Development of novel antioxidants: Design, synthesis, and reactivity

Article abstract of DOI:10.1021/jo0301090

We are attempting to develop novel synthetic antioxidants aimed at retarding the effects of free-radical induced cell damage. In this paper we discuss the design strategy and report the synthesis of seven novel antioxidants, including six catechols and a

Full text of DOI:10.1021/jo0301090

Develop m en t of Novel An tioxid a n ts: Design , Syn th esis, a n d
Rea ctivity
Helmi H. Hussain,† Gordana Babic,† Tony Durst,*,† J ames S. Wright,*,‡ Mihaela Flueraru,‡
Alexandru Chichirau,‡ and Leonid L. Chepelev‡
Department of Chemistry, University of Ottawa, D’Iorio Hall, 10 Marie Curie Street,
Ottawa, Canada K1N 6N5, and Department of Chemistry, Carleton University,
1125 Colonel By Drive, Ottawa, Canada K1S 5B6
tdurst@science.uottawa.ca; jim_wright@carleton.ca
Received March 31, 2003
We are attempting to develop novel synthetic antioxidants aimed at retarding the effects of free-
radical induced cell damage. In this paper we discuss the design strategy and report the synthesis
of seven novel antioxidants, including six catechols and a benzylic phenol. The bond dissociation
enthalpy (BDE) for the most active (weakest) OH bond in each molecule was calculated by theoretical
methods, as well as the BDE for the semiquinone radical. Reaction rates with the nitrogen-centered
free radical DPPH• were measured in ethyl acetate. The log of kDPPH for bimolecular reaction
correlated well with the primary BDE. The correlation between rate constants and calculated BDEs
shows that the BDE is a good predictor of antioxidant activity with DPPH•, suggesting that our
design criteria are useful and that these compounds should undergo further testing in cell cultures
and in animal models.
In tr od u ction
increase in antioxidant activity.2,3 More general accounts
on how to approach the problem were given by Thomas4
and by Noriguchi and Niki.5 Some of the present authors6
recently described the synthesis of a new class of com-
pounds, the naphthalene diols, which aimed to improve
on the activity of such antioxidants while at the same
time preventing toxicity due to quinone formation, and
demonstrated enhanced reactivity with a nitrogen-
centered test free radical, DOPPH• (2,2-di(4-tert-octylphe-
nyl)-1-picrylhydrazyl). DOPPH• is a more lipid-soluble
version of the related radical DPPH• (2,2-diphenyl-1-
picrylhydrazyl), which is widely used as a test radical.
These authors were able to show rate enhancements for
some naphthalene diols in heptane solution that were
over 2 orders of magnitude larger than those for R-toco-
pherol, the standard reference compound.6
There is currently under way a tremendous effort to
study the properties of naturally occurring antioxidants
because of their potential clinical significance. Thus free
radicals generated in biological systems, particularly
alkyl peroxyl (ROO•), which occurs in lipid peroxidation
and superoxide ion (O2-), which is formed during normal
metabolism, have been implicated in aging and in age-
related degenerative diseases. Many authors have dem-
onstrated protective effects of different chemical antiox-
idants in cellular systems, e.g. R-tocopherol and related
compounds, ascorbic acid, the flavonoids, stilbenes re-
lated to resveratrol, catechins which occur naturally in
tea, glutathione and related thiols, etc.1 The antioxidants
often work cooperatively, so that the radical derived from
R-tocopherol reacts with ascorbic acid (low pH) or ascor-
bate ion (higher pH) to regenerate R-tocopherol. The
ascorbyl radical that forms can then either disproportion-
ate to ascorbate plus dehydroascorbic acid or be reduced
back to ascorbic acid by interaction with other reducing
species in the cell.1
In designing synthetic antioxidants, it is helpful to
consider what kind of molecular structure is optimum
and what kind of design parameters are relevant to the
problem.2,5,7 Generally speaking compounds containing
weak C-H bonds which lead to carbon-centered radicals
R• are undesirable, because such compounds usually
rapidly add molecular oxygen and thus become chain
propagating peroxyl radicals ROO• (although some may
There has been much less work on the systematic study
of synthetic antioxidants designed to optimize the anti-
oxidant activity, while at the same time satisfying other
important criteria, including solubility, bioavailability,
and lack of toxicity (or selective toxicity in the case of
cancer chemopreventive antioxidants). Earlier work con-
sidered variations on the vitamin E polyalkylchromanol
class of compounds, and was able to demonstrate some
(2) Burton, G.; Ingold, K. U. Acc. Chem. Res. 1986, 194, 194-201.
(3) Mukai, K.; Okabe, K.; Hosose, H. J . Org. Chem. 1989, 54, 557-
560.
(4) Thomas, C. E. In Handbook of synthetic antioxidants; Packer,
L., Cadenas, E., Eds.; Marcel Dekker: New York, 1997; pp 1-52.
(5) Noguchi, N.; Niki, E. Free Radicals Biol. Med. 2000, 28, 1538-
1546.
(6) Foti, M. C.; J ohnson, E. R.; Vinqvist, M. R.; Wright, J . S.; Barclay,
L. R. C.; Ingold, K. U. J . Org. Chem. 2002, 67, 5190-5196.
(7) Wright, J . S. Chem. Br. 2003, 39, 25-27.
† University of Ottawa.
‡ Carleton University.
(1) Halliwell, B.; Gutteridge, J . M. C. Free Radicals in Biology and
Medicine; Oxford University Press: Oxford, UK, 1999.
10.1021/jo0301090 CCC: $25.00 © 2003 American Chemical Society
Published on Web 08/13/2003
J . Org. Chem. 2003, 68, 7023-7032
7023

Hussain et al.
not).8,9 Sulfur-centered radicals are of interest and play
an important role in biochemistry, although they are also
capable of forming peroxyl radicals of the type RSOO•
by reaction with O2.1 Nitrogen-centered radicals are a
possibility and can be stable with respect to oxygen
addition, as in DPPH•, but by far the most studied
molecules involve oxygen-centered radicals. The virtue
of the latter is that oxygen-centered radicals (e.g. ArO-H
f ArO•) do not react with oxygen to form a trioxyl radical
ArOOO• since this step is endoergic by ca. 20 kcal/mol.10
The second design point is that the molecule should react
rapidly with peroxyl radicals. Since ROO• forms a bond
in ROO-H with a bond dissociation enthalpy (BDE) of
ca. 88 kcal mol-1, this requires that an antioxidant
ArO-H have a BDE below that value, to provide an
exoergic and therefore rapid reaction. Since R-tocopherol
has a (experimental) BDE of 77 kcal/mol2, we should
design an antioxidant with a BDE less than that value.
Ascorbate ion has a BDE of 68.5 kcal/mol11 and if
ascorbate is to regenerate the synthetic ArO-H, its BDE
must be less than that of ArO-H. This places the useful
design window at about 68-75 kcal/mol. Third, the
radical that is formed should be relatively unreactive
with lipid, protein, or other biological substrates. Fourth,
the antioxidant or radicals derived from it should not
react with molecular oxygen at an appreciable rate to
produce superoxide anion or its conjugate acid, hydrop-
eroxyl radical (autoxidation). Finally, to design lipid-
soluble antioxidants, it may be necessary to add func-
tional groups whose only purpose is to improve solubility,
e.g. the phytl tail in R-tocopherol, or to improve transport
through the cell wall, e.g. by esterifying the alcohol
groups.
SCHEME 1a
a
Reagents and conditions: (a) K2CO3, acetone/CH3l; (b) tBuLi/
THF, 78 °C/CH3l; (c) Pb(OAc)4/benzene, 50 °C; (d) 80% HOAc; (e)
Na2S2O4/ether/water.
we obtained a BDE for phenol of 87.10 kcal mol-1 in good
agreement with a current experimental gas-phase value of 87.3
( 1 kcal mol-1 obtained by Wayner et al.,15a although some-
what below the value of 88.3 ( 0.8 kcal mol-1 obtained by
Pedulli and co-workers15b in benzene solution. For the present
calculations, we estimate our BDE values to be accurate to
within ca. 1-2 kcal mol-1 of gas-phase experiments, i.e., to
essentially within experimental error. For determination of
the ionization potential we calculated the adiabatic ionization
potential at 0 K. This value corresponds to the internal energy
change ∆E°0 for ArOH (g) f ArOH+ (g) + e- (see Experimental
Section for details of the calculation).
