Aminosilanes as two-electron donors: A technological application of radical cation chemistry

Article abstract of DOI:10.1139/v03-073

Aminosilanes possess the appropriate structural features for use as two-electron sensitizers in silver halide photography. Here, we describe studies of the nucleophile-assisted cleavage reactions of the C - Si bonds in their radical cations. Water is identified as a useful nucleophile. It is found that the kinetics of these reactions are best described by taking into account a radical cation - water complex. The cleavage reactions are also controlled by steric effects at silicon and by the proximity of a carboxylate group that can modify the nucleophilicity of the water. Cleavage forms an α-amino radical that can donate a second electron to 9,10-dicyanoanthracene as a solution-phase electron acceptor.

Full text of DOI:10.1139/v03-073

777
Aminosilanes as two-electron donors: A
technological application of radical cation
chemistry
Ian R. Gould, Stephen A. Godleski, Paul A. Zielinski, and Samir Farid
Abstract: Aminosilanes possess the appropriate structural features for use as two-electron sensitizers in silver halide
photography. Here, we describe studies of the nucleophile-assisted cleavage reactions of the C—Si bonds in their radi-
cal cations. Water is identified as a useful nucleophile. It is found that the kinetics of these reactions are best described
by taking into account a radical cation – water complex. The cleavage reactions are also controlled by steric effects at
silicon and by the proximity of a carboxylate group that can modify the nucleophilicity of the water. Cleavage forms
an α-amino radical that can donate a second electron to 9,10-dicyanoanthracene as a solution-phase electron acceptor.
Key words: silane, radical cation, two-electron sensitization, fragmentation, steric effects.
Résumé : Les aminosilanes possèdent les caractéristiques structurales appropriées qui permettent de les utiliser comme
sensibilisateurs à deux électrons dans la photographie à l’halogénure d’argent. Nous décrivons les études portant sur les
réactions de clivage, avec assistance nucléophile, de la liaison C—Si dans leurs cations radicalaires. L’eau est identifiée
comme un nucléophile utile. On a trouvé que les cinétiques de ces réactions sont mieux décrites en tenant compte d’un
complexe cation radicalaire – eau. Les réactions de clivage sont contrôlées également par les effets stériques du sili-
cium, et par la proximité du groupe carboxylate qui peut modifier la nucléophilicité de l’eau. Le clivage conduit à un
radical α-amino qui peut donner un second électron au 9,10-dicyanoanthracène en tant qu’accepteur d’électron dans la
phase en solution.
Mots clés : silane, cation radicalaire, sensibilisation à deux électrons, fragmentation, effets stériques.
[Traduit par la Rédaction] Gould et al. 788
Introduction
second electron into the silver halide, potentially doubling
the efficiency of the photographic system (3, 4). This appli-
cation of radical cation chemistry is termed two-electron
sensitization (TES) of silver halide, and the amine precursor
to the radical cation is a two-electron sensitizer (3, 4).
For proper operation, a two-electron sensitizer must meet
several criteria. One of these is that the radical cation should
fragment on the microsecond timescale, preferably with a
rate constant, kfr, that is controllable (3). The fragmentation
reactions of the TES compounds described to date are
decarboxylation reactions of amine carboxylates, as shown
schematically in eq. [1]. The rate constants for fragmenta-
tion, kfr, of several of these radical cations were reported (3).
Fragmentation reactions of radical cations yield interest-
ing and useful catalytic species, such as radicals, protons,
and electrophiles (1). Pioneering work on these reactions by
Arnold and co-workers explored their scope and defined
their thermodynamics (2), and qualitative and quantitative
studies on a wide variety of radical cation fragmentation re-
actions have now been reported (1, 2). Recently, a techno-
logical application for radical cation fragmentation was
discovered in silver halide photography (3, 4). The primary
event in the photographic process is excitation of a sensitiz-
ing dye that results in one-electron injection from the ex-
cited dye into the silver halide (5). Trapping of the resultant
oxidized dye by electron transfer from an organic amine
forms the amine radical cation. Fragmentation of this radical
cation forms an α-amino radical that is capable of injecting a
[1]
Received 16 January 2003. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 25 June 2003.
They exhibit a strong dependence on the oxidation poten-
tial of the amine. A useful method of controlling the frag-
mentation rate constant that does not depend upon oxidation
potential or, even better, that could compensate for the effect
of oxidation potential, would provide a greater degree of
control in the application of these compounds. The develop-
ment of structures that provide such control is the subject of
the present paper, which describes fragmentation reactions
of aminosilane radical cations. These reactions are subject to
control by steric factors and respond to the medium in a
Dedicated to Donald R. Arnold, a pioneer in the chemistry of
radical cations.
I.R. Gould.1 Department of Chemistry and Biochemistry,
Arizona State University, Tempe, AZ 85287-1604, U.S.A.
S.A. Godleski, P.A. Zielinski, and S. Farid. Research
Laboratories, Eastman Kodak Company, Rochester, NY
14650-2109, U.S.A.
1Corresponding author (e-mail: igould@asu.edu).
Can. J. Chem. 81: 777–788 (2003)
doi: 10.1139/V03-073
© 2003 NRC Canada

