Transformation of α-tocopherol (vitamin E) and related chromanol model compounds into their phenoxonium ions by chemical oxidation with the nitrosonium cation

Article abstract of DOI:10.1021/jo0517951

α-Tocopherol (α-TOH), the main oil component making up vitamin E, and its nonnatural solid 6-hydroxy-2,2,5,7,8-pentamethylchroman and 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid structurally related analogues were oxidized quantitatively with 2 mol equiv of NO ⁺SbF₆^- in CH₃CN at 233 K to form phenoxonium cations (α-TO⁺SbF₆^-) in a chemically reversible two-electron/one-proton process. Solution-phase infrared spectroscopy, ¹H and ¹³C NMR spectroscopy, and corresponding theoretical calculations of the spectroscopic data using density-based and wave-function-based models support the identity of the remarkably stable phenoxonium cations. The presence of an oxygen atom in the para position to the hydroxyl group and the chromanol ring structure appear to be important factors in stabilization of the phenoxonium ions, which raises the interesting possibility that the cations play a crucial role in the mode of action of vitamin E in biological systems. Although the phenoxonium cations are reactive toward nucleophiles such as water, they may be moderately stable in the hydrophobic (lipophilic) environment where vitamin E is known to occur naturally.

Full text of DOI:10.1021/jo0517951

Transformation of r-Tocopherol (Vitamin E) and Related
Chromanol Model Compounds into Their Phenoxonium Ions by
Chemical Oxidation with the Nitrosonium Cation
Stephen B. Lee, Ching Yeh Lin, Peter M. W. Gill, and Richard D. Webster*
Research School of Chemistry, Australian National University, Canberra ACT 0200, Australia
webster@rsc.anu.edu.au
Received August 26, 2005
R-Tocopherol (R-TOH), the main oil component making up vitamin E, and its nonnatural solid
6-hydroxy-2,2,5,7,8-pentamethylchroman and 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic
acid structurally related analogues were oxidized quantitatively with 2 mol equiv of NO+SbF6- in
CH3CN at 233 K to form phenoxonium cations (R-TO+SbF6-) in a chemically reversible two-electron/
one-proton process. Solution-phase infrared spectroscopy, 1H and 13C NMR spectroscopy, and
corresponding theoretical calculations of the spectroscopic data using density-based and wave-
function-based models support the identity of the remarkably stable phenoxonium cations. The
presence of an oxygen atom in the para position to the hydroxyl group and the chromanol ring
structure appear to be important factors in stabilization of the phenoxonium ions, which raises
the interesting possibility that the cations play a crucial role in the mode of action of vitamin E in
biological systems. Although the phenoxonium cations are reactive toward nucleophiles such as
water, they may be moderately stable in the hydrophobic (lipophilic) environment where vitamin
E is known to occur naturally.
1. Introduction
(R-TO•). Second, the R-TO• radical reacts with another
LOO• radical so that, overall, one R-TOH molecule is able
to inhibit two LOO• sites. The localization and mobility
of R-TOH within the lipids are critical to its ability to
function, as well as are synergistic interactions with other
species such as ascorbic acid, hydroquinones, and
catechols.1d,g In addition to vitamin E sacrificially pre-
venting cell damage, there is evidence that it may
function instead, or as well, by a mechanism that is not
directly related to the inhibition of oxidation2 and has a
different function in mammals and plants.3 To date, the
important biological chemistry of vitamin E is thought
to involve only the neutral compound (R-TOH) and its
related phenoxyl radical (R-TO•). Nevertheless, voltam-
metry experiments4,5 have demonstrated the existence
R-Tocopherol (R-TOH) is the most naturally abundant,
fully methylated, and biologically active of the four (R,
â, γ, and δ) structurally related phenolic lipophilic
compounds that are labeled vitamin E. R-TOH is thought
to play an important role as an antioxidant in mam-
malian tissues by inhibiting the free radical chain au-
tooxidation of polyunsaturated fatty acid esters (LH) in
two principal steps.1 First, R-TOH reacts with an oxidized
site on a lipid cell wall (LOO•) to yield a molecule of lipid
hydroperoxide (LOOH) and the tocopheroxyl radical
* Fax: + 61 2 6125 0750.
(1) The following recent citations are representative examples of the
numerous reviews describing the mechanisms of inhibition of lipid
peroxidation by vitamin E. (a) Packer, L.; Witt, E. H.; Tritschler, H. J.
Free Radical Biol. Med. 1995, 19, 227-250. (b) Bowry, V. W.; Ingold,
K. U. Acc. Chem. Res. 1999, 32, 27-34. (c) Traber, M. G.; Arai, H.
Annu. Rev. Nutr. 1999, 19, 343-355. (d) Wang, X.; Quinn, P. J. Prog.
Lipid Res. 1999, 38, 309-336. (e) Finkel, T.; Holbrook, N. J. Nature
2000, 408, 239-247. (f) Wang, X.; Quinn, P. J. Mol. Membr. Biol. 2000,
17, 143-156. (g) Niki, E.; Noguchi, N. Acc. Chem. Res. 2004, 37, 45-
51.
(2) For example, see: (a) Azzi, A.; Stocker, A. Prog. Lipid Res. 2000,
39, 231-255. (b) Ricciarelli, R.; Zingg, J.-M.; Azzi, A. Biol. Chem. 2002,
383, 457-465. (c) Azzi, A.; Ricciarelli, R.; Zingg, J.-M. FEBS Lett. 2002,
519, 8-10.