Mea su r em en t of R a t e Con st a n t for R ea ct ion w it h
DP P H•. Kinetic measurements with DPPH• were similar to
those described previously by Foti et al.,6 except that we used
the more polar DPPH• and ethyl acetate solvent instead of
DOPPH• in heptane (for details, see the Experimental Section).
The experiments were carried out in a stopped-flow apparatus
at the National Research Council in Ottawa. The decay of
DPPH• was monitored at 519 nm in the presence of varying
(excess) concentrations of antioxidant at room temperature,
giving the second-order reaction (eq 1):
In the present paper we describe the synthesis and
testing of seven novel antioxidants, six of which fall into
or near our design window and one that deliberately lies
outside it. BDEs are calculated for first and second
oxidations (i.e. for successive loss of H-atoms) for each,
and it is shown that the log of the reaction rate constant
with DPPH• correlates inversely with the primary BDE,
as expected. Results are compared with the phenols and
naphthalene diols described previously and it is shown
that these novel compounds in general validate the
reactivity predictions based on BDE.
ArO-H + DPPH• f ArO• + DPPH2
(1)
Under these pseudo-first-order conditions the observed rate
constant kobs is given by eq 2:
kobs ) k0 + kDPPH[ArOH]
(2)
where k0 is the intercept and the second-order rate constant
kDPPH is the slope of the plot of kobs vs [ArOH].
Th eor etica l a n d Exp er im en ta l Meth od s
Ca lcu la tion of BDE a n d IP . For calculation of the BDE
we used the lowest level method (LLM) described by DiLabio
et al.12 Details of this calculation, which uses a B3LYP density
functional approach, are described in the Experimental Sec-
tion. The BDE corresponds to the standard gas-phase enthalpy
change at 298 K (∆H°298) for ArO-H (g) f ArO• (g) + H• (g).
In cases where there are two OH groups as in the catechols,
we use the term BDE1 for loss of the first (most weakly bound)
H-atom to form the semiquinone and BDE2 for loss of the
second H-atom to form the quinone. Using this procedure13,14
Resu lts a n d Discu ssion
A. Syn th esis. The synthesis of the catechol (2, 3-meth-
yl-4-methoxy-1,2-dihydroxybenzene) starting from sesa-
mol (3) is outlined in Scheme 1. Deprotection of the
methylendioxy group in 4 was accomplished by treatment
with Pb(OAc)4 in refluxing benzene, which generated the
intermediate ortho ester 5. Hydrolysis of 5 in aqueous
acetic acid gave in 85% yield a mixture of the desired
catechol 2 and the corresponding o-quinone 6 as a reddish
oil, accompanied by 15% of the carbonate 7. The red
mixture, dissolved in ether, when stirred with aqueous
(8) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1976, 9, 13-19.
(9) Scaiano, J . C.; Martin, A.; Yap, G. P. A.; Ingold, K. U. Org. Lett.
2000, 2, 899-901.
(10) Wright, J . S.; Chepelev, L. Unpublished results.
(11) DiLabio, G. A.; Wright, J . S. Free Radicals Biol. Med. 2000,
29, 480-485.
(14) Wright, J . S.; J ohnson, E. R.; DiLabio, G. A. J . Am. Chem. Soc.
2001, 123, 1173-1183.
(15) (a) Wayner, D. D. M.; Lusztyk, E.; Page, D.; Ingold, K. U.;
Mulder, P.; Laarhoven, L. J . J . A.; Aldrich, H. S. J . Am. Chem. Soc.
1995, 117, 8737-8744. (b) Lucarini, M.; Pedulli, G. F.; Cipollone, M.
J . Org. Chem. 1994, 59, 5063.
(12) DiLabio, G. A.; Pratt, D. A.; LoFaro, A. D.; Wright, J . S. J . Phys.
Chem. A 1999, 103, 1653-1661.
(13) Wright, J . S.; Carpenter, D. J .; McKay, D. J .; Ingold, K. U. J .
Am. Chem. Soc. 1997, 119, 4245-4252.
7024 J . Org. Chem., Vol. 68, No. 18, 2003

Development of Novel Antioxidants
SCHEME 2a
SCHEME 4
a
Reagents and conditions: (a) tBu(Me)2SiCl/imidazole/DM; (b)
K2CO3/acetone/CH3l; (c) Bu4NF/CH2Cl2; (d) SnCl4, Bu3N; (CH2O)n,
°
toulene, 95 C; (e) H2O2/NaOH.
SCHEME 3a
SCHEME 5
a
Reagents and conditions: (a) allyl bromide/acetone/CH3l/
K2CO3; (b) decalin, 190 °C; (c) H(+)/benzene; (d) SnCl4, Bu3N,
(CH2O)n, toluene, 95 °C; (e) H2O2/NaOH; (f) Na2S2O4.
aqueous sodium dithionite. The spectroscopic properties
of 12 and 17 are in agreement with the assigned
structure; see the Experimental Section.
sodium dithionite gave 2 as a yellowish oil in 91% yield.
This oil takes on a reddish color shortly after exposure
to air, indicative of possible oxidation to its o-quinone 6.
However, the 200-MHz proton NMR spectrum taken
after several days of storage in a closed vessel continued
to indicate the presence of more than 98% of the catechol
structure.
2,3,5-Trimethylhydroquinone, 8, was chosen as the
starting material for the preparation of the fully substi-
tuted monocyclic catechol 11 (Scheme 2). Protection of
the less hindered OH group in 8, methylation of the
remaining phenolic group, and deprotection afforded the
methoxyphenol 9 in greater than 80% overall yield.
o-Formylation with the Vilsmaier-Hack reagent gave the
salicyladeyde 10, which upon reaction with H2O2/NaOH
treatment yielded the desired 11.
Formylation of γ-tocophorol, 18, under the SnCl4/Bu3N/
(CH2O)n conditions afforded the fully substituted salicy-
laldehyde 19 in 19% yield (Scheme 4). Treatment with
H2O2 and NaOH gave 50% of isolated o-quinone 20.
Reduction with aqueous sodium dithionite as before
yielded the catechol 21 as an oily material that quickly
turned reddish due to its ease of oxidation back to 20
(Scheme 4).
Finally, the symmetrical methylenedioxycatechol 22
was obtained by Fremy’s salt oxidation of sesamol 3 to
the o-quinone 23 followed by reduction with sodium
dithionite in a two-phase ether-water system. The
methylated analogue 24 was also prepared from 3 via
conversion to the MOM ether, ortho-directed metalation
and methylation, Fremy’s salt oxidation to 25, and finally
Na2S2O4 reduction (Scheme 5).
We decided to target the bicyclic catechol 12 since
reduction of the size of the nonaromatic ring in tocophoryl
analogues from the dihydropyran to dihydrofuran is
known to increase the antioxidant capability.16,17 Thus,
2,3-dimethyl-1,4-dihydroxybenzene,13, was monoally-
lated and the resultant ether was subjected to a Claisen
rearrangement at 190 °C in decalin to give 14 (Scheme
3). Treatment with TsOH in refluxing benzene caused
cyclization to 15. o-Formylation and treatment with basic
hydrogen peroxide, as in the formation of 11, afforded a
4:1 mixture of the desired 12 and the corresponding
o-quinone 17 in 51% yield (Scheme 3). The reddish
crystalline quinone 17 was readily reduced to the nearly
colorless catechol 12 upon stirring an ether solution with
BDE, IP , a n d Kin etics. To study the DPPH• kinetics
for the various molecules tested, consider that all the
compounds tested except R-tocopherol have two exchange-
able hydroxyl group H-atoms, and could in principle
undergo reaction with two molecules of DPPH•, compli-
cating the kinetics. However, the pseudo-first order rate
constants are determined by exponential fits to the decay
curves near time zero and we assume that we are
measuring the rate of reaction of only the first (weaker)
OH abstraction, i.e., eq 1. For antioxidant activity, on
the other hand, participation of the second OH group is
a distinct possibility and depends on the BDE. For 1, the
two possibilities are shown in Scheme 6 below, where it
is clear that the weaker bond is the phenolic OH and the
stronger bond is the benzylic OH, since aliphatic alcohols
have BDEs in the range 100-105 kcal mol-1. In this case
the final product would be a (triplet) diradical that does
not have significant stabilization.