778
Can. J. Chem. Vol. 81, 2003
different way than do the decarboxylation reactions. Frag-
mentation reactions of aminosilane radical cations have been
described previously by Mariano and co-workers, and ki-
netic measurements of these reactions were reported (6).
Here, we describe more detailed kinetic studies; we further
explore the factors that control the rate constants for frag-
mentation of these radical cations; and we provide solution-
phase evidence that aminosilanes can act as TES reagents.
Fragmentation of the C—Si bond in silane radical cations is
usually not a unimolecular reaction for many systems but
occurs via a nucleophilic-assisted mechanism, as indicated
in eq. [2], where X–SiR3•+ represents a generic silane radical
cation (6, 7). One consequence of this is that the bimolecular
rate constants for the desilylation reactions have been found
to be sensitive to steric effects, particularly with respect to
substitution at silicon (7).
substituent. Here, we describe a series of useful TES donors,
based on desilylation chemistry, which satisfy these various
requirements.
Results and discussion
Structural features and synthetic approach
The basic structure that has the important required fea-
tures is shown in eq. [3].
[3]
[2]
Cleavage of the C—Si bond generates the desired α-
amino radical. The β-propionate group serves as a water-
solubilizing functionality. Decarboxylation, as in eq. [1],
does not occur because this would generate a primary,
unstabilized alkyl radical.
For example, the reactivity of triethyl derivatives (1, R3 =
Et3) was observed to be almost an order of magnitude lower
than the corresponding trimethyl analogues (R3 = Me3) (7).
The reactivity of the derivative R3 = Me2-t-Bu was still
lower, and the R3 = i-Pr3 compound was dramatically less
reactive than the triethylsilane compound, R3 = Et3. Because
the substituents on silicon have only a small influence on the
oxidation potential of the compound, these donors offer an
obvious way to change, kfr, without changing the oxidation
potential of the radical cation precursor.
In the TES scheme, the radical X• should be reducing
enough to inject an electron into the conduction band of sil-
ver halide, i.e., it should have an oxidation potential that is
equal to, or more negative than, ca. –0.9 V vs. SCE (3).
From previous work on the carboxylate TES donors, a set of
structural features to meet this requirement for X• have been
identified (3). Specifically, a radical site that is α to nitrogen
in a dialkylaniline (Structure 2) is sufficiently reducing —
even when the radical is primary — so long as the aniline is
substituted with a strong electron-donating group, e.g., 4-
methoxy. However, with weakly electron-donating groups or
with electron-withdrawing groups in the 4 position, such a
radical must have secondary substitution (Structure 3) to be
sufficiently reducing.
The synthetic route employed is outlined in Scheme 1.
The tert-butyl ester of aniline β-propionate, prepared via Mi-
chael addition of tert-butyl acrylate to aniline, was reacted
with the appropriate trialkylsilane triflates (see experimental
section) to provide the tert-butyl ester silyl aniline deriva-
tives 4E and 5E. The tert-butyl group was removed with
trifluoroacetic acid, and the sodium carboxylate salt 4S was
obtained using sodium methoxide or sodium hydroxide. The
4-formyl tert-butyl ester derivatives 6E and 7E were ob-
tained from 4E and 5E via a Vilsmeier reaction. Removal of
the tert-butyl group as before gave the carboxylate 6S. The
structures of the compounds are summarized in Table 1.
Solution phase studies on the radical cations
Experimental approach
The rate constants for fragmentation of the radical cations
were obtained using transient absorption spectroscopy, using
9,10-dicyanoanthracene (DCA) as the excited-state electron
acceptor. The radical cations of the donor molecules, X–
SiR3•+, were formed by oxidation of the donor using a
biphenyl radical cation (C•+) in a cosensitization scheme
(Scheme 2) (8).
•+
Fragmentation of X–SiR3 leads to the formation of the
radical, X•. If X• is sufficiently reducing, it can donate an
electron to another DCA molecule so that two reduced ac-
ceptors are formed for one absorbed photon in an exact solu-
tion phase analogue of the TES scheme (3). Formation of
this second-reduced DCA can easily be observed in a time-
resolved absorption experiment when kfr is not so small that
its formation is significantly slower than the processes that
remove the DCA•–. Indeed, observation of two DCA reduc-
tion processes represents a convenient method for confirma-
tion of fragmentation of the radical cations, because it
provides evidence for formation of the (X•) radical cleavage
product. In general, reduction of two DCA molecules is ob-
served for all of the silane radical cations studied here, when
kfr is greater than ca. 106 s–1 (see for example, Fig. 1 below).
As discussed further below, in some cases the X• radicals
could be directly observed. When evidence of radical forma-
To generate such radicals via desilylation reactions,
trialkylsilicon groups should be placed at the α positions of
dialkylanilines. In addition, to be useful for photographic
purposes, such compounds require a water-solubilizing
© 2003 NRC Canada

Gould et al.
779
Scheme 1.
Scheme 2.
tion was obtained in this way, the rate constant for fragmen-
tation of the radical cation, kfr, was simply equated to the
observed rate constant for its pseudo-first-order decay.
The decay kinetics of the donor radical cation can be
monitored in different ways, depending on the system under
investigation. One problem is that there is no analyzing
wavelength where only the radical cation absorbs, i.e., con-
tributions from the DCA radical anion (DCA•–) to the overall
absorption decay kinetics are always observed. The simplest
way to deal with this is to saturate the solution with oxygen,
which scavenges DCA•– to form the nonabsorbing super-
oxide (O2•–). Under these conditions, a fast-absorption decay
is observed, owing to the reaction of DCA•–, with a pseudo-
With the 4-formyl derivatives 6 and 7, however, the radicals
can be observed at long wavelengths (>730 nm). In this
wavelength region, the radical cation and radical anion ab-
sorb only weakly, and grow-in of X• can be observed as a
result of fragmentation of the X–SiR3•+. As discussed above,
however, the radicals also react with the DCA sensitizer to
form DCA•–, so that the exponential grow-in is followed by
an exponential absorption decay with a pseudo-first-order
rate, kDCA, that depends upon the DCA concentration,
(Scheme 2). One interesting feature of such a double expo-
nential kinetic scheme is that the rates can become “re-
versed” (9). This is the case if the rate of decay of the
radical caused by the reaction with DCA, kDCA, is faster than
the rate at which it is formed owing to the fragmentation re-
action, kfr. Under these conditions, the rate of growth of the
absorption signal is equal to kDCA (i.e., the process that re-
moves the X•) and the rate of the decay of the absorption
signal is equal to kfr (i.e., the process that forms the X•) (9).
The two processes can be distinguished, and the fragmenta-
tion rate correctly identified, by changing the concentration
of DCA, which affects only the pseudo-first-order kDCA and
not kfr.
first-order rate constant, kO , which is usually ca. 7 × 106 s–1
2
in aqueous acetonitrile. This is followed by a slower pseudo-
first-order decay as a result of the reaction of the donor radi-
cal cation. This method is easily implemented when the X–
•+
SiR3 fragmentation rate is much slower than the rate of re-
action of the DCA•– with oxygen. As kfr increases to ca. 2–3 ×
106 s–1, it becomes necessary to analyze the observed ab-
sorption decay as a sum of two exponentials.
In principle, the fragmentation rate can also be determined
from the rate of growth of the radical product, X•
(Scheme 2). The radical products derived from 4 and 5 can-
not be readily observed in the visible region of the spectrum.
Finally, under certain conditions, an estimate for kfr can be
obtained from analysis of the growth of the second DCA•–
signal because of the reaction of X• with DCA. When kfr and
© 2003 NRC Canada

780
Can. J. Chem. Vol. 81, 2003
Table 1. Kinetic parameters for fragmentation of aminosilane radical cations in acetonitrile, using water as a nucleophile, at room tem-
perature.a
d
KeqkNu
Keq (M–1)
kNu (s–1)
k0 (s–1)
e
b
c
Structure
Compound number
(M–1 s–1)
1.4 × 105
~1 × 105
~5 × 104
~5 × 104
~4 × 105
4S
0.1
1.4 × 106
5.1 × 106
4.3 × 105
1.4 × 107
4E
5E
6S
0.08
0.15
0.1
4.1 × 105
6.5 × 104
1.4 × 106
6E
7E
0.1
0.1
3.6 × 107
5.4 × 106
3.6 × 106
5.4 × 105
~5 × 105
~4 × 105
aThe parameters are obtained from fitting the data according to eq. [7]; see text for details and a discussion of errors.
bEquilibrium constant for complex formation between radical cation and water, kC/k–C
.
cUnimolecular rate constant for nucleophile-assisted cleavage of radical cation in water complex.
dThe effective bimolecular rate constant for water-assisted fragmentation of the radical cation.
eFirst-order rate constant for decay of the radical cation in acetonitrile.
kDCA are of similar magnitude, a double exponential analysis
of the data can be used to estimate both rate constants, as il-
lustrated in Fig. 1, below.
is clearly around 530 nm (Fig. 2 and ref. 6), and wave-
lengths in this region were chosen for direct observation of
these species.
Shown in Fig. 3 is a transient absorption decay at 530 nm
for excitation of DCA–biphenyl in the presence of amino-
silane 6E. A fast-absorbance decay caused by reaction of
DCA•– with oxygen is observed, followed by a slower decay
as a result of fragmentation of the silyl radical cation, 6E•+.
The total absorbance, A, is fitted to an equation of the form
shown in eq. [4], where t is time and A1 and A2 represent the
In the present study, we have used all three techniques as
kfr varied as a function of the concentration of the nucleo-
phile (usually water, see below). The absorbance as a func-
tion of time, in each case, is analyzed according to a specific
type of double exponential analysis. Shown in Fig. 2 are rep-
resentative transient absorption spectra observed upon exci-
tation of a solution of DCA–biphenyl in the presence of the
aminosilanes, in this case silane 4S. According to Scheme 2,
after oxidation of the silane by the biphenyl radical cation
and before its fragmentation, equal concentrations of the sil-
ane radical cation and the DCA radical anion are present.
From the previously determined extinction coefficient of the
DCA radical anion (8), the extinction coefficient of the sil-
ane radical cations can thus be obtained by comparing the
signal sizes due to the radical anion and radical cation. The
optimum wavelength for observing the silane radical cations
1
2
[4]
A = A e−k t + A2e−k t + A3
1
absorbances owing to DCA•– and 6E•+, respectively, and A3
represents small amounts of residual absorbance at infinite
time. In this case, k1 corresponds to kO and k2 corresponds
2
to kfr.
Figure 4 shows the formation and decay of the radical, X•,
formed from fragmentation of the radical cation of the 4-
formyl-substituted silylaniline 7E. The absorbance is fitted
© 2003 NRC Canada