(3) Munne-Bosch, S.; Alegre, L. Crit. Rev. Plant Sci. 2002, 21, 31-
57.
10.1021/jo0517951 CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/16/2005
10466
J. Org. Chem. 2005, 70, 10466-10473

Transformation of R-TOH into Its Phenoxonium Ion
SCHEME 1. Electrochemically Induced
In acetonitrile, at approximately neutral pH, R-TOH
can be electrochemically oxidized by one electron at solid
electrodes to form the cation radical (R-TOH+•). R-TOH+•
quickly dissociates into R-TO• (and H+), which is im-
mediately further oxidized at the electrode surface by one
electron to form the brightly orange/red- (λmax ) 452 nm;
ꢀ ) 2.5 × 103 L cm-1 mol-1)4 colored R-TO+ (Scheme 1,
pathway 2). R-TO+ is stable for at least several hours at
low temperatures and can be reduced back to R-TOH on
both the cyclic voltammetry (seconds) and the controlled
potential electrolysis (CPE) (hours) timescales,4b although
it is unstable in the presence of nucleophiles and has been
reported to react with water/moisture to form 9-hydroxy-
R-tocopherone (1), which slowly rearranges to the quinone
(2) (eq 1).8 The addition of an equimolar amount of
Transformations of r-Tocopherola,b
organo-soluble base (such as Et4NOH) to solutions of
R-TOH immediately forms the phenolate anion (R-TO-),5
which can be oxidized in two sequential one-electron
steps to also form R-TO+ (Scheme 1, pathway 3). In the
presence of ∼g1-2% CF3SO3H (or another organo-
soluble acid4a), R-TOH is oxidized in two sequential one-
electron steps to form R-TOH+• first and then, at more
positive potentials, the dication (R-TOH2+) is most likely
formed (Scheme 1, pathway 1).4
Although electrooxidation of R-TOH is a proven path-
way to the R-TO+ cation in high yield in CH3CN,4 it
requires at least a 30-fold excess of supporting electrolyte
to avoid the effects of migration, which otherwise results
in the charged species produced in the working electrode
compartment moving to the counter electrode compart-
ment and reacting irreversibly with species produced at
the auxiliary electrode.9 We were interested in testing
the stability of R-TO+ under low ionic strength conditions,
closer to those encountered in a lipophilic environment,
and in testing whether the cation could form by a
homogeneous electron-transfer mechanism in addition to
the heterogeneous reactions at solid electrodes. NO+ was
chosen as the chemical oxidant because it has a reduction
potential of +0.87 V vs Fc/Fc+ in CH3CN,10 which exceeds
the oxidation potential of the phenols by at least ∼+0.3-
0.4 V, and because it is easily obtained as high-purity
dry salts. Furthermore, NO+ is useful as an in situ
oxidant because it is NMR (1H and 13C) silent and because
its reduced form is easily removed from solution as a
gaseous product. The closely related higher melting point
a Refs 4 and 7. bOne resonance structure is displayed for each
compound. The listed potentials (vs Fc/Fc+) were obtained by
voltammetry and are the approximate values necessary to bring
about oxidation of the phenolic compounds but do not necessarily
correspond to the formal potential. The counterions for the charged
species are the supporting electrolyte cation [Bu4N+] and anion
[PF6-].
of several other forms of R-tocopherol that are easily
electrochemically accessible through proton and electron
transfers (Scheme 1). The phenoxonium closed-shell
cation (R-TO+) that will be described in more detail here
has recently been found to be surprisingly stable under
mild conditions in solution (a dry aprotic organic solvent),
although evidence for its existence has been restricted
to infrared and UV-vis spectroscopies.4 Phenoxonium
cations are thought to be critical intermediates in oxida-
tion processes associated with phenols but are usually
very unstable, with few reports of compounds stable
enough to characterize.6 Therefore, it is a significant
observation that the phenoxonium cation derived from a
naturally occurring compound is stable in solution.
(4) (a) Svanholm, U.; Bechgaard, K.; Parker, V. D. J. Am. Chem.
Soc. 1974, 96, 2409-2413. (b) Williams, L. L.; Webster, R. D. J. Am.
Chem. Soc. 2004, 126, 12441-12450.
Electrochemistry, 3rd ed.; Lund, H., Baizer, M. M., Eds.; Marcel
Dekker: New York, 1991; Chapter 16. (c) Rieker, A.; Beisswenger, R.;
Regier, K. Tetrahedron 1991, 47, 645-654. (d) Eickhoff, H.; Jung, G.;
Rieker, A. Tetrahedron 2001, 57, 353-364.
(5) (a) Nanni, E. J., Jr.; Stallings, M. D.; Sawyer, D. T. J. Am. Chem.
Soc. 1980, 102, 4481-4485. (b) Webster, R. D. Electrochem. Commun.
1999, 1, 581-584.
(8) (a) Du¨rckheimer, W.; Cohen, L. A. J. Am. Chem. Soc. 1964, 86,
4388-4393. (b) Marcus, M. F.; Hawley, M. D. J. Org. Chem. 1970, 35,
2185-2190.