(16) Burton, G. W.; LePage, Y.; Gabe, E. J .; Ingold, K. U. J . Am.
Chem. Soc. 1980, 102, 7791-7792.
(17) Burton, G. W.; Hughes, L.; Ingold, K. U. J . Am. Chem. Soc.
1983, 105, 5950-51.
J . Org. Chem, Vol. 68, No. 18, 2003 7025

Hussain et al.
SCHEME 6
now competitive with the first due to the resonance
stabilization in the quinone and BDE2 being only 75.8
kcal mol-1. Thus the semiquinone can be relatively easily
further oxidized since its BDE2 is comparable to the
primary BDE in R-tocopherol.
Compound 11 completes the substitution of the ben-
zene ring, and shows how use of additivity values must
be tempered with caution. Simple additivity predicts that
∆BDE for the first abstraction is -9 (catechol), -2 (o-
methyl), -1.0 (two m-methyl), and -6.0 (p-methoxy),
giving a predicted BDE of 71 kcal mol-1 vs the calculated
73.6 kcal mol-1. This exceeds our usual error from
additivity values because there is a new interaction that
must be considered, which has been discussed previ-
ously14 for a similar case. Figure 2 shows the geometry
of 11 prior to loss of an H-atom. The methoxy group is
forced out-of-plane by almost 90°. This reduces optimal
overlap of the methoxy oxygen relative to its normal
position, rendering this a less electron-donating substitu-
ent; a corrected additivity value for this substituent is
now only -2.6 kcal mol-1, leading to a much better
estimate of the effect of this functional group on the BDE.
The secondary BDE is even lower at 72.6 kcal mol-1, so
the second oxidation (not measured) should be even more
rapid than the first.
Compound 12 was designed to take advantage of the
enhanced planarity of the five-membered ring, which was
shown previously16,17 to improve the activity relative to
R-tocopherol. At the same time introducing the catechol
functionality has been introduced and the phytyl tail
truncated to a single methyl group. This compound has
the very low BDE of 68.7 kcal mol-1, right at the lower
limit of our design window. The second BDE is also very
low at only 71.7, indeed it was necessary to protect this
compound from air or it rapidly autoxidized to form the
red quinone (not characterized).
SCHEME 7
For the catechols, on the other hand, the first oxidation
leads to a semiquinone and the second to a quinone, as
shown in Scheme 7. Here it is not obvious which step
has the lower BDE since catechols are known to be prone
to oxidation, with formation of quinones being common,
even in the presence of the weak oxidant molecular
oxygen.
Results of the kinetic studies are collected in Table 1.
Reproducibility in the kinetic runs was excellent with
very little deviation in the linear fits to the data, as
shown by the high values of R2. Values of the second-
order rate constant kDPPH range over more than 3 orders
of magnitude, with the benzylic compound 1 being least
reactive and 22 and 24 the most reactive, both of the
latter being almost 2 orders of magnitude more reactive
with DPPH• in ethyl acetate than R-tocopherol.
The calculated BDE and IP values are also shown in
Table 1. Compound 1 contains an internal H-bond that
donates a hydrogen atom leaving the phenoxyl radical,
shown in Figure 1. For the radical that forms (at right,
Figure 1), by performing an initial conformer search we
allow the OH group remaining to rotate to form a new
intramolecular H-bond, i.e., the conformer search as-
sumes that there is sufficient energy available at 298 K
to make this rotation possible (the barrier is ca. 3 kcal
mol-1). This “free rotation in the radical” assumption is
made throughout all calculations where the barriers are
expected to be low, as in torsional motion about a single
bond. Compound 1 has a primary BDE of 79.5 kcal mol-1
(for the phenolic O-H) that lies 7.8 kcal mol-1 below
phenol, and is therefore predicted to react more slowly
than R-tocopherol (expt BDE 77 kcal mol-1, calcd BDE
75.0 kcal mol-1), which it does. Additivity values for
substituents14 can also be used to predict the BDE; it
appears that the combined effects are simply that of a
p-methoxy group (-6 kcal mol-1) and an o-methyl group
(-2 kcal mol-1) leading to the observed ∆BDE ) -8 kcal
mol-1. A BDE for the second exchangeable OH was
calculated to be above 100 kcal mol-1 and therefore would
not participate in any exchange reaction.
Compound 21 is similar to 12. It is R-tocopherol with
a catechol group replacing the o-methylphenol group in
R-tocopherol. We refer to this molecule as “Super-E” since
it closely resembles the original molecule but now has
low primary and secondary BDE values of 69.3 and 70.5
kcal mol-1, respectively. This compound is also prone to
air oxidation. Compound 22 is similar to 21 since it lies
at the lower edge of the design window with a BDE1 value
of 69.0 kcal mol-1. Note that in this case the BDE2 value
falls below the window at 67.5 kcal mol-1. Finally, 24
lies slightly below our design window with both BDEs
at ca. 67 kcal mol-1; regeneration by ascorbate will be
problematic in this case. Also, at some point when the
BDE is low enough the compound will become a pro-
oxidant via direct reaction with oxygen, and it will be of
interest to see when this behavior is reached. To study
this problem in a biological context, it would be of interest
to extend our calculations to include free energy changes
in the presence of the appropriate concentration of
oxygen, as well as solvation effects. One possible reaction
is direct hydrogenation, according to ArOH + O2 f ArO•
Compound 2 is the least-substituted member of the
catechol series. The weaker OH bond is para to the
methoxy group, corresponding to BDE1. From additivity
we obtain -9 for the catechol group, -0.5 for m-methyl,
and -6 for p-methoxy, giving a predicted ∆BDE ) -15.5
kcal mol-1, in reasonable agreement with the observed
∆BDE ) -14.6 kcal mol-1 (using 87.10 - 72.5). The
second oxidation will now produce the corressponding
quinone (as in Scheme 7). Loss of the second H-atom is
•
+ HO2 , but there are other variations including electron
transfer from the corresponding anion ArO- to oxygen
to generate superoxide O2-; this process will depend on
the pK of the acid.
Table 1 also shows the IP values relative to phenol.
From prior work we know that R-tocopherol normally
reacts by H-atom transfer and not electron transfer, i.e
7026 J . Org. Chem., Vol. 68, No. 18, 2003

Development of Novel Antioxidants
TABLE 1. Ra te Con sta n t for DP P H• + Com p ou n d in Eth yl Aceta te Solven t, Ca lcu la ted Bon d Dissocia tion En th a lp y of
th e F ir st (BDE1) a n d th e Secon d (BDE2) Exch a n gea ble (OH) Hyd r ogen Atom , a n d Ca lcu la ted Ion iza tion P oten tia l
Rela tive to P h en ol (∆IP )
BDE1 calcd
BDE2 calcda
∆IP calcdb
compd
k (M-1 s-1
)
R2
(kcal mol-1
)
(kcal mol-1
)
(kcal mol-1
)
1
2
2.9
200
210
0.9942
0.9995
0.9993
0.9999
0.9980
0.9988
0.9984
0.9987
0.9992d
79.5
72.5
73.6
68.7
69.3
69.0
66.9
81.0
75.0e
111.0
75.8
72.6
71.7
70.5
67.5
67.3
-22.1
-25.5
-24.9
-33.7
-36.0
-28.7
-31.3
ca. -22
-36.2
11
12
21
22
24
3000
4500
5040
8870
1.1
p-methoxyphenol
R-tocopherol
160d
a
BDE for loss of the second H-atom after the first has already been lost. For a catechol, BDE2 would be the loss of an H-atom from the
semiquinone radical. Relative to phenol in the same basis set, for which we obtain 184.06 kcal mol-1
.
c In this case BDE2 corresponds
b
to forming a triplet product, since quinone formation is not possible. Literature value from ref 19. e The C16H33 (phytyl tail) was replaced
d
by a methyl group in the calculation.
F IGURE 1. (a) Parent benzyl alcohol (structure 1) showing
starting conformation. (b) Radical formed after first H-atom
abstraction, showing benzylic OH rotated toward the phenoxyl
radical center.
F IGURE 3. Correlation between calculated gas-phase BDEs
and measured solution-phase BDEs (see ref 14 for data) for
substituted phenols.