Gould et al.
781
Fig. 2. Transient absorption spectra of (open circles) the radical
Fig. 1. Optical density (in units of 10–3, mOD) at 705 nm, as a
function of time, t, observed upon excitation of DCA–biphenyl
and 4S in argon-purged acetonitrile containing 9.3 M water,
showing two-electron reduction of DCA. The absorbance at this
wavelength is mainly due to the radical anion DCA•–. The jump
in optical density at zero time initial is owing to absorbance by
4S•+ and DCA•–. The growth in absorbance is owing to reduction
of a second DCA after fragmentation of the 4S•+. The change in
absorbance with time is described by a consecutive a → b → c
process in which c is monitored. The smooth curve represents
the best fit to the data according to 15.3 – 12.2 exp(–6.1 ×
105 t) + 5.0 exp(–1.5 × 106 t). The slower component corre-
sponds to the 4S•+ reaction, kfr under the present conditions, and
the faster component to reduction of the second DCA.
anion of 9,10-dicyanoanthracene (DCA•–) in acetonitrile solvent
and (closed circles) the radical cation of a representative amino-
silane 4S (4S•+) in acetonitrile in the presence of 20% water,
showing overlap of the spectra. The spectrum of 4S•+ in aceto-
nitrile in the presence of less water is somewhat broader. The
figure indicates that the optimum wavelength for direct observa-
tion of the aminosilane radical cations is around 530 nm.
rate becoming less dependent at higher water concentrations.
The kinetics of reactions involving high concentrations of
hydroxylic quenchers, such as water, often exhibit nonlinear
behavior that has been subject to a variety of explanations
(6, 10). For example, with increasing water concentration,
the character of the solvent gradually changes from polar
aprotic to polar protic, which would be expected to decrease
the nucleophilicity of the water (11). However, the radical
cations are strong Lewis acids, and an alternate explanation
of the behavior, in this case, may be due to reversible forma-
tion of a complex. Previously, Mariano and co-workers sug-
gested a similar explanation for the nonlinear dependence of
this reaction on water concentration, in terms of a covalent
complex between the oxygen of the nucleophile and the sili-
con on the radical cation (6). Such a complex, whatever its
structure, can be characterized by a second-order rate con-
stant for formation, kC, and a first-order rate constant for
separation, k–C, Scheme 3. The ratio of these rate constants,
kC/k–C, defines an overall equilibrium constant, Keq.
Nucleophile-assisted fragmentation can occur within the
complex, with a first-order rate constant, kNu, Scheme 3. In
this scheme, k0 is the pseudo-first-order decay rate constant
of the radical cation in the absence of added nucleophile.
Based on Scheme 3, the decay of the radical cation and the
growth of the radical X• are properly described by two expo-
nential components. However, when k–C is much larger than
kNu, the time constant for one of these components is much
larger than the other, and this fast component also has a
small amplitude. Under these conditions, only the larger and
slower component is actually observed. The time constant
to an equation of the form shown in eq. [5], where A1 repre-
sents the absorbance of X• at the concentration of the X–
SiR3•+ precursor extrapolated to zero time and A2 represents
1
2
[5]
A = (A1 k1)/(k2 – k1) (e−k t – e−k t) + A2
the residual absorbance at infinite time. In this case, the
“normal” situation applies, i.e., k1 corresponds to kfr, and k2
corresponds to kDCA. Figure 1 shows the growth in absorp-
tion caused by the formation of the second DCA•–, as a re-
sult of reaction of X• with DCA, for excitation of DCA–
biphenyl with silane 4S. The absorbance is fitted according
to eq. [6], where A1 represents the absorbance of the second
DCA•– at infinite time, A2 represents the absorbance of the
first DCA•– that appears at essentially zero time, and k1 cor-
responds to kfr and k2 to kDCA
.
1
2
[6]
A = A1[1 + (k2e−k t – k1e−k t)/(k2 – k1)] + A2
Fragmentation rate constants
Experiments were performed in acetonitrile at a varying
concentrations of water because, in a photographic disper-
sion, water is the likely nucleophile to assist the fragmenta-
tion process. In all cases, the rate of fragmentation of the
radical cations increased with increasing water concentration
(6, 7). Typical data are shown in Fig. 5 for the radical cat-
ions of silanes 6E and 4E.
As observed previously (6), the dependence on added wa-
ter as a nucleophile is somewhat nonlinear, with the reaction
© 2003 NRC Canada

782
Can. J. Chem. Vol. 81, 2003
Scheme 3.
Fig. 4. Optical density (in units of 10–3, mOD) at 740 nm, as a
function of time, t, observed upon excitation of DCA–biphenyl
and 7E in argon-purged acetonitrile containing 2.1 M water. The
absorbance at this wavelength is mainly owing to the radical re-
sulting from fragmentation of the radical cation, 7E•+. The absor-
bance is described by a consecutive a → b → c process in which
the intermediate b is monitored. The smooth curve represents the
best fit to the data and is given by: –21.05 exp(–2.1 × 106 t) +
19.1 exp(–3.4 × 105 t) + 3.9.4 The negative component corre-
sponds to radical growth, kfr under the present conditions, and
the positive component to radical decay caused by reduction of
DCA. The spike at early time is owing to the biphenyl (cosen-
sitizer) radical cation (see text), which reacts very rapidly with 7E.
Fig. 3. Optical density (in units of 10–3, mOD) at 530 nm, as a
function of time, t, observed upon excitation of DCA–biphenyl
and 6E in oxygen-purged acetonitrile containing 1.1 M water.
Both the radical anion of DCA (DCA•–) and the radical cation
6E•+ contribute to the total absorbance, and the decay is thus a
sum of two independent pseudo-first-order exponentials. The
smooth curve represents the best fit to the data and is given by:
5.1 exp(–8.1 × 106 t) + 32.3 exp(–4.0 × 106 t) + 0.6. The faster
decay corresponds to reaction of DCA•– with oxygen and the
slower decay to reaction of 6E•+ with water, i.e., kfr under the
present conditions.
for this component is given by eq. [7] and is equated to kfr.2
Using eq. [7], values for Keq and kNu that reproduce the ob-
served dependence of kfr on water concentration can be ob-
tained.3
ion (Table 1). Although Table 1 summarizes the best values
for the parameters, we find that reasonable fits to the data
can actually be obtained using both slightly higher and
lower values for Keq (by ca. 30%) than those given in Ta-
ble 1, with corresponding changes in kNu in the opposite di-
rection. Although there is some uncertainty associated with
absolute values of Keq and kNu, the product KeqkNu is deter-
mined quite accurately and represents a good measure of the
relative reactivities of the different compounds, Table 1.
KeqkNu has dimensions of M–1 s–1, is equivalent to the initial
slope of the observed rate constant vs. water concentration,
[7]
kfr = 0.5 {kC [H2O] + k0 + k–C + kNu
– 0.5 {(kC [H2O] + k0 – k–C – kNu
+ 4kC [H2O]k–C)}0.5
}
2
)
The data are summarized in Table 1. The values for Keq
are roughly the same for all of the reactions studied here and
are small, ca. 0.1 M–1. In contrast to Keq, however, kNu de-
pends quite strongly upon the substituents on the radical cat-
2 Equation [7] gives the time constant of the slower observed component as a difference between two functions. The time constant of the
faster component is given by the sum of the same two functions, which is too large (>108 s–1) to be resolved under the experimental condi-
tions.
3 In principle, fitting of the dependence of kfr on [H2O] using eq. [7] allows determination of values for both kC and k–C. However, the kfr cal-
culated using eq. [7] are insensitive to the absolute values of kC and k–C when kC is ~108 M–1 s–1 or larger.
4 The form of this equation is not identical to that of eq. [5] due to a small absorption at this wavelength of DCA•–, the kinetics of which are
superimposed and are described by eq. [6].
© 2003 NRC Canada