(6) (a) Speiser, B.; Rieker, A. J. Chem. Res. (S) 1977, 314-315. (b)
Vigalok, A.; Rybtchinski, B.; Gozin, Y.; Koblenz, T. S.; Ben-David, Y.;
Rozenberg, H.; Milstein, D. J. Am. Chem. Soc. 2003, 125, 15692-15693.
(7) For comprehensive reviews on the anodic electrochemistry of
phenols, see: (a) Yoshida, K. Electrooxidation in Organic Chemistry.
The Role of Cation Radicals as Synthetic Intermediates; Wiley: New
York, 1983; pp 174-186. (b) Hammerich, O.; Svensmark, B. In Organic
(9) The supporting electrolyte also lowers the solution resistance to
enable electrolysis to be performed in organic solvents with low
dielectric constants.
(10) (a) Kochi, J. K. Acc. Chem. Res. 1992, 25, 39-47. (b) Connelly,
N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877-910.
J. Org. Chem, Vol. 70, No. 25, 2005 10467

Lee et al.
replacement of the phytyl group in R-TOH with a methyl
group in (CH3)R-TOH. The anodic peak current was
slightly less for R-TOH because of a smaller diffusion
coefficient brought about by the bulky phytyl chain. The
anodic-cathodic peak-to-peak separation (∆Epp) was
smaller on GC compared to that on Pt, although the peak
shapes on both electrodes were complicated. As the
temperature was lowered, the ∆Epp values increased
substantially, far more than can be accounted for by the
effects of uncompensated resistance. The voltammogram
for (COOH)R-TOH showed only a small reverse peak on
the cyclic voltammetry timescale, which might be caused
by less chemical reversibility (at T ) 293 K) compared
to the other phenols.
Digital simulation techniques allowed approximate
kinetic parameters to be calculated for the electrochemi-
cally induced oxidation reaction of R-TOH to produce
R-TO+ that is represented through the reversible series
of steps given in eqs 2-4 (Scheme 1, pathway 2).
Simulations were performed on the data obtained at 243
K (Figure 2) because CPE experiments had previously
indicated that at this temperature the chemical trans-
formations were completely reversible (at higher tem-
peratures, this might not be the case).4b
R-TOH - e- a R-TOH+•
FIGURE 1. Voltammetric data for 2 mM substrates obtained
at 293 K in CH3CN with 0.2 M Bu4NPF6 at 1 mm diameter
planar Pt (s) and GC (- - -) electrodes.
E° ) +0.55 V; ks ) 0.005 cm s-1 (2)
R-TOH+• a R-TO• + H+
nonnatural compounds 6-hydroxy-2,2,5,7,8-pentameth-
ylchroman [(CH3)R-TOH] and 6-hydroxy-2,5,7,8-tetra-
methylchroman-2-carboxylicacid[(COOH)R-TOH](Scheme
1) were also tested.
Keq ) 1 × 10-7 mol L-1
;
kf ) 1 × 103 s-1; kb ) 1 × 1010 L mol-1 s-1 (3)
R-TO• - e- a R-TO+
E° ) +0.13 V; ks ) 0.005 cm s-1 (4)
2. Results and Discussion
2.1. Mechanistic Considerations Derived From
Voltammetry Experiments. Cyclic (CV) and square
wave (SWV) voltammograms of R-TOH, (CH3)R-TOH, and
(COOH)R-TOH in CH3CN at 293 K are shown in Figure
1. The voltammograms are more complicated than ex-
pected for a simple reversible electron transfer because
they are affected by homogeneous reactions and surface-
based interactions of the solutes with the electrodes4b and
are associated with the pathway 2 series of electron
transfers and the homogeneous reaction given in Scheme
1. In each case, the peak values on the forward scan
direction for both CV and SWV experiments are at less
positive potentials on GC compared to those on Pt.
Because the potential of the voltammetric wave is
influenced by the kinetics of the homogeneous reactions,
the peak potentials (or half-wave potentials) do not
correspond to the thermodynamic formal potentials (E°).
The SWV shows larger anodic peak currents on GC
compared to those on Pt (especially for R-TOH), which is
thought to be associated with a rapid adsorption/desorp-
tion process on GC.4b The anodic peak currents observed
for (COOH)R-TOH appeared to be the same on Pt and
GC electrodes, suggesting less surface-based interactions
compared to (CH3)R-TOH and R-TOH.
R-TO• + R-TO• a products
Keq ) 1 × 104 L mol-1
;
kf ) 1 × 103 L mol-1 s-1; kb ) 1 × 10-1 s-1 (5)
The reversibility of the complete electrochemical trans-
formation (eqs 2-4) is critically determined by the
equilibrium constant for the proton dissociation reaction
in eq 3 because the Keq value must be sufficiently low to
enable the back-reaction to occur on the voltammetric
timescale. The values of the homogeneous rate constants
in eq 3 are also very important because R-TO• is not
stable in solution and decays by a bimolecular self-
reaction (eq 5).11 Therefore, the backward transformation
step in eq 3 must be faster than the forward bimolecular
self-reaction in eq 5, otherwise R-TO• would react before
it was further oxidized to the phenoxonium ion. The kf
value in eq 5 was previously calculated over a range of
temperatures by simulation of CV data obtained during
the oxidation of the phenolate anion (R-TO-) in CH3CN.5b
The Keq, kf, and kb values in eq 3 were estimated from
CV experiments with the addition of varying concentra-
tions of CF3SO3H because the addition of acid shifts the
equilibrium toward the protonated species (Figure 2)
The CVs observed for R-TOH and (CH3)R-TOH at the
same electrode and temperature were very similar,
indicating that the mechanism and kinetics of oxidation
were also likely to be very similar and unaffected by the
(11) Bowry, V. W.; Ingold, K. U. J. Org. Chem. 1995, 60, 5456-
5467.