F IGURE 2. Structure of the fully substituted catechol 11,
showing that the methoxy group is rotated 90° out of plane
due to interaction with adjacent methyl groups.
from gas to benzene solution, and for the use of locally
dense instead of full basis sets, there should nevertheless
the mechanism ArOH f [ArOH]+ + e- to form the radical
be a good correlation between gas- and solution-phase
BDEs. In Figure 3 we have plotted the BDE (calculated,
gas) vs BDE (experimental, benzene) for a set of 13
substituted phenols, omitting only the o,o-di-tert-bu-
tylphenol case for which the B3LYP functional fails.14
Given that the experimental errors are between 0.12 and
0.8 kcal mol-1 for this data set and that the data span a
range over 10 kcal mol-1 with a standard deviation of
only ca. 0.5 kcal mol-1, it can be seen that the theoretical
calculation compares well with this highly precise set of
experimental measurements. In extending the calculation
to more complex substituted phenols it is hard to be
absolutely sure that there are no anomalies, but we
believe from this and other examples that the calculated
cation is not observed. For R-tocopherol the IP lies 36.0
kcal mol-1 below phenol in our calculation. Other com-
pounds such as aromatic amines which are prone to
formation of radical cations in solution have ∆IP values
at least 40 kcal mol-1 below those of phenol.14 Since none
of the compounds in Table 1 are more easily ionized than
R-tocopherol, we do not expect them to react by an
electron-transfer mechanism.
Accu r a cy of BDE a n d IP Ca lcu la tion s. In ref 14
and elsewhere we have discussed the accuracy of the
calculated BDE data using the (RO)B3LYP/6-311+G(2d,-
2p)//AM1/AM1 method. Our BDE calculations for phenol
and substituted phenols can be compared to accurate
solution-phase (benzene) values obtained by Pedulli and
co-workers.15b Allowing for a slight shift due to the change
∆BDEs are accurate to within ca. 1.0 kcal mol-1
.
J . Org. Chem, Vol. 68, No. 18, 2003 7027

Hussain et al.
F IGURE 5. Plot of log kDPPH vs change in IP relative to
phenol. The ∆IP axis has reversed sign (see Table 1).
F IGURE 4. Plot of log kDPPH vs change in BDE relative to
phenol (87.1 kcal mol-1). The ∆BDE axis has reversed sign,
e.g. p-methoxyphenol has BDE ) -6.1 kcal mol-1, log kDPPH
)
1.1 M-1 s-1
.
paper are comparable to the most reactive molecules
described by Foti et al., e.g. their 4-methoxy 1,8-
naphthalene diol, which had a calculated BDE value of
67.5 kcal mol-1, very close to our compound 24 with BDE
) 66.9 kcal mol-1. Foti et al.6 also looked at kinetics of
reactions with peroxyl radicals and obtained a good linear
correlation with BDE values; this adds confidence that
tests with the nitrogen radical DPPH• will mimic behav-
ior in systems which contain peroxyl radicals ROO•. The
latter are generated biologically via normal metabolism
and are known to play a role in lipid peroxidation1.
Cor r ela tion betw een DP P H• Kin etics a n d IP . As
described above, we believe that the kinetics of reaction
between DPPH• radicals and the catechols in ethyl
acetate should proceed by H-atom transfer rather than
by single-electron transfer to form the radical cation, i.e.,
the reaction DPPH• + ArOH f DPPH2 + [ArOH]+. Since
the radical cation [ArOH]+ rapidly equilibrates to lose a
proton and generate the phenoxyl radical, it can be
difficult to distinguish between the hydrogen-atom and
single electron-transfer mechanisms. One instructive
approach to try to distinguish between mechanisms is
to plot the ionization potential relative to phenol, ∆IP,
vs log kDPPH and examine the quality of the correlation.
This graph is shown in Figure 5. It is clear from the large
scatter that the BDE shows by far the better correlation
than the IP (e.g. R2 values of 0.98 or 0.52 for BDE and
IP, respectively), consistent with our argument that the
reaction proceeds by H-atom transfer.
The IP calculations were monitored in an extensive
series of comparisons on phenol with a variety of electron-
donating and electron-withdrawing substituents. By us-
ing the method described below, values of the IP relative
to phenol with the same method of calculation are
expected to provide a set of ∆IP values accurate to ca. 2
kcal mol-1
.
Cor r ela tion betw een DP P H• Kin etics a n d BDE.
Results from the present experiment are shown in Figure
4. The data now extend over nearly 4 orders of magnitude
in kDPPH and thus provide a stringent test of using BDEs
to correlate rate data. There is a very good linear
correlation between log kDPPH and ∆BDE , with the R2
value being 0.978 for nine data points. This correlation
includes an additional value we determined for p-meth-
oxyphenol (see Table 1), which allowed us to extend the
correlation over a wider range in BDE. The slope in ethyl
acetate is somewhat less than that for the Foti correlation
in heptane,6 being 0.284 log unit per kcal mol-1 in BDE,
whereas Foti et al. obtained 0.352 log unit per kcal mol-1
in BDE. Note that the rates are also much lower in ethyl
acetate than in heptane due to hindering of the H-atom
abstraction by H-bonding in the more polar solvent.18 An
example that relates the two series of data is p-methox-
yphenol, which has a rate constant of 240 in heptane but
only 1.1 in ethyl acetate.
The quality of the correlations between the present
data and the Foti data, which span a somewhat wider
range in BDE (ca. 30 kcal mol-1), is also similar. The
molecules containing the lowest BDEs in the present
It should be noted that these conclusions apply to the
relatively nonpolar solvent ethyl acetate, in which the
DPPH measurements were made in this paper. In water,
one would have to examine whether the mechanism may
involve electron transfer from an anion, that originating
(18) Valgimigli, L.; Banks, J . T.; Ingold, K. U.; Lusztyk, J . J . Am.
Chem. Soc. 1995, 117, 9966-9971.
7028 J . Org. Chem., Vol. 68, No. 18, 2003

Development of Novel Antioxidants
from either DPPH or from the catechols,24 which have
pK values near 10. It is generally the case that when an
anionic form is present, as in the case of ascorbate where
at physiological pH 7.4 the ascorbate anion is predomi-
nant, then we expect electron transfer to be the most
important mechanism of reduction of a free radical. Even
in that extreme case, however, H-atom transfer is a
competing mechanism and cannot be disregarded.
Ca tech ol Toxicity. Catechols are known to exert toxic
effects in cells, mainly via two mechanisms.22 In the first,
they become oxidized to quinones which then act as
Michael acceptors. The nucleophiles may be proteins,
DNA, or glutathione (especially as the glutathione anion
GS-). Glutathione depletion in particular leads to a
change in the redox environment of the cell, generally
with negative consequences.23 In the second mechanism,
the quinones are reduced to semiquinones by reducing
agents or enzymes in the cell. These semiquinones then
redox cycle to the quinone and back to semiquinone,
reducing oxygen to superoxide in the process, and hence
the system acts as a free radical generator. This mech-
anism can lead to oxidative stress in cells, causing lipid
peroxidation or other oxidative damage.
The rate of reaction of quinones as Michael acceptors
is variable. It depends on whether the â-carbon position
is unhindered sterically, and also whether the ring
contains electron donors near the â-carbon. The latter
effect reduces the electrophilicity of the carbon under
attack and may prevent reaction altogether. Of the
compounds we have described only 2 should be a strong
Michael acceptor. The remainder are either hindered
sterically or have strong electron donors at this position,
and hence should be relatively unreactive.
cases BDE2 > BDE1, suggesting that these compounds
should not be particularly active at redox cycling. In
menadiol, by comparison, whose corresponding quinone
is thought to induce oxidative stress by redox cycling, we
have calculated10 that the second BDE is much lower
than the first, consistent with the observed toxicity. Note,
however, that menadione also has an open ring position
that can undergo nucleophilic attack, so toxicity can also
result from glutathione depletion and the changing redox
status of the cell.