Gould et al.
783
Fig. 6. Observed rate constants for pseudo-first-order decay of
Fig. 5. Observed rate constants for pseudo-first-order decay of
the radical cations 6E•+ and 4E•+, kfr, as a function of water con-
centration, in acetonitrile at room temperature, illustrating the ef-
fect of the oxidation potential of the precursor. The curves are
calculated as described in the text, with values for effective bi-
molecular rate constants, KeqkNu, of 3.6 × 106 M–1 s–1 for 6E•+
and 4.1 × 105 M–1 s–1 for 4E•+.
the radical cations 6E•+ and 7E•+, kfr, as a function of water con-
centration, in acetonitrile at room temperature, illustrating the ef-
fect of steric hindrance at silicon. The curves are calculated as
described in the text, with values for effective bimolecular rate
constants, KeqkNu, of 3.6 × 106 M–1 s–1 for 6E•+ and 5.4 ×
105 M–1 s–1 for 7E•+.
and represents the “effective” bimolecular rate constant.
This value can thus be compared to the bimolecular rate
constants reported for nucleophile-assisted fragmentation of
silane radical cations reported elsewhere (6, 7). Importantly,
the use of KeqkNu allows for the nonlinear dependence on
water concentration to be properly accounted for.
that the positive charge is likely to be more localized on the
nitrogen in the aminosilane radical cations, i.e., close to the
site of nucleophilic attack. In the anisyl silane, 8, the posi-
tive charge may be located more on the oxygen, i.e., away
from the site of nucleophilic attack.
The effect of a 4-formyl substituent is to significantly en-
hance the reactivity toward water as a nucleophile. This is
expected based upon previous work on silane and other radi-
cal cations (1, 2, 6, 7) and is easily explained using the ther-
modynamic cycle approach to understanding radical cation
reactivity, pioneered by Arnold and co-workers (12). Using
such cycles, it can easily be shown that when all other fac-
tors are equal, the thermodynamics of a radical cation reac-
tion should become more favorable with increasing
oxidation potential of the precursor, leading to a faster reac-
tion (12). In the present case, the electron-withdrawing
formyl group increases the oxidation potential of the
aminosilane and faster reaction results, as observed in the
comparison between 6E and 4E in Fig. 5. The data of Ta-
ble 1 (see also 6S vs. 4S and 7E vs. 5E) indicate that the ef-
fect of the 4-formyl substituent is to increase KeqkNu by
approximately one order of magnitude, as compared with the
corresponding unsubstituted compound. Comparable in-
creases in rate constant for cleavage using methanol as the
nucleophile were observed previously by Mariano and co-
workers for similar silane radical cations with electron-
withdrawing groups on nitrogen (6).
In addition to oxidation potential, the silanes offer a
means to control the rate of fragmentation via steric effects,
as illustrated in Fig. 6, where the dependence on water con-
centration is shown for trimethylsilyl 6E and triethylsilyl
7E. The KeqkNu for the triethylsilanes are smaller than those
of the trimethyl analogues by a factor of ca. 6 1 for both
the unsubstituted and the 4-formyl anilines (Table 1, see 5E
vs. 4E and 7E vs. 6E). Steric effects of similar magnitude
were observed in corresponding reactions of the radical cat-
ions of benzyltrialkylsilanes (7). Interestingly, a comparison
of the tert-butyl esters and the corresponding carboxylate
salts shows that the carboxylates react with water more
slowly by a factor of ca. 3 (Fig. 7 and Table 1, see 4S vs. 4E
and 6S vs. 6E). A definitive explanation of this effect cannot
be given, but it may be attributed to hydration of the car-
boxylate group, which will decrease the nucleophilicity of
the water in the microenvironment of the C—Si bond (11).
The literature experiment that most closely relates to the
present work is the water-assisted radical cation fragmenta-
The aminosilanes are considerably easier to oxidize than
the aryltrialkylsilanes studied previously (7), yet the reactivi-
ties of their radical cations towards water as a nucleophile
are almost as high. This may be a consequence of the fact
© 2003 NRC Canada