10468 J. Org. Chem., Vol. 70, No. 25, 2005

Transformation of R-TOH into Its Phenoxonium Ion
FIGURE 2. (a) Cyclic voltammograms of 15 mM substrates
recorded in CH3CN with 0.5 M Bu4NPF6 at a 1 mm diameter
planar Pt working electrode at 243 K. The CV recorded in the
presence of 0.5 M CF3SO3H is offset by -40 µA. R-TOH (s).
R-TO+ (- - -), synthesized by exhaustive two-electron oxidation
of R-TOH in a controlled potential electrolysis cell. (b) DigiSim
simulations of the experimental data in (a), using the param-
eters given in eqs 2-5, with a solution resistance of 600 Ω.
The CV simulated in the presence of 0.5 M H+ is offset by -40
µA.
(CF3SO3H was assumed to fully dissociate in CH3CN).
At concentrations of acid above ca. 0.5 M, the forward
electrochemical transformation mainly produces R-TOH+•,
which can be reduced back to the starting material when
the scan direction is reversed; hence, the CV appears
similar to what is expected for a simple reversible
electron transfer (Figure 2). Simulations were also per-
formed on CV data obtained in solutions of R-TO+ (from
the bulk oxidation of R-TOH) that aided the estimation
of the kinetic parameters (Figure 2, dashed lines). The
rate and equilibrium constants in eq 2-5 are specific for
CH3CN at 243 K and are possibly quite different in lower
dielectric media like CH2Cl2 (see below) and at higher
temperatures. Importantly, the simulations confirm that
the back-reaction in eq 3 is fast and the equilibrium
constant is very low, which is the reason the transforma-
tion from phenol to a phenoxonium cation occurs readily
in both directions in CH3CN.
FIGURE 3. (a) Background subtracted solution-phase ATR-
FTIR spectra obtained at 233 K in CH3CN with between 15
and 40 mM phenols before (‚‚‚) and after (s) the addition of 2
mol equiv of NOSbF6 to form the phenoxonium cations. Each
spectrum represents 128 accumulated scans recorded at 2 cm-1
resolutions. (b) Theoretical infrared spectra calculated at the
EDF2/6-31+G* level.
solutions that within 2-3 min transformed into bright
orange/red solutions (Supporting Information Figure S1)
with very similar IR spectra between 1500 and 1800 cm-1
(Figure 3a, solid lines), indicating the formation of the
phenoxonium ions in high yields.4b Also consistent with
the formation of the phenoxonium cations was the
observation that the strong OH stretches in the FTIR
spectra of the phenolic starting materials at ca. 3400 cm-1
were missing from the spectrum of the oxidized com-
pounds.12
2.2. FTIR Spectroscopy. The chemical oxidation
reactions were performed at 233 K and monitored with
in situ attenuated total reflection Fourier transform
infrared (ATR-FTIR) spectroscopy. Oxidation of the three
phenols with 2 mol of NO+ (eq 6) produced dark green
The “H+” released during oxidation of the phenol
probably exists coordinated to the solvent and weakly ion
-
paired with the SbF6 anion (eq 6). In addition to the
bands detected at 1605 and 1649 cm-1, (CH3)R-TO+ and
(COOH)R-TO+ showed a strong band at 1670 cm-1, which
was also present in the spectrum of R-TO+ but was of
2NO+SbF6
-
R-TOH y
z
CH3CN
(12) Pouchert, C. J. The Aldrich Library of FT-IR Spectra, 1st ed.;
Aldrich Chemical Co., Inc.: Milwaukee, WI, 1985; Vol. 1, pp 452-457.
R-TO+SbF6- + [CH3CNH]+SbF6- + 2NO(g) (6)
J. Org. Chem, Vol. 70, No. 25, 2005 10469

Lee et al.
lower intensity and so was not highlighted in an earlier
study that used electrolysis procedures to generate
R-TO+.4b (COOH)R-TOH and (COOH)R-TO+ showed extra
bands at ∼1750 cm-1 due to the presence of the carboxylic
acid functional group. In each case, the spectra of the
oxidized compounds and the solution color remained
constant over a period of at least 2 h (at 233 K), indicating
substantial stability of the phenoxonium cations.