Con clu sion s
In this paper we discussed a design strategy to create
antioxidants which could be potentially useful for biologi-
cal purposes, and decided on a “design window” in the
range 68-75 kcal mol-1. On the basis of this strategy
we examined a series of target compounds containing a
weak O-H bond, with most of the molecules within the
design window and some outside it. We report the
synthesis of seven novel compounds, including a benzylic
phenol and six catechols. BDE values were calculated for
loss of the first and second H-atom in these compounds
by using density functional techniques with the B3LYP
functional, as well as ionization potentials, and we believe
these calculations to be accurate to within ca. 2 kcal
mol-1. Primary BDE values were in the range 67-81 kcal
mol-1, with secondary BDEs ranging from 67.5 to >100
kcal mol-1. The compounds were then tested for reactivity
with the free radical DPPH•. A very good linear correla-
tion was obtained between log kDPPH• and the BDE,
showing that the BDE can be used to predict reactivity
in this series. On the other hand, a poor correlation was
obtained between log kDPPH• and the IP, showing that the
ionization mechanism is not active in this reaction, which
was carried out in ethyl acetate. Aspects of catechol
toxicity were discussed, which are relevant to biological
experiments. Because of the high reactivity with DPPH•
the compounds are potentially interesting in biological
applications, and a testing program is now under way to
determine toxicity and protective effects in cell cultures
using these novel synthetic antioxidants.
The question of redox cycling and the “tuning” of the
catecholic antioxidants to retard the rate of reaction with
molecular oxygen remains an open question, which also
relates to when an antioxidant becomes a pro-oxidant.
In general the catechols are thought to act as an anti-
oxidant by transfer of the first H-atom to a free radical
(BDE1, Table 1) but as a pro-oxidant by transfer of a
second H-atom from the semiquinone (BDE2, Table 1)
with subsequent redox cycling between quinone and
semiquinone and the resulting creation of oxidative stress
via generation of superoxide ion.1 On the other hand,
Table 1 shows that the catechols in this paper are
generally comparable in BDE1 and BDE2, in fact in most
Exp er im en ta l Section
Ca lcu la tion of BDE. For the calculation of the BDE
according to the LLM method, we used the AM1 semiempirical
method to obtain the optimized geometry and frequencies of
parent and radical (BDE1), or radical and quinone (BDE2),
where T ) 298.15 K, P ) 1.00 atm, and the AM1 frequencies
are scaled by the factor 0.973. At the geometry minimum a
single-point calculation was done with (RO)B3LYP/6-311+G-
(2d,2p), where RO indicates that for the radical the restricted
open-shell B3LYP method was used. All electronic energies
are then corrected by the thermal contribution to the enthalpy
to obtain H°298, the standard gas-phase enthalpy at 298 K. To
complete the specification of the method, we set the electronic
energy of the H-atom to its exact value of -0.50000 hartree,
and obtain its enthalpy H°298 ) -0.50000 + 5/2RT ) -0.49764
hartree. Starting geometries were generally obtained with the
PC-Spartan20 builder module, using AM1; coordinates were
then sent to the Gaussian 98 program21 for all subsequent
calculations.
(19) Valgimigli, L.; Banks, J . T.; Lusztyk, J .; Ingold, K. U. J . Org.
Chem. 1999, 64, 3381-3383.
(20) Spartan ’02 for Windows, Wavefunction, Inc.: Irvine, CA, 2002.
(21) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J . A. Gaussian 98, Revision A.7, Gaussian, Inc., Pittsburgh PA,
1998.
(22) Bolton, J . L.; Trush, M. A.; Penning, T. M.; Dryhurst, G.; Monks,
T. J . Chem. Res. Toxicol. 2000, 13, 135-160.
Ca lcu la tion of IP . IPs are only reported for parent
hydrocarbons (i.e. the first ionization) and not for the radicals
which can form in the catechols. Geometries and frequencies
were calculated and scaled with use of AM1 as above. To obtain
(23) Nelson, S. D.; Pearson, P. G. Annu. Rev. Pharmacol. Toxicol.
1990, 30, 169-195.
(24) K. U. Ingold, personal communication.
J . Org. Chem, Vol. 68, No. 18, 2003 7029

Hussain et al.
the adiabatic ionization potential we took the optimized
geometry for parent and (radical) cation and performed a (RO)-
B3LYP/6-31G(d) calculation, as described in ref 25. The
difference between parent and cation at 0 K corresponds to
∆E°0 for ArOH (g) f [ArOH]+ (g) + e-. Reported values are
relative to the calculated value for phenol, for which we obtain
∆E°0 ) 184.06 kcal mol-1. This lies well below the experimen-
tal value (see ref 25 for an extensive list of experimental data)
of 196.2 ( 0.02 kcal mol-1; however, we found previously25 that
there is a systematic error arising from the treatment of the
benzene ring, and that use of ∆IP values for substituted
phenols caused a large error cancellation, resulting in ∆IP
tion (10 mL), extraction with ethyl acetate (3 × 20 mL), and
drying and evaporating the organic extracts. The crude
product, 5.29 g of brown oil, was purified via gradient flash
chromatography (125 g of silica gel, hexanes-ethyl acetate
eluent, polarity increase by 5%). The yield of 4, a white solid
1
(mp 39-41 °C), was 3.60 g (72%). H NMR (CDCl3, 300 MHz)
δ 2.12 (s, 3H, methyl), 3.76 (s, 3H, methoxy), 5.89 (s, 2H,
methylenedioxy), 6.25 (d, J ) 8.4 Hz, 1H, H-6), 6.58 (d, J )
8.4 Hz, 1H, H-7); 13C NMR (CDCl3, 300 MHz) δ 153.88 (C-5),
147.07 (C-3), 141.42 (C-1), 109.67 (C-4), 104.62 (C-7), 102.12
(C-6), 101.24 (C-2), 56.46 (methoxy), 9.13 (methyl); MS [EI,
m/z (%)] 166 [MH+] (98), 151 (100), 121 (47), 93 (16), 65 (17);
HRMS calcd for C9H10O3 166.0630, found 166.0626.
values accurate to ca. 2 kcal mol-1
.
DP P H• Kin etics. The stopped-flow apparatus is an Applied
Photophysics SX 18 MV spectrometer with a xenon 150 arc
light source. DPPH• was obtained from Northern Sources, Inc.,
and was used without further purification. All compounds were
dissolved in ether, washed with aqueous sodium dithionite,
and then re-isolated prior to the kinetic measurements. In a
typical run, DPPH• (ca. 5 × 10-5 M) was deoxygenated under
nitrogen, and then mixed 1:1 with (deoxygenated) solutions
of antioxidants, where the antioxidant ArOH concentration
was usually 2 orders of magnitude greater, so as to obtain
pseudo-first-order rate constants. The decay of DPPH• was
monitored at 519 nm in the presence of a given (large excess)
concentration of antioxidant at room temperature. Under these
conditions the decay curve is pseudo-first order, and it was
well fitted by a single-exponential function [DPPH] ) [DPPH]0
exp(-kobst) + constant. From plots of kobs vs [ArOH] at five
concentrations the second-order rate constant was obtained
as the slope of the plot.
Syn th eses: 2-Hyd r oxym eth yl-4-m eth oxyp h en ol (1). A
solution of 2-hydroxy-5-methoxybenzaldehyde (1.00 g, 66
mmol) and sodium borohydride (0.201 g, 53 mmol) in 30 mL
of THF was stirred for 3 h at 0 °C. Aqueous sodium hydroxide
(10%, 10 mL) was added and the mixture extracted with ethyl
acetate (3 × 20 mL). The combined organic portions were dried
over magnesium sulfate, filtered, and concentrated in vacuo.
The crude product was purified by gradient flash chromatog-
raphy (30 g of silica gel, ethyl acetate-hexane eluent, 5%
polarity increase) to afford 0.671 g, 66% of 1 as a white solid:
mp 77-79 °C.