784
Can. J. Chem. Vol. 81, 2003
Fig. 7. Observed rate constants for pseudo-first-order decay of
the radical cations 6E•+ and 6S•+ as a function of water concen-
tration, in acetonitrile at room temperature, illustrating the effect
of a proximal carboxylate group. The curves are calculated as
described in the text, with values for effective bimolecular rate
constants, KeqkNu, of 3.6 × 106 M–1 s–1 for 6E•+ and 1.4 ×
106 M–1 s–1 for 6S•+.
tion reaction shown in eq. [8], studied by Mariano and co-
workers (6c). As mentioned above, a nonlinear dependence
on water concentration was also observed in this previous
work, and a rate constant was estimated at low water con-
centrations to be 1.3 × 106.
[8]
This value is somewhat higher than the (KeqkNu) for either
of the p-H-substituted trimethylsilyl-radical cations included
in the present study (Table 1). The reasons for this small ef-
fect is not exactly clear. It may further illustrate the subtle
effects that substituents on nitrogen can have on the
micropolarity and, thus, the nucleophilicity of the water. Al-
ternatively, it may reflect the somewhat different methods of
kinetic analysis used in the two studies.
The nature of the decay process of the radical cation in
the absence of added water, k0, has not been established.
These values are either measured directly in the absence of
water or are obtained as part of the fitting procedure de-
scribed below. Pseudo-first-order rate constants less than ca.
106 s–1 are often difficult to measure accurately in experi-
ments such as these because of competing second-order re-
combination reactions. They are also the least accurately
determined parameters obtained from the fitting procedure.
Nevertheless, it is clear that the decay constants of the
unsubstituted, aminosilane radical cations 4E and 5E, in the
absence of water, are in the range of ~5 × 104 s–1, whereas
those of the 4-formyl analogues 6E and 7E are nearly an or-
der of magnitude larger, ~4–5 × 105 s–1. The higher reactiv-
ity of the 4-formyl derivatives, relative to the unsubstituted
analogues, is understandable based on the thermodynamic
arguments given above. It is not possible to confirm frag-
mentation by the observation of two-electron reduction un-
der these conditions, because the kfr would be too small. It is
possible that the decays under these conditions may be due
to deprotonation (6) or perhaps to reaction with impurities.
were performed using
a nanosecond pulsed excimer
(Questek model 2620, 308 nm, ca. 20 ns, ca. 100 mJ)
pumped dye laser (Lambda Physik model FL 3002). The la-
ser dye was DPS in dioxane (410 nm, ca. 20 ns, ca. 10 mJ).
The analyzing light source was a pulsed 150W xenon arc
lamp (Osram XBO 150/W). The arc lamp power supply was
a PRA model 302, and the pulser was a PRA model M-306.
The pulser increased the light output by ca. 100-fold for a
time period of ca. 2–3 ms. The analyzing light was focused
through a small aperture (ca. 1.5 mm) in a cell holder de-
signed to hold 1 cm2 cuvettes. The laser and analyzing
beams irradiated the cell from opposite directions and
crossed at a narrow angle (ca. 15°). After leaving the cell,
the analyzing light was collimated and focused onto the slit
(1 mm, 4 nm bandpass) of an ISA H-20 monochromator.
The light was detected using a Hamamatsu model R446
photomultiplier. The output of the photomultiplier tube was
terminated into 50 ohm and captured using a Tektronix
DSA-602 digital oscilloscope. The experiments were per-
formed in 1 cm2 cuvettes, equipped with arms for degassing.
The DCA and biphenyl concentrations in all of the time-
resolved experiments were ca. 10–4 and ca. 0.1 M, respec-
tively. The concentrations of the aminosilanes was typically
ca. 2–4 × 10–3 M.
Conclusions
The properties of a series of aminosilanes have been in-
vestigated as potential two-electron donors for silver halide
photography. Solution-phase studies have been performed
that confirm the radical cations of the donors fragment to
give appropriately reducing radicals in the presence of an
appropriate nucleophile. By varying the steric bulk of the
substituents on the silicon, the rate of fragmentation of the
radical cations can be varied substantially with minimal
changes in the oxidation potential of the donor. Pseudo-first-
order rate constants for fragmentation varying over ca. 2 or-
ders of magnitude have been measured by varying the silane
oxidation potential, the size of the substituents at silicon,
and the concentration of water as an external nucleophile.
NMR spectra were run on a Varian 300 MHz spectrome-
ter, generally at a concentration of ca. 100 mg mL–1. For se-
lected examples, additional NMR data obtained from
measurements at concentrations of 10–20 mg mL–1, includ-
ing coupling constants, are given in Table 2. Preparation of
the materials was performed as follows:
Experimental
The solvents were spectrograde and used as received.
9,10-Dicyanoanthracene and biphenyl were available from
previous studies (3). The laser flash photolysis experiments
© 2003 NRC Canada

Gould et al.
785
Table 2. Detailed NMR data of representative compounds.
R
Me
H
Et
Me
Me
R′
H
CHO
CHO
–
–
R′ ′
CO2
CO2-t-Bu
5E (CDCl3)
CO2
CO2-t-Bu
6E (CDCl3)
Compound (solvent)
4S (D2O)
6S (CD3OD)
3.49
3.38
2.41
2.31
3.34
1.14
6.83
7.20
6.65
—
3.63
3.47
2.59
2.41
3.51
1.24
6.73
7.20
6.65
—
3.75
3.57
2.57
2.37
3.62
1.36
6.84
7.67
—
9.52
—
0.09
—
3.73
3.54
2.62
2.43
3.49
1.31
6.72
7.69
—
9.7
1.46
0.1
—
—
Ha
Hb
Hc
Hd
He
Hf
o-H
m-H
p-H
CHO
t-Bu
Si(CH3)3
Si(CH2CH3)3
Si(CH2CH3)3
J values
*Jab*
—
–0.01
—
1.47
—
0.65
0.97
—
—
15.0
5.2
15.5
5.2
14.8
4.8
15.1
5.2
Jac
Jad
Jbc
Jbd
10.2
9.7
5.8
10.4
9.7
5.8
11.0
11.1
5.5
10.5
10.2
5.8
*Jcd*
Jef
14.9
7.4
14.9
7.4
14.8
7.3
15.7
7.4
anhydride (23.9 g, 84.6 mmol) was added over 2 min. A
thick slurry resulted. 1-Trimethylsilylethanol (10.0 g,
84.6 mmol) was added and the mixture stirred for 1 h while
the temperature was allowed to rise to 25°C. The resulting
pyridine trifluoromethanesulfonate was removed by filtra-
tion. The filtrate was concentrated in vacuo at 25°C to give
18.2 g (86 %) of the sulfonate.
tert-Butyl 3-(N-phenylamino)propionate
Aniline (126.7 g, 1.36 mol), tert-butyl acrylate (174.5 g,
1.36 mol), and methanesulfonic acid catalyst (4 mL) were
combined and stirred at gentle reflux (135°C) for 4 h. The
mixture was cooled and petroleum ether (1 L) was added.
The resulting precipitated salt was removed by filtration and
discarded. The filtrate was concentrated in vacuo at 90°C to
an oil (232 g). The desired pure ester was obtained by vac-
uum distillation (138 g, 46%, 115–125°C at 0.3–0.5 mmHg
(1 mmHg = 133.322 Pa)).
1H NMR (300 MHz, CDCl3) δ: 1.45 (s, 9H, -C(CH3)3),
2.50 (t, 2H, -CH2CH2CO-), 3.40 (t, 2H, -NHCH2CH2CO-),
4.20 (s, 1H, -NH-), 6.60 (d, 2H, o-H), 6.70 (t, 1H, p-H), 7.30
(t, 2H, m-H).
13C NMR (75 MHz, CDCl3) δ: 8.11 (-CH3), 35.16
(-CH2CH2CO-), 39.69 (-NHCH2CH2-), 80.82 (-O-C(CH3)3),
113.06 (2,6-aromatic C), 117.60 (4-aromatic C), 129.25
(3,5-aromatic C), 147.80 (1-aromatic C), 171.69 (-COO-).
1H NMR (300 MHz, CDCl3) δ: 0.12 (s, 9H, -Si(CH3)3),
1.55 (d, 3H, CH3CH-), 4.92 (q, 1H, -SiCHOTfCH3).
13C NMR (75 MHz, CDCl3) δ: –4.47 (-Si(CH3)3), 17.38
(CH3CH-), 85.36 (CH3CH-), 118.0 (q, -SO2CF3).
1-Triethylsilylethanol
Borane–tetrahydrofuran complex (1.0 M solution in tetra-
hydrofuran, 290 mL, 0.29 mol) was cooled to 0°C. Triethyl-
vinylsilane (41.2 g, 0.29 mol) was added over 10 min,
keeping the reaction temperature between 0 and 10°C. The
mixture was stirred for 1 h, and the temperature was allowed
to rise to 25°C. Water (10 mL) was cautiously added
dropwise, followed by 3 M aqueous sodium hydroxide
(97 mL, 0.29 mol) over 10 min. Hydrogen peroxide (30%,
99 mL, 0.96 mol) was added over 10 min, keeping the reac-
1-Trimethylsilylethyl trifluoromethanesulfonate
Pyridine (6.7 g, 84.6 mmol) and dichloromethane
(200 mL) were cooled to –25°C. Trifluoromethanesulfonic
© 2003 NRC Canada