When 200 mM of water was added to the solutions
containing the phenoxonium cations at 233 K, the color
faded from bright orange/red to pale yellow over a period
of approximately 10 min. During the color change, the
infrared peaks at 1605, 1649, and 1670 cm-1 disappeared
completely while a new band appeared at 1641 (oxidized
R-TOH and (CH3)R-TOH) or 1645 cm-1 (oxidized
(COOH)R-TOH). The spectroscopic results indicate that
a transformation occurs in the presence of relatively low
levels of water with the strong absorbance at 1641 or
1645 cm-1, indicating the formation of new carbonyl-
containing compound(s) (such as 1 or 2 in eq 1).
starting material showed no absorbancies above 1600
cm-1 and showed several strong absorbancies between
1550 and 1100 cm-1 that coincided reasonably well with
the experimental spectrum, albeit shifted to higher
wavenumbers by approximately 50 cm-1. The calcula-
tions predicted three new strong bands between 1800 and
1600 cm-1 in the theoretical spectrum of the cation, which
correlate with three strong bands observed in the experi-
mental spectrum of (CH3)R-TO+, although with closer
spacing between the bands. The theoretical absorptions
correspond to a CdO stretch, a symmetric CdC ring
stretch, and an asymmetric CdC ring stretch, from
highest to lowest wavenumber, respectively. (Ball-and-
stick diagrams showing the stretching modes are given
in Supporting Information Figure S4.) It is possible that
the highest intensity absorption observed in the experi-
mental spectrum at 1649 cm-1 is associated with the
carbonyl stretch, whereas the relatively weaker bands
at 1670 and 1605 cm-1 are due to the CdC ring sym-
metric and asymmetric stretching modes, respectively.
1
2.3. NMR Spectroscopy. H and 13C NMR experi-
Chemical oxidation experiments were also performed
on 2,4,6-tri-tert-butylphenol and 2,6-di-tert-butyl-4-meth-
oxyphenol, two compounds that are known to form stable
phenoxyl radicals.13 Oxidation of 2,4,6-tri-tert-butylphe-
nol at 233 K produced darkly colored solutions that
became pale yellow within 1 min. A strong band at 1656
cm-1 did not change when water was added, suggesting
that it was associated with a relatively stable product
but not a phenoxonium cation (the rapid color change also
suggested that the phenoxonium cation quickly decom-
posed, if it actually formed). Oxidation of 2,6-di-tert-butyl-
4-methoxyphenol produced a darkly colored solution that
transformed to an orange/red color and then within 5 min
faded to a pale yellow color. In situ ATR-FTIR spectros-
copy detected transitory absorbancies at 1640 and 1575
cm-1 that diminished as the orange/red color faded, which
is consistent with the presence of a phenoxonium cation,
albeit of limited stability (Supporting Information Figure
S2). The final pale yellow solution had strong bands at
1655 and 1602 cm-1, consistent with a stable species with
a quinone structure. It is possible that the oxygen atom
in the para position to the carbonyl group aids in the
stabilization of the phenoxonium cations, but other
factors must also be involved to account for the high
stability of the phenoxonium cation of vitamin E and its
analogues compared to only limited suggested stability
of the 2,6-di-tert-butyl-4-methoxyphenoxonium cation.
When the solvent was removed from solutions contain-
ing the (CH3)R-TO+ and (COOH)R-TO+ cations at 243 K
with a diffusion pump (p ) 1 × 10-5 mmHg), red solid
materials were obtained that consisted of the phenoxo-
nium cation and acid (eq 6). The solid material could be
handled in a drybox at room temperature and appeared
stable for around 24 h, although it became sticky and
changed to a much darker red color with increasing time.
Infrared spectra of the fresh solid compounds showed
bands at 1670, 1649, and 1605 cm-1, indicating that the
phenoxonium cations were semistable at room temper-
ature in the solid state (Supporting Information Figure
S3) and that the infrared spectra in the carbonyl region
were not influenced by solvation effects.
ments were conducted on oxidized solutions of (CH3)R-
TOH in CD3CN at 233 K because the high solubility
minimized the required 13C NMR spectral collection
times, compared to those for R-TOH and (COOH)R-TOH,
and decreased the likelihood of decomposition occurring.
Furthermore, the presence of two methyl groups in
(CH3)R-TO+ removes the chirality at the quaternary
carbon and simplifies the 1H NMR spectra (compared to
those of (COOH)R-TO+ and R-TO+) and the absence of
the phytyl chain simplifies the 13C NMR spectrum
compared to that of R-TO+. Both the 1H and the 13C NMR
spectra of (CH3)R-TO+ (Figure 4) were very clean and free
of any decomposition products, indicating that the chemi-
cal oxidation occurs quantitatively and that the oxidized
products were completely stable in solution on the
timescale of the experiment (several hours). The 1H NMR
line width of (CH3)R-TO+ was greater than that of
(CH3)R-TOH (Figure 4) which was most likely caused by
the presence of a small amount of the (CH3)R-TOH+•
radical. Previous in situ electrochemical-NMR (1H) ex-
periments performed during the oxidation of R-TOH
using a 10 nm thick Au film working electrode positioned
within the radio frequency (RF) coils of the NMR mag-
net14 (where electrolysis was not exhaustive) detected the
rapid disappearance of all the peaks in the cyclic portion
of the molecule due to rapid electron exchange with the
cation radical.4b
In the 13C NMR spectrum, the quaternary carbon shifts
from 73.1 ppm in the neutral molecule up to 99.9 ppm in
the cation, consistent with a decrease in electron density
in the region of the quaternary carbon. Also, in the 13C
NMR spectrum, the cationic complex shows two bands
in the carbonyl region (194.8 and 184.5 ppm) consistent
with a quinoid complex and with the downfield band
associated with the strongly electropositive carbon in the
quinoid ring bonded to the chromanol oxygen (atom d in
Figure 4). The 1H NMR spectra also support an increase
in positive charge in the region around the quaternary
carbon, with all proton resonances in the vicinity of the
Theoretical infrared spectra were calculated for (CH3)R-
TOH and (CH3)R-TO+ (Figure 3b) using the EDF2/6-
31+G* model. The unscaled theoretical spectrum of the
(13) Webster, R. D. Electrochem. Commun. 2003, 5, 6-11, and
references therein.