4-Meth oxyben zo[1,3]d ioxole. Anhydrous potassium car-
bonate (35.0 g, 0.253 mol) was added to the solution of sesamol,
3 (5.01 g, 36 mmol), in 80 mL of acetone at room temperature
under a nitrogen atmosphere. The mixture was stirred vigor-
ously for 1 h. Methyl iodide (24.8 mL, 40 mmol) was added
dropwise and the reaction mixture was refluxed at 55 °C for
42 h. The resulting slurry was filtered through Celite and the
filtrate evaporated in vacuo. The residue was taken up in 40
mL of water and extracted with ether (3 × 30 mL). The
combined ether fractions were washed with 25 mL of 10%
sodium hydroxide solution, followed by 2 × 25 mL of water,
dried, filtered and concentrated in vacuo to afford anisole (4.61
g, 84%) of 5 methoxybenzo[1,3]dioxole as a yellow liquid. This
material was used for the next step without further purifica-
tion. 1H NMR (CDCl3, 200 MHz) δ 3.73 (s, 3H, methoxy), 5.89
(s, 2H, methylene dioxy), 6.31 (dd, J ) 8.5, 2.6 Hz, 1H, H-6),
6.47 (d, J ) 2.6 Hz, 1H, H-4), 6.69 (d, J ) 8.5 Hz, 1H, H-7);
MS [EI, m/z (%)] 152 [MH+] (100), 137 (74), 107 (32), 79 (31),
51 (17); HRMS calcd for C8H8O3 152.0473, found 152.0477.
5-Meth oxy-4-m eth ylben zo[1,3]d ioxole (4). A 1.7 M solu-
tion of tert-butyllithium in pentane (27.0 mL, 0.046 mmol) was
added dropwise to a solution of 5-methoxybenzo[1,3]dioxole
(4.60 g, 30 mmol) in dry THF (20 mL), under nitrogen at 0 °C.
The mixture was stirred for 30 min and methyl iodide (3.80
mL, 61 mmol) was added dropwise. The reaction mixture was
kept at 0 °C for 1 h and an additional 2 h at rt. Workup
involved quenching with saturated ammonium chloride solu-
Acetic Acid 5-Meth oxy-4-m eth ylben zo[1,3]d ioxol-2-yl
Ester (5). Lead tetraacetate (14.4 g, 32.5 mmol), recrystallized
from glacial acetic acid, was added to a solution of 5-methoxy-
4-methylbenzo[1,3]dioxole (4, 3.60 g, 21.7 mmol) in benzene
(60 mL), under nitrogen, and stirred at reflux (75-80 °C) for
20 h. The reaction mixture was cooled, diluted with 30 mL of
water, and extracted with ether (3 × 25 mL). The combined
organic extracts were dried over Mg2SO4, filtered, and con-
centrated in vacuo to afford a brown oil (4.48 g, 92% crude).
This material was reacted as described below without further
purification. 1H NMR (CDCl3, 200 MHz) δ 2.09 (s, 3H, acetate
methyl), 2.12 (s, 3H, methyl), 3.77 (s, 3H, methoxy), 6.36 (d, J
) 8.0 Hz, 1H, H-6), 6.70 (d, J ) 8.0 Hz, 1H, H-7), 7.64 (s, 1H,
H-2); 13C NMR (CDCl3, 300 MHz) δ 170.16 (acetate -CdO),
151.51, 146.55, 142.79, 113.96, 108.76, 103.14, 102.10, 56.79
(methoxy), 22.32, 9.09 (methyl).
3-Met h yl-4-m et h oxy-1,2-ben zoq u in on e (6), 3-Met h yl-
4-m eth oxy-1.2-d ih yd r oxyben zen e (2), a n d 5-Meth oxy-4-
m eth ylben zo[1,3]d ioxol-2-on e (7). Crude acetate from above
(4.48 g, 20 mmol) was dissolved in 80% aq acetic acid (100
mL) and stirred at room temperature for 48 h. The solvent
was evaporated and the remaining residue washed with
toluene (4 × 1 mL) and condensed to dryness in vacuo.
Gradient flash chromatography (125 g silica gel, hexanes-
ethyl acetate eluent with 5% polarity increase) afforded a
mixture of 2 and 6 (2.62 g, 85%) as a red liquid and 7 (0.540
g, 15%) as an off-white solid: mp 110-111 °C. For 7: 1H NMR
(CDCl3, 300 MHz) δ 2.17 (s, 3H, methyl), 3.80 (s, 3H, methoxy),
6.59 (d, J ) 8.8 Hz, 1H, H-6), 6.93 (d, J ) 8.8 Hz, 1H, H-7);
13C NMR (CDCl3, 300 MHz) δ 155.5 (C-2), 152.3 (C-5), 143.0
(C-1), 137.3 (C-3), 111.3 (C-4), 107.1 (C-7), 105.7 (C-6), 56.6
(methoxy), 9.2 (methyl); MS [EI, m/z (%)] 180 [MH+] (100), 135
(12), 121 (84), 93 (44), 65 (17); HRMS calcd for C9H8O4
180.0422, found 180.0392.
3-Meth yl-4-m eth oxy-1,2-d ih yd r oxyben zen e (2). A solu-
tion of sodium dithionite (20.75 g, 11.9 mmol) in 75 mL of
water was added to a rapidly stirred solution (2.62 g, 17.2
mmol) of the o-quinone 6/catechol 2 mixture from above in 25
mL of ether. The color of the reaction mixture changed from
orange to bright yellow within 3 min. Aqueous 10% hydro-
chloric acid (10 mL) was added and the reaction mixture was
extracted with ether (4 × 20 mL). The combined ether extracts
were washed with saturated sodium bicarbonate (2 × 10 mL),
dried (MgSO4), filtered, and concentrated to give 2 (2.39 g,
1
91%) as a yellowish oil. H NMR (CDCl3, 300 MHz) δ 2.12 (s,
3H, methyl), 3.74 (s, 3H, methoxy), 5.83 (br s, 2H, hydroxyl),
6.28 (d, J ) 8.7 Hz, 1H, H-5), 6.60 (d, J ) 8.7 Hz, 1H, H-6);
13C NMR (CDCl3, 300 MHz) δ 152.8 (C-4), 143.5 (C-1), 137.7
(C-2), 114.2 (C-3), 112.4 (C-6), 103.0 (C-5), 56.7 (methoxy), 8.9
(methyl); MS [EI, m/z (%)] 154 [MH+] (91), 139 (100), 121 (38),
93 (15), 65 (22); HRMS calcd for C8H10O3 154.0630, found
154.0612.
4-ter t-Bu tyldim eth ylsilyloxy-2,3,5-tr im eth ylph en ol. tert-
Butyldimethylsilyl chloride (1.19 g, 1.2 mol equiv) in dry DMF
(5 mL) was added to a solution of 2,3,5-trimethylhydroquinone
(8, 1.0 g, 1.0 mol) and imidazole (1.79 g, 4.0 mol equiv) in dry
DMF (10 mL) at -20 °C under nitrogen. The reaction mixture
was warmed to 20 °C and stirred for 2 h, poured into water,
and extracted with CH2Cl2 (3 × 20 mL). The combined organic
(25) DiLabio, G. A.; Pratt, D. A.; Wright, J . S. Chem. Phys. Lett.
1999, 311, 215-220.
7030 J . Org. Chem., Vol. 68, No. 18, 2003

Development of Novel Antioxidants
layers were washed with water (2 × 20 mL) and dried (MgSO4)
and the solvent was removed under vacuo. The crude product
was subjected to a flash chromatography with use of ethyl
acetate and hexane to give the title compound as a solid (1.5
g, 85%). 1H NMR (CDCl3, 200 MHz) δ 0.23 (s, 6H), 1.07 (s,
9H), 2.16 (s, 3H), 2.19 (s, 3H), 2.21 (s, 3H), 4.46 (s, 1H), 4.50
(s, 1H); 13C NMR (CDCl3, 50 MHz) δ 146.9, 143.5, 125.9, 123.8,
120.5, 118.1, 25.9, 18.3, 16.1, 13.0, 12.4, -4.2; HRMS calcd
for C15H26O2Si 266.1702, found 266.1703.
washed with hexane (20 mL). The yield of 14, a white solid
(mp 124-126 °C), was 0.81 g (81%). 1H NMR (CDCl3, 200 MHz)
δ 2.14 (s, 3H), 2.16 (s, 3H), 3.31 (br d, J ) 6.2 Hz, 2H), 4.44 (s,
1H), 4.61 (s, 1H), 5.15 (m, 2H), 5.97 (m, 1H), 6.41 (s, 1H); 13C
NMR (CDCl3, 125 MHz) δ 147.1, 146.2, 136.4, 124.7, 122.5,
122.0, 116.5, 116.0, 113.7, 112.5, 35.5, 12.2, 11.9; HRMS calcd
for C11H14O2 178.0994, found 178.0994.