786
Can. J. Chem. Vol. 81, 2003
tion temperature at ca. 30°C. The mixture was refluxed for
1 h. Cooling to 25°C gave three liquid phases. The addition
of anhydrous potassium carbonate (150 g, 1.09 mol) pro-
duced two clear liquid phases. The top organic phase was
dried over magnesium sulfate and concentrated in vacuo at
25°C to give a ca. 1:1 mixture (53.6 g) of the desired 1-
triethylsilylethanol and the unwanted isomer 2-triethylsilyl-
ethanol. The crude mixture was distilled, and the fraction
boiling at 90–100°C, 12 mmHg, was collected (26.8 g). This
material was distilled again, and the fraction boiling at 87–
90°C, 12 mmHg, was the desired pure isomer (10.0 g, 22%).
1H (300 MHz, CDCl3) δ: 0.60 (q, 6H, -Si(CH2CH3)3),
0.90 (t, 9H, -Si(CH2CH3)3), 1.25 (s, 1H, -OH), 1.30 (d, 3H,
CH3CH-), 3.62 (q, 1H, CH3CH-).
2H, -NCH2CH2-), 6.65 (t, 1H, p-CH), 6.75 (d, 2H, o-CH),
7.20 (m, 2H, m-CH).
13C NMR (75 MHz, CDCl3) δ: –1.89 (-Si(CH3)3), 13.61
(CH3CH-), 28.13 (-C(CH3), 34.87 (-CH2CH2CO-), 44.47
(-NCH2CH2-), 47.05 (CH3CH-), 80.50 (-C(CH3)3), 112.89
(o-CH), 115.98 (p-CH), 129.12 (m-CH), 148.63 (aromatic
C-N), 171.47 (-CO-).
Compound 4S
The tert-butyl ester 4E (3.70 g, 11.5 mmol) and trifluoro-
acetic acid (10 mL, ca. 130 mmol) were stirred at 25°C for
16 h. The reaction mixture was concentrated in vacuo at
25°C. The oil was dissolved in ethyl ether (100 mL) and
washed with water (2 × 25 mL). The ether layer was washed
with aqueous sodium bicarbonate (3.5 g, 42 mmol, 25 mL
water). The bottom aqueous layer (pH ca. 7) was discarded
and the top ether layer treated with anhydrous sodium bicar-
bonate – magnesium sulfate. The dried organic extract was
concentrated in vacuo at 25°C to give the free acid as a
golden oil (2.7 g, 88%). The acid (2.50 g, 9.4 mmol) was
stirred with a mixture of diethyl ether (50 mL), water
(50 mL), and sodium hydroxide (0.37 g, 9.3 mmol). The top
ether layer was discarded. The bottom aqueous layer
(pH 10) was washed with fresh ether and the ether layer
again discarded. The bottom aqueous layer was treated with
decolorizing carbon and silica gel, resulting in a pH ca. 9. A
few drops of acetic acid were added to bring the pH to ca.
7–8. The solution was concentrated in vacuo at 25°C to give
a brown oil (2 g, 74%). The oil was dissolved in acetonitrile
and treated with anhydrous sodium sulfate and decolorizing
carbon. The solvent was removed at 25°C to give 4S as a hy-
groscopic gum.
13C (75 MHz, CDCl3) δ: 1.46 (-Si(CH2CH3)3), 7.42
(-Si(CH2CH3)3), 20.21 (-CHCH3), 60.07 (-CHCH3).
1-Triethylsilylethyl trifluoromethanesulfonate
A mixture of pyridine (4.9 g, 62 mmol) and dichloro-
methane (150 mL) was cooled to –25°C. Trifluoromethane-
sulfonic acid anhydride (17.5 g, 62 mmol) in
dichloromethane (25 mL) was added, resulting in a thick
slurry. A solution of 1-triethylsilylethanol (9.96 g, 62 mmol)
and dichloromethane (25 mL) was added and the reaction
mixture stirred for 1 h while the temperature was allowed to
rise to 25°C. Most of the solid had dissolved. Additional
pyridine (ca. 1.5 mL) was added to raise the pH of the reac-
tion mixture from ca. 3 to 5, resulting again in a thick slurry.
The salt (14.7 g) was removed by filtration and the filtrate
concentrated in vacuo at 25°C to a semi-solid (20 g). Petro-
leum ether (200 mL) was added, and more resulting salt was
removed by filtration. The filtrate was treated with
decolorizing carbon and concentrated in vacuo at 25°C to an
1H NMR (300 MHz, D2O) δ: 0.00 (s, 9H, -Si(CH3)3), 1.23
(bs, 3H, CH3CH-), 2.20–2.60 (bm, 2H, -NCH2CH2CO-),
3.30 (m, 1H, CH3CH-), 3.40–3.80 (bd, 2H, -NCH2CH2-),
4.80 (s, HOD), 6.65 (bm, 1H, p-CH), 6.77 (bs, 2H, o-CH),
7.07 (bm, 2H, m-CH). See also Table 2.
1
oil (13.12 g). The oil was determined by H NMR to be a
70:30 mixture of the desired triflate – starting 1-triethyl-
silylethanol and was used without further purification.
1H NMR (300 MHz, CDCl3) δ: 0.50–0.72 (m, -Si(CH2CH3)3),
0.90–1.05 (m, -Si(CH2CH3)3), 1.30 (d, -CH(OH)CH3),
1.60 (d, -CH(OTf)CH3), 3.65 (q, -CH(OH)CH3), 5.08 (q,
-CH(OTf)CH3).
13C NMR (75 MHz, D2O) δ: 0.00 (-Si(CH3)3), 15.12
(CH3CH-), 36.84 (-CH2CH2CO-), 50.23 (-NCH2CH2-),
52.11 (CH3CH-), 118.69 (o-CH), 122.55 (p-CH), 132.40 (m-
CH), 148.80 (aromatic C-N), 182.38 (-COO-).
13C NMR (75 MHz, CDCl3) δ: 1.17 (-CHOTfSi(CH2CH3)3),
1.47 (-CHOHSi(CH2CH3)3), 6.92 (-CHOTfSi(CH2CH3)3),
7.42 (-CHOHSi(CH2CH3)3), 17.96 (-CHOTfCH3), 20.17
(-CHOHCH3), 60.07 (-CHOHCH3), 83.97 (-CHOTfCH3),
118.0 (q, -SO2CF3).
Compound 5E
tert-Butyl 3-(N-phenylamino)propionate (9.96 g, 45 mmol),
diisopropylethylamine (5.80 g, 45 mmol), and acetonitrile
(25 mL) were combined. 1-Triethylsilylethyl triflate
(13.12 g, 45 mmol, ca. 70% pure, the remainder being the
parent alcohol) was added, followed by additional aceto-
nitrile (25 mL). The mixture was stirred for 2 h at 25°C,
then at reflux for 3 h. The mixture was concentrated in
vacuo at 40°C to give an oil. Petroleum ether (100 mL) was
added and the resulting salt filtered and discarded. The fil-
trate was concentrated in vacuo at 40°C to give an oil
(13.5 g). Flash chromatography (silica gel, 9 parts petroleum
ether : 1 part ethyl acetate) gave the crude product, 5E
(4.6 g) with trace silyl impurities. A second flash chromato-
graphic operation gave the pure desired product (3.0 g, 18%).
1H NMR (300 MHz, CDCl3) δ: 0.65 (q, 6H, -Si(CH2CH3)3),
1.00 (t, 9H, -Si(CH2CH3)3), 1.28 (d, 3H, -CHCH3), 1.50 (s,
9H, -C(CH3)3), 2.40–2.50 and 2.55–2.70 (m, m, 1H, 1H,
Compound 4E
A mixture of 1-trimethysilylethyl trifluoromethanesulfo-
nate (6.38 g, 25.5 mmol), tert-butyl 3-anilinopropionate
(5.64 g, 25.5 mmol), potassium carbonate (3.52 g, 25.5 mmol),
and butyronitrile (25 mL) was stirred at reflux for 16 h. The
mixture was filtered to remove salt, and the filtrate was con-
centrated in vacuo at 50°C to give 6.9 g of a yellow oil. The
desired product was the first eluting material (3.7 g, 45%)
obtained via flash chromatography (silica gel, 9 parts petro-
leum ether : 1 part ethyl acetate).
1H NMR (300 MHz, CDCl3) δ: 0.12 (s, 9H, -Si(CH3)3),
1.25 (d, 3H, CH3CH-), 1.48 (s, 9H, -C(CH3)3), 2.40–2.65
(m, 2H, -CH2CH2CO-), 3.35 (q, 1H, CH3CH-), 3.45–3.7 (m,
© 2003 NRC Canada