(14) Webster, R. D. Anal. Chem. 2004, 76, 1603-1610.
10470 J. Org. Chem., Vol. 70, No. 25, 2005

Transformation of R-TOH into Its Phenoxonium Ion
FIGURE 5. HF/6-31+G* potential-derived electrostatic charges
for (CH3)R-TO+. Hydrogen atoms are omitted for clarity.
cations are possibly not as thermally stable in solution
as (CH3)R-TO+, and thus, decomposition occurred while
transferring the sample to the spectrometer. The 13C
NMR spectrum of R-TO+ showed bands at 102.4, 187.6,
and 195.1 ppm, in close agreement with the quaternary
carbon and carbonyl-type resonances observed in the
spectrum of (CH3)R-TO+. The phenoxonium cations are
substantially more soluble in CH3CN than their parent
compounds (R-TOH and (COOH)R-TOH are soluble at
∼15 mM concentrations in CH3CN, and (CH3)R-TOH is
soluble up to at least 1 M at 293 K) and are consequently
difficult to precipitate from solution as pure compounds
free from the acid byproduct. Furthermore, thermal and
moisture instability have made additional purification
and crystallization of the cations extremely difficult.
It is likely that other oxidizing agents will also produce
the phenoxonium cations, providing they can be used in
dry conditions. However, it was found that Ag+ (used as
AgNO3 or AgClO4) was not a sufficiently strong oxidant
in CH3CN to oxidize R-TOH to R-TO+. The choice of the
counteranion was also found to be important in stabiliz-
ing the phenoxonium cations. Cyclic voltammetry experi-
ments indicated that the cation radical or phenoxonium
ion were unstable in the presence of chloride (possibly
due to homogeneous oxidation of the halide by R-TOH+•
or R-TO+).
FIGURE 4. NMR spectra (500 MHz 1H and 125.721 MHz 13C)
at 233 K in CD3CN of 0.57 M (CH3)R-TOH and its associated
phenoxonium cation, (CH3)R-TO+, formed by reacting the
1
starting material with 2 mol of NOSbF6 at 233 K (eq 6). H
spectra represent 128 accumulated scans, and 13C spectra
represent 2000 accumulated scans.
nonaromatic chromanol ring shifting upfield. As expected,
1
the OH resonance in the H NMR spectrum of (CH3)R-
TOH at 5.4 ppm was missing from the spectrum of
(CH3)R-TO+. A broad band occurred in the 1H NMR
spectrum of (CH3)R-TO+ at ca. 10 ppm at 233 K (and ca.
12 ppm at 298 K), consistent with the presence of the
acidic byproduct, [CD3CNH]+SbF6
.
-
Extra peaks appeared in the 1H and 13C NMR spectra
as the (CH3)R-TO+ solution was warmed to room tem-
perature, although the phenoxonium cation could still be
detected after several hours at 298 K. When 10 µL of
water (∼1 M) was added to the NMR tube, the 1H NMR
peaks corresponding to (CH3)R-TO+ disappeared im-
mediately and the NMR spectra became very compli-
cated, indicating that multiple products were being
formed (thin-layer chromatography (TLC) analysis also
indicated that numerous products were formed) (Sup-
porting Information Figure S5). Therefore, the decom-
position of phenoxonium cations in CD3CN containing low
levels of water occurs by a mechanism more complicated
than the simple hydrolysis mechanism depicted in eq 1.
The NMR spectra of oxidized solutions of R-TOH and
especially (COOH)R-TOH were not as clean as the
spectra of (CH3)R-TO+ and indicated that some decom-
position of the phenoxonium cation was occurring on the
experimental timescale. The R-TO+ and (COOH)R-TO+
1H and 13C NMR predictions using fully empirical
correlations usually provide an excellent match to ex-
perimental NMR spectra. However, empirical predictions
of the NMR spectra of the phenoxonium cation were not
productive because of the scarcity of data on related
compounds (we are aware of no published NMR data on
phenoxonium cations) and because of delocalization of the
positive charge. Potential-derived charges (HF/6-31+G*)
for each of the non-hydrogen atoms in the phenoxonium
cation are shown in Figure 5. The assumption that the
calculated 13C NMR spectrum of the phenoxonium cation
(Supporting Information Figure S6) is acceptable at the
HF/6-31G* level of theory is supported by the accuracy
of the theoretical spectrum of the starting phenol com-
J. Org. Chem, Vol. 70, No. 25, 2005 10471

Lee et al.
pared to that of the experimental spectrum (compare
Figure 4 and Supporting Information Figure S6). The
theoretical calculations predict that the chemical shift
of the quaternary carbon increases from 61.1 ppm (73.1
ppm experimental) in the neutral molecule up to 86.4
ppm (99.9 ppm experimental) in the cation. The theoreti-
cal calculations predict chemical shifts of 199.6 ppm
(194.8 ppm experimental) and 173.5 ppm (184.5 ppm
experimental) for the carbon atoms bonded to oxygen
atoms in the phenoxonium cation.