2,6,7-Tr im eth yl-2,3-d ih yd r oben zofu r a n -5-ol (15). p-Tol-
uenesulfonic acid (3.5 g) was added to a solution of 5-allyl-
2,3-dimethylbenzene-1,4-diol 14 (3.0 g, 16.8 mmol) in benzene
(40 mL) and the reaction mixture was refluxed for 8 h and
then cooled to room temperature. Workup involved addition
of saturated NaHCO3 (20 mL), extraction with ethyl acetate
(3 × 20 mL), drying (MgSO4), and evaporating the solvents.
The solution was filtered and condensed in vacuo to give a
solid. Flash chromatography of the crude product with ethyl
acetate-hexane eluent afforded 15 (2.1 g, 69%) as a white
solid: mp (126-128 °C). 1H NMR (CDCl3, 200 MHz) δ 1.43 (d,
J ) 6.2 Hz, 3H), 2.11 (s, 3H), 2.13 (s, 3H), 2.74 (dd, J ) 14.4,
7.0 Hz, 1H), 3.21 (dd, J ) 15.0, 8.8 Hz, 1H), 4.33 (s, 1H), 4.82
(m, 1H), 6.49 (s, 1H); 13C NMR(CDCl3, 50 MHz) δ 152.1, 147.2,
123.5, 121.9, 118.9, of 109.1, 78.9, 37.8, 21.7, 12.3, 11.8; HRMS
calcd for C11H14O2 178.0994, found 178.0994.
4-ter t-Bu tyld im eth ylsilyloxy-1-m eth oxy-2,3,5-tr im eth -
ylben zen e. A mixture of the above phenol (0.76 g, 1.0 mol
equiv), methyl iodide (0.9 mL, 5.0 mol equiv), and potassium
carbonate (0.79 g, 2.0 mol equiv) in 10 mL of acetone was
refluxed for 24 h. The solvent was evaporated and water (20
mL) was added to the residue. The resulting mixture was
extracted with ethyl acetate (3 × 20 mL). Flash chromatog-
raphy (ethyl acetate-hexane eluent) of the crude product
1
yielded 0.76 g (95%) of the desired product as a yellow oil. H
NMR (CDCl3, 50 MHz) δ 0.18 (s, 6H), 1.00 (s, 9H), 2.08 (s,
3H), 2.16 (s, 3H), 2.20 (s, 3H), 3.63 (s, 3H), 6.42 (s, 1H); 13C
NMR(CDCl3, 50 MHz) δ 150.8, 149.2, 130.4, 127.7, 125.8,
118.1, 60.0, 25.8, 18.2, 16.0, 12.8, -4.3; HRMS calcd for
C
16H28O2Si 280.1860, found 280.1859.
4-Meth oxy-2,3,5-tr im eth ylp h en ol (9). A solution of the
5-H yd r oxy-2,6,7-t r im et h yl-2,3-d ih yd r ob en zofu r a n -4-
ca r ba ld eh yd e (16). A solution of hydroxybenzofuran 15 (3.5
g, 19.6 mmol), tin tetrachloride (0.28 mL, 0.12 mol equiv), and
tributylamine (1.87 mL, 0.4 mol equiv) in 40 mL of anhydrous
toluene in a three-neck flask was stirred at room temperature
for 20 min and then paraformaldehyde (1.3 g, 2.2 mol equiv)
was added. The resulting yellowish solution was heated for 2
h at 90-95 °C, cooled to room temperature, poured into water
(50 mL), acidified to pH 2 with 2 N HCl, and extracted with
ether (3 × 50 mL). The combined organic layers were washed
with water (50 mL), dried over MgSO4, and filtered and the
solvent was removed in vacuo. The crude residue was purified
by flash chromatography with ethyl acetate and hexane to give
a bright yellow solid 16 (2.1 g, 52%). 1H NMR (CDCl3, 200
MHz) δ 1.47 (d, J ) 6.2 Hz, 3H), 2.09 (s, 3H), 2.15 (s, 3H),
2.98 (dd, J ) 15.8, 8.0 Hz, 1H), 3.51 (dd, J ) 15.6, 8.6 Hz,
1H), 4.91 (m, 1H), 9.84 (s, 1H), 11.06 (s, 1H); 13C NMR (CDCl3,
50 MHz) δ: 173.9, 154.3, 151.3, 133.4, 130.2, 124.1, 114.2, 79.3,
35.5, 21.7, 13.3, 10.7; HRMS calcd for C12H14O3 206.0943, found
206.0943.
above silylated phenol (0.8 g, 1.0 mol equiv) in dry THF (8
mL) was treated with Bu4NF (4.27 mL, 1.0 M, 1.5 mol equiv)
at room temperature under nitrogen for 1 h. The reaction
mixture was quenched with 10 mL of water, extracted with
ethyl acetate, and processed in the usual manner. Flash
chromatography of the crude product [ethyl acetate and
hexane] gave the pure 9 (0.45 g, 96%). 1H NMR (CDCl3, 50
MHz) : 2.11 (s, 3H), 2.19 (s, 6H), 3.65 (s, 3H), 4.98 (br s, 1H),
6.42 (s, 1H); 13C NMR (CDCl3, 50 MHz) δ 150.1, 149.7, 130.6,
128.2, 121.3, 114.5, 60.2, 15.9, 12.7, 11.9; HRMS calcd for
C
10H14O2 166.0994, found 166.0996.
2-Hydr oxy-5-m eth oxy-3,4,6-tr im eth ylben zaldeh yde (10).
The same formylation procedure as for the preparation of 16
1
(with 15 as a substrate) gave 57% of 10 as a yellow solid. H
NMR (CDCl3, 50 MHz) δ 2.07 (s, 3H), 2.20 (s, 3H), 2.44 (s,
3H), 3.58 (s, 3H), 10.14 (s, 1H), 12.13 (s, 1H); 13C NMR(CDCl3,
200 MHz) δ 194.5, 157.8, 148.9, 141.7, 129.9, 123.8, 116.1, 60.5,
13.7, 10.9, 10.1; HRMS calcd for C11H14O3 194.0943, found
194.0953.
4-Met h oxy-3,4,6-t r im et h ylb en zen e-1,2-d iol (11). The
same procedure as for the preparation of 12 (with 16 as a
substrate) gave 11 (62%) as a pinkish solid: mp 114-115 °C.
1H NMR (CDCl3, 500 MHz) δ 2.13 (s, 6H), 2.16 (s, 3H), 3.62
(s, 3H), 4.81 (s, 1H), 4.93 (s, 1H); 13C NMR (CDCl3, 500 MHz)
δ 150.2, 140.0, 138.1, 121.3, 120.5, 114.4, 60.4, 12.0, 11.9, 8.9;
HRMS calcd for C10H14O3 182.0943, found 182.0943.
2,6,7-Tr im eth yl-2,3-d ih yd r o-ben zofu r a n -4,5-d ion e (11)
a n d 2,6,7-Tr im eth yl-2,3-d ih yd r oben zofu r a n -4,5-d iol (12).
Hydrogen peroxide (0.2 mL, 30%) was added to a rapidly
stirred solution of aldehyde 16 (1.0 g, 4.9 mmol) and sodium
hydroxide (0.19 g, 1.0 mol equiv) in 16 mL of water-THF (9:
1) solvent mixture, under nitrogen, at 28-30 °C. Aqueous 10%
HCl was added after the solution had been stirred for 30 min
at that temperature followed by 30 min at room temperature.