Gould et al.
787
-CH2CH2CO-), 3.45–3.70 (m, 3H, -NCH2CH2- and -NCHCH3),
6.65 (t, 1H, p-H), 6.75 (d, 2H, o-H), 7.20 (m, 2H, m-H). See
also Table 2.
13C NMR (75 MHz, CDCl3) δ: 3.75 (-Si(CH2CH3)3), 7.62
(-Si(CH2CH3)3), 12.79 (-CHCH3), 28.13 (-C(CH3)3), 34.86
(-CH2CH2CO-), 44.85 (-NCH2CH2-), 45.47 (-CHCH3),
80.49 (-O-C(CH3)3), 113.63 (o-CH), 115.98 (p-CH), 129.15
(m-CH), 148.42 (aromatic C-N), 171.52 (-CO-).
13C NMR (75 MHz, D2O) δ: 0.02 (-Si(CH3)3), 16.48
(CH3CH-), 39.02 (-CH2CH2CO-), 48.62 (-NCH2CH2-),
49.76 (CH3CH-), 114.59 (o-CH), 125.37 (p-C-CHO), 135.93
(m-CH), 157.02 (aromatic C-N), 183.03 (-COO-), 195.47
(-CHO).
Compound 7E
DMF (40 mL) was cooled to –10°C. Phosphorous
oxychloride (2.11 g, 13.7 mmol) was added, followed by the
tert-butyl ester 5E. The mixture was allowed to warm to
25°C and stirred for 16 h. The yellow solution was added to
a solution of sodium acetate (11.2 g, 137 mmol) and water
(150 mL) and the resulting mixture stirred for 1 h. The prod-
uct was extracted with petroleum ether (3 × 50 mL), the ex-
tracts treated with magnesium sulfate and decolorizing
carbon and concentrated in vacuo at 35°C to an oil (2.9 g).
Flash chromatography (silica gel: 9 parts petroleum ether : 1
part ethyl acetate) gave the purified 4-formyl tert-butyl ester,
7E (2.3 g, 85%).
Compound 6E
Dimethylformamide (DMF, 50 mL) was chilled to –10°C.
Phosphorous oxychloride (3.20 g, 21 mmol) was added. A
solution of tert-butyl 3-(N-phenylamino)propionate (3.36 g,
10.4 mmol) and DMF (15 mL) was added and the mixture
stirred at 25°C for 16 h. The reaction mixture was added to a
solution of sodium acetate (17.22 g, 220 mmol) and water
(230 mL) and allowed to stir for 2 h. The product was ex-
tracted with petroleum ether (3 × 50 mL). The petroleum
ether extracts were dried over magnesium sulfate and treated
with carbon. The treated extracts were concentrated in vacuo
at 25°C to give 3.4 g of oil. The pure 6E, 3.04 g (84%), was
obtained via flash chromatography (silica gel, 9 parts petro-
leum ether : 1 part ethyl acetate).
1H NMR (300 MHz, CDCl3) δ: 0.10 (s, 9H, -Si(CH3)3),
1.32 (d, 3H, CH3CH), 1.45 (s, 9H, -C(CH3)3), 2.40–2.70 (m,
2H, -NCH2CH2CO-), 3.50 (q, 1H, CH3CH-), 3.55–3.80 (m,
2H, -NCH2CH2-), 6.70 (d, 2H, o-CH), 7.70 (d, 2H, m-CH),
9.70 (s, 1H, -CHO). See also Table 2.
1H NMR (300 MHz, CDCl3) δ: 0.65 (q, 6H, -Si(CH2CH3)3),
0.90 (t, 9H, -Si(CH2CH3)3), 1.30 (d, 3H, -CHCH3), 1.45 (s,
9H, -C(CH3)3), 2.35–2.50 and 2.60–2.75 (m, m, 1H, 1H,
-CH2CH2CO-), 3.45–3.60, 3.65, and 3.70–3.80 (m, q, m, 1H,
1H, 1H, -NCH2CH2-, -CHCH3), 6.73 (d, 2H, o-H), 6.80 (d,
2H, m-H), 9.70 (s, 1H, -CHO).
13C NMR (75 MHz, CDCl3) δ: 3.25 (-Si(CH2CH3)3), 7.47
(-Si(CH2CH3)3), 15.12 (-CHCH3), 28.06 (-C(CH3)3), 34.45
(-CH2CH2CO-), 45.13 (DEPT confirmed, -NCH2CH2- and
-NCHCH3), 81.07 (-O-C(CH3)3), 111.61 (o-CH), 125.06
(p-C-CHO), 131.80 (m-CH), 152.64 (aromatic C-N), 170.68
(-COO-), 189.72 (-CHO).
13C NMR (75 MHz, CDCl3) δ: –2.03 (-Si(CH3)3), 14.12
(CH3CH-), 28.03 (-C(CH3)3), 34.40 (-CH2CH2CO-), 44.68
(-NCH2CH2-), 47.10 (CH3CH-), 81.04 (-C(CH3)3), 111.61
(o-CH), 125.01 (p-C-CHO), 132.01 (m-CH), 152.84 (aro-
matic C-N), 170.63 (-COO-), 189.69 (-CHO).
Acknowledgments
The authors would like to thank Jerome Lenhard and
Annabel Muenter (Eastman Kodak Company) for valuable
discussions.
Compound 6S
The 4-formyl tert-butyl ester 6E (3.04 g, 8.7 mmol) was
cooled to –25°C, trifluoroacetic acid (10 mL, ca. 130 mmol)
was added, and the solution was stirred at 25°C for 16 h.
The reaction mixture was concentrated in vacuo at 25°C to a
dark oil (9.23 g). The oil was partitioned between ethyl ether
(100 mL) and aqueous sodium bicarbonate (4.92 g,
58.6 mmol, 100 mL water). The bottom aqueous layer had a
pH ca. 7–8. The top organic layer was treated with magne-
sium sulfate and decolorizing carbon and concentrated in
vacuo at 25°C to an oil (1.9 g). Pure free carboxylic acid
(1.46 g, 57%) was obtained via flash chromatography (silica
gel: 9 parts dichloromethane : 1 part methanol). The free
acid (1.46 g, ca. 5 mmol) was combined with methanolic so-
dium methoxide (0.5 M, 9.5 mL, 4.75 mmol) to give a pH of
ca. 8. Acetic acid was added to lower the pH to ca. 7. The
solution was concentrated in vacuo at 25°C to give an oil
(1.96 g). The oil was triturated with acetonitrile (3 × 5 mL)
and the solvent again removed in vacuo at 25°C to give 6S
as a hygroscopic gum (0.83 g).
References
1. (a) E. Baciocchi, M. Bietti, and O. Lanzalunga. Acc. Chem.
Res. 33, 243 (2000); (b) E.R. Gaillard and D.G. Whitten. Acc.
Chem. Res. 29, 292 (1996); (c) A. Albini, E. Fasani, and M.
Freccero. Adv. Electron Transfer Chem. 5, 103 (1996); (d) J.-M.
Saveant. Adv. Electron Transfer Chem. 4, 53 (1994); (e) P.
Maslak. Top. Curr. Chem. 168, 1 (1993); (f) R. Popielarz and
D.R. Arnold. J. Am. Chem. Soc. 112, 3068 (1990); (g) S.L.
Mattes and S. Farid. Org. Photochem. 6, 233 (1983).
2. (a) D. Mangion and D.R. Arnold. Can. J. Chem. 77, 1655
(1999); (b) L.A. Eriksson, C.M. Jansson, D.R. Arnold, and
R.J. Boyd. Can. J. Chem. 76, 1817 (1998); (c) X. Du, D.R.
Arnold, R.J. Boyd, and Z. Shi. Can. J. Chem. 69, 1365 (1991);
(d) R. Popielarz and D.R. Arnold. J. Am. Chem. Soc. 112,
3068 (1990); (e) D.R. Arnold and L.J. Lamont. Can. J. Chem.
67, 2119 (1989); (f) A. Okamoto, M.S. Snow, and D.R.
Arnold. Tetrahedron, 42, 6175 (1986); (g) A. Okamoto and
D.R. Arnold. Can. J. Chem. 63, 2340 (1985).
1H NMR (300 MHz, D2O) δ: 0.12 (bs, 9H, -Si(CH3)3),
1.38 (bd, 3H, CH3CH-), 2.0 (bs, acetonitrile), 2.15 (bs, ace-
tic acid), 2.40–2.70 (bm, 2H, -CH2CH2CO-), 3.55–3.85 (bq
and bm, 3H, CH3CH-, NCH2CH2-), 6.90 (bd, 2H, o-CH),
7.75 (bd, 2H, m-CH), 9.45 (bs, 1H, -CHO). See also Table 2.
3. (a) I.R. Gould, J.R. Lenhard, A.A. Muenter, S.A. Godleski,
and S. Farid. J. Am. Chem. Soc. 122, 11 934 (2000); (b) I.R.
Gould, J.R. Lenhard, A.A. Muenter, S.A. Godleski, and S.
Farid. Pure Appl. Chem. 73, 455 (2001).
© 2003 NRC Canada