2.4. Chemical Oxidation with NOSbF6 in CH2Cl2.
The addition of NOSbF6 to R-TOH, (CH3)R-TOH, and
(COOH)R-TOH in pure CH2Cl2 resulted in the formation
of dark-green colored solutions with electron paramag-
netic resonance (EPR) experiments indicating the pres-
ence of paramagnetic compounds. The obtained EPR
spectra of the three oxidized phenols in CH2Cl2 were
similar, indicating that the structures of the radicals were
closely related (Supporting Information Figure S7a), and
were the same as the spectrum obtained by electrochemi-
cal oxidation of R-TOH in CH3CN in the presence of CF3-
SO3H (that also produced green solutions).4b The radical
in the green solutions had previously been assigned to
R-TOH+• on the basis of the observation that the signal
intensity and stability increased as acid was added to
the solution as a result of the equilibrium in eq 3 shifting
toward the protonated species. The R-TO• radical can be
electrochemically formed by one-electron oxidation of the
phenolate anion (Scheme 1)5 but was not detected in this
work (it has a substantially different EPR spectrum from
the spectra assigned to the phenol cation radicals (Sup-
porting Information Figure S7b)).5b,11,15 The R-TO• radical
was not expected to be stable under the conditions used
for this study because its low oxidation potential (∼+0.2
V vs Fc/Fc+) necessitates that it will undergo immediate
further oxidation with NO+ to form R-TO+ (Scheme 1)
(no attempt was made at trapping the R-TO• radical with
spin markers). In contrast, R-TOH+• cannot be oxidized
until at least +1.4 V vs Fc/Fc+ (Scheme 1);4 therefore, it
is stable in the presence of NO+ and its concentration in
solution is determined by Keq in eq 3. The relative amount
of cation radicals was difficult to quantify, but it is likely
they exist together in solution with the phenoxonium
cations because of the equilibrium constant in eq 3
shifting toward the protonated species in the low dielec-
tric constant solvent CH2Cl2 (similar to that observed in
acidified CH3CN). The cation radicals are more intensely
colored than the phenoxonium cations and so dominate
the overall solution color (for R-TOH+•, λmax ) 464 nm
and ꢀ g 1.3 × 104 L cm-1 mol-1).4
FIGURE 6. Voltammetric data for 2 mM substrates obtained
at 293 K in CH2Cl2 with 0.2 M Bu4NPF6 at 1 mm diameter
planar Pt (s) and GC (- - -) electrodes.
reduction of the cation radical) and at ∼ +0.15 V vs Fc/
Fc+ (associated with the reduction of the phenoxonium
cation). The reverse peak at +0.4 V was larger in
solutions containing (CH3)R-TOH compared to those
containing R-TOH, suggesting a higher proportion of the
cation radical. Although the CV behavior in CH2Cl2
appeared simpler than that observed in CH3CN, particu-
larly at higher temperatures (in the sense that the
voltammetry was very similar on both electrode surfaces
in CH2Cl2), the longer-term bulk oxidation experiments
were simpler in CH3CN because the reaction proceeded
quantitatively to the phenoxonium cation (as proven by
the NMR and IR experiments). It was interesting to
observe that the CV behavior of (COOH)R-TOH was
significantly different from that of the two other phenols
when the scan direction was reversed (in CH3CN and
CH2Cl2), but the bulk species produced by chemical
oxidation were the same (as indicated by IR experiments).
3. Conclusions
CV experiments performed on R-TOH and (CH3)R-TOH
in CH2Cl2 were similar on GC and Pt electrodes and
clearly showed the presence of the cation radicals when
the scan directions were reversed (that were not detected
in pure CH3CN) (Figure 6). On the forward scan, one
anodic peak was detected at ∼ +0.5 V vs Fc/Fc+ (similar
to that in CH3CN), and on the reverse scan, cathodic
peaks were observed at ∼ +0.4 V (associated with the
Chemical oxidation of R-TOH with the structurally
similar nonnatural phenols (CH3)R-TOH and (COOH)R-
TOH with NOSbF6 in CH3CN at 233 K is a straightfor-
ward method for preparing the phenoxonium cations in
high concentrations in solution in accordance with het-
erogeneous electrochemical electrolysis procedures.4 The
chemical oxidation reaction occurs in three steps in CH3-
CN: initial one-electron oxidation to form the phenol
cation radical, deprotonation of the cation radial to form
the neutral radical, and further one-electron oxidation
of the neutral radical to form the phenoxonium cation.
When the oxidation reaction is conducted in the low
dielectric constant solvent CH2Cl2, the deprotonation
(15) (a) Kohl, D. H.; Wright, J. R.; Weissman, M. Biochim. Biophys.
Acta 1969, 180, 536-544. (b) Ozawa, T.; Hanaki, A.; Matsumoto, S.;
Matsuo, M. Biochim. Biophys. Acta 1978, 531, 72-78. (c) Mukai, K.;
Tsuzuki, N.; Ishizu, K.; Ouchi, S.; Fukuzawa, K. Chem. Phys. Lipids
1981, 29, 129-135. (d) Matsuo, M.; Matsumoto, S.; Ozawa, T. Org.
Magn. Reson. 1983, 21, 261-264.
