleavage of a C–X bond
0.20a)
0.37a)
17
18
[Ir Cl6]2−
Ⅳ
Alkyl halide (RX)
Peroxide
0.22–0.46b) 14,15
Cleavage of an O–O bond
0.30b)
0.35a)
0.50a)
0.6a)
13,20
19
26b
23
Nitroxide
H+-transfer from oxidized flavin to reduced nitroxide
Ascorbic acid (HA−)
Ketene silyl acetal
[Fe L3]3+
SET from HAO to Fe
Ⅲ
Ⅲ
Trityl cation
C–C bond formation
a) Calculated from the kinetic data reported. b) The values given in the referred papers.

S. Yasui et al.
Bull. Chem. Soc. Jpn., 75, No. 6 (2002) 1317
2
3
R. A. Marcus, J. Phys. Chem., 72, 891 (1968).
S. Yasui, Rev. Heteroatom Chem., 12, 145 (1995), and ref-
Experimental
erences cited therein.
Instruments. GC and GCMS analyses were performed, re-
spectively, with a Shimadzu GC-14B gas chromatograph and with
a Shimadzu GCMS-QP2000A gas chromatograph-mass spectrom-
eter equipped with a Shimadzu GC-MSPAC 200S data processor.
UV/visible spectra were recorded on a Hitachi U-3212 spectro-
photometer. Kinetics was carried out with a Union Giken RA-401
stopped-flow spectrophotometer. Cyclic voltammetry was carried
out on a Cypress Systems OMNI90 potentiostat.
4
a) S. Yasui, K. Shioji, M. Tsujimoto, and A. Ohno, Chem.
Lett., 1995, 783. b) S. Yasui, M. Tsujimoto, K. Shioji, and A.
Ohno, Chem. Ber./Recl., 130, 1699 (1997). c) S. Yasui, K. Shioji,
M. Tsujimoto, and A. Ohno, J. Chem. Soc., Perkin Trans. 2, 1999,
855.
5
S. Yasui, K. Shioji, M. Tsujimoto, and A. Ohno, Heteroat-
om Chem., 11, 152 (2000).
6
(1969).
7
R. D. Powell and C. D. Hall, J. Am. Chem. Soc., 91, 5403
Materials. Triphenylphosphine (1a) was purchased (Nacalai
Tesque). Trivalent phosphorus compounds 1b–g, j, n were also
commercially available (Tokyo Chemical Industry). These mate-
rials were purified by recrystallization or distillation, if neces-
sary. Phosphinites 1h and 1k were obtained by the condensation
of chlorodiphenylphosphine with 2-propanol and 1-butanol, re-
spectively, in the presence of pyridine, followed by purification
through distillation. Phosphonites 1i and 1m were obtained in a
similar manner by the condensation of dichlorophenylphosphine
with 2-propanol and 1-butanol, respectively. Tris(1, 10-phenan-
throline)iron(Ⅲ) hexafluorophosphate (2H) and tris(5-chloro-1,
10-phenanthroline)iron(Ⅲ) hexafluorophosphate (2Cl) were syn-
thesized according to literature procedure.23
G. Pandey, S. Hajra, and M. K. Ghorai, Tetrahedron Lett.,
35, 7837 (1994).
8
S. Takagi, T. Okamoto, T. Shiragami, and H. Inoue, J. Org.
Chem., 59, 7373 (1994).
9
Z. B. Alfassi and P. Neta, J. Phys. Chem. A, 101, 2153
(1997).
10 M. Nakamura, M. Miki, and T. Majima, J. Chem. Soc., Per-
kin Trans. 2, 2000, 1447.
11 S. Yasui, M. Tsujimoto, K. Itoh, and A. Ohno, J. Org.
Chem., 65, 4715 (2000).
12 R. Ballardini, G. Varani, M. T. Indelli, F. Scandola, and V.
Balzani, J. Am. Chem. Soc., 100, 7219 (1978).
13 G. B. Schuster, J. Am. Chem. Soc., 101, 5851 (1979).
14 S. Bank and D. A. Juckett, J. Am. Chem. Soc., 97, 567
(1975).
15 J. Grimshaw, J. R. Langan, and G. A. Salmon, J. Chem.
Soc., Faraday Trans., 90, 75 (1994).
16 J.-M. Savéant, Acc. Chem. Res., 26, 455 (1993).
17 M. J. Thomas and C. S. Foote, Photochem. Photobiol., 27,
683 (1978).
Product Analysis. A solution of 1 ((0.25 – 1.00) × 10−4 M),
2 (1.00 × 10−4 M), and ethanol (1.00 × 10−2 M) in acetonitrile
was allowed to react at 25 °C under an argon atmosphere for 5
min. The resulting solution was analyzed on a spectrophotometer
to determine the yield of 4 based on the absorbance at 508 nm.23
To determine the yield of 3a, the reaction of 1a (1.00 × 10−2 M)
with 2 (1.00 × 10−2 M) in the presence of ethanol (1.00 × 10−1
M) was carried out under identical conditions. After 5 min, an ali-
quot of the reaction mixture was added to toluene containing
dodecane as an internal standard, and the organic layer was ana-
lyzed by GC and GCMS.
18 H. C. Gardner and J. K. Kochi, J. Am. Chem. Soc., 97, 1855
(1975).
19 T. W. Chan and T. C. Bruice, J. Am. Chem. Soc., 99, 7287
(1977).