788
4. (a) S.Y. Farid, J.R. Lenhard, C.H. Chen, A.A. Muenter, I.R.
Can. J. Chem. Vol. 81, 2003
Freccero, A. Albini, D.E. Falvey, and P.S. Mariano. J. Am.
Chem. Soc. 116, 4211 (1994).
Gould, S.A. Godleski, and P.A. Zielinski. U.S. Patent
5 747 235, May 5, 1998; (b) S.Y. Farid, J.R Lenhard, C.H.
Chen, A.A. Muenter, I.R. Gould, S.A. Godleski, P.A. Zielinski,
and C.H. Weidner. U.S. Patent 5 747 236, May 5, 1998;
(c) I.R. Gould, S. Farid, S.A. Godleski, J.R. Lenhard, A.A.
Muenter, and P.A. Zielinski. U.S. Patent 5 994 051, Nov. 30,
1999; (d) S.Y. Farid, J.R. Lenhard, C.H. Chen, A.A. Muenter,
I.R. Gould, S.A. Godleski, P.A. Zielinski, and C.H. Weidner.
U.S. Patent 6 010 841, Jan. 4, 2000; (e) A. Adin, J.J. Looker,
S.Y. Farid, I.R. Gould, S.A. Godleski, J.F. Lenhard, A.A.
Muenter, L.C. Vishwakarma, and P.A. Zielinski. U.S. Patent
6 054 260, Apr. 25, 2000; (f) S.Y. Farid, I.R. Gould, S.A.
Godleski, J.R. Lenhard, A.A. Muenter, P.A. Zielinski, and
C.H. Weidner. U.S. Patent 6 153 371, Nov. 28, 2000; (g) A.
Adin, J.J. Looker, S.Y. Farid, I.R. Gould, S.A. Godleski, J.R.
Lenhard, A.A. Muenter, L.C. Vishwakarma, and P.A. Zielinski.
U.S. Patent 6 306 570, Oct. 23, 2001.
7. (a) H.J.P de Lijser, D.W. Snelgrove, J.P. Dinnocenzo. J. Am.
Chem. Soc. 123, 9698 (2001); (b) J.P. Dinnocenzo, S. Farid,
J.L. Goodman, I.R. Gould, W.P. Todd, and S.L. Mattes. J. Am.
Chem. Soc. 111, 8973 (1989); (c) J.P. Dinnocenzo, S. Farid,
J.L. Goodman, I.R. Gould, and W.P. Todd. Mol. Cryst. Liq.
Cryst. 194, 151 (1991); (d) W.P. Todd, J.P. Dinnocenzo, S.
Farid, J.L. Goodman, and I.R. Gould. J. Am. Chem. Soc. 113,
3601 (1991); (e) W.P. Todd, J.P. Dinnocenzo, S. Farid, J.L.
Goodman, and I.R. Gould. Tetrahedron Lett. 34, 2863 (1993);
(f) K.P. Dockery, J.P. Dinnocenzo, S. Farid, J.L. Goodman, I.R.
Gould, and W.P. Todd. J. Am. Chem. Soc. 119, 1876 (1997).
8. I.R. Gould, D. Ege, J.E. Moser, and S. Farid. J. Am. Chem.
Soc. 112, 4290 (1990).
9. J.W. Moore and R.G. Pearson. Kinetics and mechanism. 3rd
ed. Wiley-Interscience, New York. 1981. p. 290.
10. See, for example (a) L.M. Tolbert and J.E. Haubrich. J. Am.
Chem. Soc. (1994) 116, 10 593; (b) G.W. Robinson. J. Phys.
Chem. 95, 10 386 (1991).
5. J.R. Fyson, P.J. Twist, and I.R. Gould. Electron transfer pro-
cesses in silver halide photography. In Electron transfer in
chemistry. Vol. 5. Edited by V. Balzani. Wiley-VCH, New
York. 2001. p. 285.
11. N. Isaacs. Physical organic chemistry. 2nd ed. Longman,
England. 1987.
6. (a) Z. Su, D.E. Falvey, U.C. Yoon, and P.S. Mariano. J. Am.
Chem. Soc. 119, 5261 (1997); (b) Z. Su, P.S. Mariano, D.E.
Falvey, U.C. Yoon, and S.W. Oh. J. Am. Chem. Soc. 120,
10 676, (1998); (c) X. Zhang, S.-R. Yeh, S. Hong, M.
12. (a) R. Popielarz and D.R. Arnold. J. Am. Chem. Soc. 112,
3068 (1990); (b) P. Maslak, T.M. Vallombroso, W.H.
Chapman, Jr., and J.N. Narvaez. Angew. Chem. Int. Ed. Engl.
33, 73 (1994).
© 2003 NRC Canada