10472 J. Org. Chem., Vol. 70, No. 25, 2005

Transformation of R-TOH into Its Phenoxonium Ion
4.3. Theoretical Calculations. Cyclic voltammetry simu-
lations were performed using DigiSim 3.03.17 Molecular orbital
calculations were performed using a development version of
the Q-Chem 2.1 software package18 and the Spartan 04
software package.19 Harmonic vibrational frequencies were
calculated by Q-Chem using the EDF2/6-31+G* density
functional model20 and the SG-1 quadrature grid,21 and no
empirical scale factors were applied. NMR chemical shifts
(without spin-spin coupling) were calculated by Spartan at
the HF/6-31G* level using molecular structures optimized at
the same level. Electrostatic potential maps and potential-
derived atomic charges were computed by Spartan at the HF/
6-31+G* level using structures optimized at the same level.
reaction is partially inhibited resulting in solutions
containing the phenol cation radicals that are detectable
by EPR spectroscopy. Therefore, direct oxidation of
vitamin E in a lipophilic membrane could result in the
formation of both R-TOH+• and R-TO+, depending on the
dielectric properties of the medium. The formation of
R-TO• in a lipophilic membrane is more likely to involve
a hydrogen atom abstraction mechanism rather than an
oxidation mechanism, unless the oxidation occurs through
the phenolate anion, R-TO- (Scheme 1).
CH3CN solutions containing each of the phenoxonium
cations display three strong infrared absorbancies at
1670, 1649, and 1605 cm-1. Theoretical calculations
predict that the fundamental vibrational modes corre-
sponding to the absorbancies are a CdO stretch, a
symmetric CdC ring stretch, and an asymmetric CdC
ring stretch. 13C NMR spectroscopy experiments and
theoretical calculations performed on (CH3)R-TO+ indi-
cate that the positive charge is shared between the
carbon atoms bonded to oxygen atoms and the quater-
nary carbon in the chromanol ring. The kinetic stability
of the intermediate phenol cation radicals and neutral
radicals (R-TOH+• and R-TO•) does not appear to be an
important prerequisite for the stability of the phenoxo-
nium cations because R-TO• is kinetically less stable than
R-TO+, especially at high concentrations. The crucial
factor in the reversibility of the chemical transformation
in eq 6 in CH3CN is that, in contrast to most phenol
cation radicals, the intermediate cation radical formed
by the initial one-electron oxidation is only weakly acidic
(Keq ∼ Ka ∼ 10-7 mol L-1).
Acknowledgment. R.D.W. thanks the ARC for the
award of a QEII Fellowship.
Supporting Information Available: Photograph showing
typical red/orange color of phenoxonium cations. Solution-
phase ATR-FTIR spectra obtained during the electrolysis of
2,6-di-tert-butyl-4-methoxyphenol and solid-state FTIR spectra
of (COOH)R-TO+. Theoretical normal modes of vibration of
(CH3)R-TO+. 1H NMR spectra showing the decomposition of
(CH3)R-TO+. Theoretical 13C NMR spectra of (CH3)R-TOH and
(CH3)R-TO+. EPR spectra of cation radicals produced by
chemical oxidation of the phenols in CH2Cl2. Summary of
results from theoretical calculations (Cartesian coordinates,
no. of imaginary frequencies, and total energies). This material
is available free of charge via the Internet at http://pubs.acs.org.
JO0517951
(16) Webster, R. D. J. Chem. Soc., Perkin Trans. 2 2002, 1882-
1888.
(17) (a) DigiSim;Bioanalytical Systems, Inc. (BAS): West Lafayette,
IN. (b) Rudolph, M.; Reddy, D. P.; Feldberg, S. W. Anal. Chem. 1994,
66, 589A-600A.
4. Experimental Section
(18) Kong, J.; White, C. A.; Krylov, A. I.; Sherrill, C. D.; Adamson,
R. D.; Furlani, T. R.; Lee, M. S.; Lee, A. M.; Gwaltney, S. R.; Adams,
T. R.; Ochsenfeld, C.; Gilbert, A. T. B.; Kedziora, G. S.; Rassolov, V.
A.; Maurice, D. R.; Nair, N.; Shao, Y.; Besley, N. A.; Maslen, P. E.;
Dombroski, J. P.; Daschel, H.; Zhang, W.; Korambath, P. P.; Baker,
J.; Byrd, E. F. C.; Van Voorhis, T.; Oumi, M.; Hirata, S.; Hsu, C.-P.;
Ishikawa, N.; Florian, J.; Warshel, A.; Johnson, B. G.; Gill, P. M. W.;
Head-Gordon, M.; Pople, J. A. J. Comput. Chem. 2000, 21, 1532-1548.
(19) Spartan 04 for Macintosh; Wavefunction Inc.: Irvine, CA.
(20) Lin, C. Y.; George, M. W.; Gill, P. M. W. Aust. J. Chem. 2004,
57, 365-370.
4.1. Reagents. HPLC grade acetonitrile and analytical
grade dichloromethane were dried and stored over calcium
hydride (under nitrogen) and distilled immediately prior to
use, and acetonitrile-d3 (99.8% D) was stored over 4 Å
molecular sieves under nitrogen. Chemical oxidation experi-
ments were conducted using conventional vacuum-line tech-
niques to ensure a moisture-free environment.
4.2. Apparatus. Solution-phase FTIR experiments were
conducted utilizing a diamond composite attenuated total
reflection (ATR) probe.16
(21) Gill, P. M. W.; Johnson, B. G.; Pople, J. A. Chem. Phys. Lett.
1993, 209, 506-512.
J. Org. Chem, Vol. 70, No. 25, 2005 10473