20 a) J.-Y. Koo and G. B. Schuster, J. Am. Chem. Soc., 99,
6107 (1977). b) J.-Y. Koo and G. B. Schuster, J. Am. Chem. Soc.,
100, 4496 (1978).
21 A. J. Bard and A. Merz, J. Am. Chem. Soc., 101, 2959
(1979).
22 M. S. Workentin and R. L. Donkers, J. Am. Chem. Soc.,
120, 2664 (1998).
23 S. Fukuzumi, K. Ohkubo, and J. Otera, J. Org. Chem., 66,
1450 (2001).
24 a) C. J. Schlesener and J. K. Kochi, J. Org. Chem., 49,
3142 (1984). b) C. J. Schlesener, C. Amatore, and J. K. Kochi, J.
Am. Chem. Soc., 106, 3567 (1984).
25 a) M. Kimura and Y. Kaneko, J. Chem. Soc., Dalton Trans.,
1984, 341. b) M. Kimura, S. Yamabe, and T. Minato, Bull. Chem.
Soc. Jpn., 54, 1699 (1981).
Cyclic Voltammetry (CV). A solution of trivalent phospho-
rus compounds 1 or amines 5 (1.0 × 10−2 M) and a solution of
iron(Ⅲ) complexes (2) (5.0 × 10−2 M) were prepared in acetoni-
trile containing tetraethylammonium tetrafluoroborate (0.10 M) as
the supporting electrolyte. The CV of these solutions was carried
out at room temperature at a scan rate of 50 mV s−1. The working
electrode was platinum, and the reference electrode was Ag/Ag+
(in a solution of silver nitrate (0.01 M) and tetraethylammonium
perchlorate (0.1 M) in acetonitrile). The values of the peak poten-
tials were read on the obtained cyclic voltammograms.
Kinetics. Stock solutions of 1 and 2 were prepared so that
these concentrations became (2.00 – 8.00) × 10−3 M and 1.00 ×
10−4 M, respectively, after mixing. The solutions were placed into
separate reservoir cells of a stopped-flow spectrophotometer,
which were maintained at 25 °C and charged with argon gas. Af-
ter mixing, the increase in the absorbance at 508 nm was moni-
tored.
26 a) E. Pelizzetti, E. Mentasti, and E. Pramauro, Irong.
Chem., 17, 1181 (1978). b) E. Pelizzetti, E. Mentasti, and E.
Pramauro, Irong. Chem., 15, 2898 (1976).
27 S. Yasui, K. Itoh, M. Tsujimoto, and A. Ohno, Chem. Lett.,
1998, 1019.
28 The Ep values of some trivalent phosphorus compounds
have been reported in S. Yasui, M. Tsujimoto, M. Okamura, and
A. Ohno, Bull. Chem. Soc. Jpn., 71, 927 (1998).
This work was financially supported in part by a Grant-in-
Aid from Scientific Research (C) (No. 09640654) from the
Ministry of Education, Science, Sports and Culture. One of
the authors (S. Y.) greatly appreciates the financial support of
Tezukayama Research Grant in 2000.
ox
References
29 When a redox process at an electrode is reversible and one
red
electron is transferred at 25 °C, E1/2 = Epox −0.03 and E1/2 = Ep
1
D. Rehm and A. Weller, Isr. J. Chem., 8, 259 (1970).

1318 Bull. Chem. Soc. Jpn., 75, No. 6 (2002)
SET from Trivalent Phosphorus Compounds
Falvey, J. Org. Chem., 61, 3195 (1996).
37 B. D. Wagner, G. Ruel, and J. Lusztyk, J. Am. Chem. Soc.,
118, 13 (1996).
38 a) C. L. Wong and J. K. Kochi, J. Am. Chem. Soc., 101,
5593 (1979). b) S. Fukuzumi, C. L. Wong, and J. K. Kochi, J. Am.
Chem. Soc., 102, 2928 (1980).
39 a) P. Siders and R. A. Marcus, J. Am. Chem. Soc., 103, 741
(1981). b) P. Siders and R. A. Marcus, J. Am. Chem. Soc., 103,
748 (1981).
+ 0.03. See, for example, H. Oevering, M. N. Paddon-Row, M.
Heppener, A. M. Oliver, E. Cotsaris, J. W. Verhoeven, and N. S.
Hush, J. Am. Chem. Soc., 109, 3258 (1987).
30 J. M. Saveant and S. K. Binh, J. Org. Chem., 42, 1242
(1977).
31 J. A. Kampmeier and T. W. Nalli, J. Org. Chem., 58, 943
(1993).
32 Water may exist in the solvent in a considerable amount
and contribute as a nucleophile significantly to decomposition of
1·+, which explains a low yield of benzyl ether.
33 When the SET is exothermic (∆G23 < 0), k23 >> k21 holds,
and then kp = k12 is obtained.
40 a) M. Bixon and J. Jortner, Faraday Discuss. Chem. Soc.,
74, 17 (1982). b) M. Bixon and J. Jortner, Chem. Phys., 176, 467
(1993).
34 J. Horiuchi and M. Polanyi, Acta Physicochim. URSS, 2,
505 (1935).
41 C. P. Andrieux, J.-M. Savéant, and C. Tardy, J. Am. Chem.
Soc., 120, 4167 (1998).
35 The observation supports the postulate that K12 and ki are
indpendent of ∆G23.
36 D. Chiapperino, G. B. Anderson, R. J. Robbins, and D. E.
42 E. D. German and A. M. Kuznetsov, J. Phys. Chem., 98,
6120 (1994).