The reaction mixture was extracted with ethyl acetate (3 ×
30 mL). Further usual workup gave a crude product that
consisted mainly of the desired catechol 12 and the related
quinone 17. Flash chromatography with ethyl acetate and
hexane as eluent gave 17 (0.1 g, 11%) as a deep red liquid
and the diol 12 (0.38 g, 40%) as a pink solid. For 17: 1H NMR
(CDCl3, 200 MHz) δ 1.45 (d, J ) 6.4 Hz, 3H), 1.89 (s, 3H),
2.00 (s, 3H), 2.54 (dd, J ) 15.0, 7.2 Hz, 1H), 3.09 (dd, J )
15.0, 10.0 Hz, 1H), 5.04 (m, 1H); 13C NMR (CDCl3, 50 MHz) δ
182.3, 174.6, 171.6, 137.0, 136.8, 111.7, 84.0, 33.0, 21.9, 13.4,
11.8; HRMS calcd for C11H12O3 192.0787, found 192.0789. For
4-Allyloxy-2,3-d im eth ylp h en ol a n d 1,4-Bis-a llyloxy-2,3-
d im eth ylben zen e. Allyl bromide (0.94 mL, 11 mmol) and
potassium carbonate (2.02 g, 14.6 mmol) were added to a
solution of 2,3-dimethylhydroxyhydroquinone (1.0 g, 7.3 mmol)
in acetone (10 mL). The resulting mixture was stirred at reflux
for 18 h, the solvent was removed in vacuo, and the remaining
residue was taken up in water and extracted with ether (3 ×
30 mL). Further workup followed by flash chromatography of
the crude product with ethyl acetate and hexane gave two
products: the desired monoallyl ether (0.45 g, 35%) as a brown
semisolid and diallyl ether (0.79 g, 50%) as a yellow oil: For
the monoallyl ether: 1H NMR (CDCl3, 200 MHz) δ 2.18 (s, 3H),
2.20 (s, 3H), 4.46 (dt, J ) 5.2, 1.6 Hz, 2H), 5.02 (br s, 1H),
5.23 (br dd, J ) 10.4, 1.6 Hz, 1H), 5.46 (br dd, J ) 17.2, 1.6
Hz, 1H), 6.07 (m, 1H), 6.55 (d, J ) 9.0 Hz, 1H), 6.61 (d, J )
9.0 Hz, 1H); 13C NMR (CDCl3, 50 MHz) δ 150.7, 147.8, 134.0,
127.4, 124.3, 116.9, 112.0, 110.7, 70.2, 12.3, 12.2; HRMS calcd
for C11H14O2 178.0994, found 178.0994:
1
12: mp 140-142 °C; H NMR (acetone-d6, 200 MHz) δ 1.35
(d, J ) 6.2 Hz, 3H), 1.96 (s, 3H), 2.07 (s, 3H), 2.67 (dd, J )
15.0, 6.8 Hz, 1H), 3.22 (dd, J ) 15.0, 8.4 Hz, 1H), 4.76 (m,
1H), 6.50 (s, 1H), 7.51 (s, 1H); 13C NMR (acetone-d6, 50 MHz)
δ 152.5, 140.5, 137.5, 123.4, 110.2, 108.9, 79.4, 35.9, 22.0, 12.2,
11.8; HRMS calcd for C11H14O3 194.0943, found 194.0943.
6-Allyl-2,3-d im eth ylben zen e-1,4-d iol (14). A solution of
allyl ether (1.0 g) in 10 mL of Decalin was heated to 190-195
°C for 5 h, using the oil bath, and then cooled to room
temperature. A solid product was separated by filtration and
2,6,7-Tr im eth yl-2,3-d ih yd r oben zofu r a n -4,5-d iol (12). A
solution of sodium dithionite (0.53 g, 6.5 mol equiv) in water
(2 mL) was added in one portion to a stirred solution of the
o-quinone 17 (0.09 g, 0.47 mmol in 4 mL of ether). The red
J . Org. Chem, Vol. 68, No. 18, 2003 7031

Hussain et al.
color of the quinone was discharged after 3 min. Aqueous
hydrochloric acid (10%) was added and the opaque mixture
extracted with ether (3 × 15 mL). The combined extracts were
washed with saturated sodium bicarbonate (2 × 10 mL), dried
over MgSO4, and filtered. Evaporation of the solvent afforded
the catechol 12 (0.08 g, 90%) as a yellow solid.
acetate (3 × 30 mL), washed with water (20 mL), dried
(MgSO4), and evaporated to give 1.1 g (65%) of 22 as a light
1
pink solid: mp 158-160 °C. H NMR ((CD3)2CO, 200 MHz) δ
5.79 (s, 2H), 6.45 (s, 2H), 7.49 (s, 2H); 13C NMR ((CD3)2CO, 50
MHz) δ 98.7, 101.3, 139.6, 140.9; HRMS calcd for C7H6O4
154.02661, found 154.02670.
6-Hyd r oxy-2,7,8-tr im eth yl-2-(4,8,12-tr im eth yltr id ecyl)-
ch r om a n -5-ca r ba ld eh yd e (19). The formylation of γ-toco-
pherol, 18, was carried out with the same procedure as was
the conversion of 15 to 16 above. The product 19 was obtained
as a yellow oil (19% yield) following column chromatography.
1H NMR (CDCl3, 200 MHz) δ 0.86 (m, 12H), 0.96-1.60 (m,
24H), 1.81 (m, 2H), 2.13 (s, 3H), 2.16 (s, 3H), 3.00 (t, J ) 6.8
Hz, 2H), 10.16 (s, 1H), 12.10 (s, 1H); 13C NMR (CDCl3, 50 MHz)
δ 193.9, 155.8, 143.9, 138.3, 124.1, 117.4, 114.4, 75.0, 39.5, 39.3,
37.3, 37.2, 32.7, 32.6, 30.7, 27.9, 24.8, 24.4, 23.6, 22.7, 22.6,
20.9, 19.7, 19.6, 18.4, 13.1, 11.0; HRMS calcd for C29H48O3
444.3604, found 444.3605.
2,7,8-Tr im eth yl-2-(4,8,12-tr im eth yltr idecyl)-3,4-dih ydr o-
2H-ch r om en e-5,6-d ion e (20) a n d 2,7,8-Tr im eth yl-2-(4,8,-
12-t r im et h ylt r id ecyl)-3,4-d ih yd r o-2H -ch r om en e-5,6-d i-
ol (21). The rearrangement of 19 to 21 and 20 was carried
out with the same procedure as was the conversion of 16 to
12, above, and gave 20 (12%) as a deep red oil and 21 (50%)
as a yellow oil. For 20: 1H NMR (CDCl3, 500 MHz) δ 0.81-
0.84 (m, 12H), 1.00-1.41 (m, 21H), 1.49 (s, 1H), 1.59 (m, 2H),
1.72 (m, 2H), 1.92 (s, 3H), 2.00 (s, 3H), 2.40 (m, 2H); 13C NMR
(CDCl3, 125 MHz) δ 180.8, 177.8, 163.2, 143.6, 134.1, 110.2,
81.3, 39.9, 39.3, 37.4, 37.3, 37.2, 37.2, 32.9, 32.7, 32.6, 32.5,
29.7, 29.1, 27.9, 25.6, 24.7, 24.4, 23.8, 22.6, 22.5, 21.3, 20.8,
19.7, 19.5, 15.3, 13.6, 11.5; HRMS calcd for C28H46O3 430.3449,
found 430.3447. For 21: 1H NMR (acetone-d6, 200 MHz) δ
0.85-0.89 (m, 12H), 1.11-1.60 (m, 24H), 1.73 (m, 2H), 2.00
(s, 3H), 2.11 (s, 3H), 2.64 (m, 2H), 6.56 (s, 1H), 6.94 (s, 1H);
13C NMR (acetone-d6, 200 MHz) δ 145.9, 141.7, 136.0, 122.9,
115.3, 107.1, 75.3, 40.4, 40.1, 33.5, 33.4, 31.7, 30.6, 30.2, 29.8,
29.4, 29.1, 28.6, 25.5, 25.1, 24.2, 23.0, 22.9, 21.6, 20.1, 17.9,
12.4, 11.5; HRMS calcd for C28H48O3 432.3605, found 432.3603.
Red u ction of 2,7,8-Tr im eth yl-2-(4,8,12-tr im eth yltr id e-
cyl)-3,4-d ih yd r o-2H-ch r om en e-5,6-d ion e (20) to 2,7,8-Tr i-
m e t h y l-2-(4,8,12-t r im e t h y lt r id e c y l)-3,4-d ih y d r o -2H -
ch r om en e-5,6-d iol (21). Sodium dithionite reduction of 20
in the two-phase system ether-water was carried out as in
the case of 16 to 12. The product 21, obtained in 89% yield,
had spectroscopic properties identical with 21, above.
4-Meth ylben zo[1,3]d ioxle-5-ol (25). Methoxymethyl chlo-
ride (4.4 g, 55 mmol) was added slowly to a solution of sessamol
(3, 5 g, 36 mmol) and diisopropyl ethylamine (7.6 mL, 44 mmol)
in dry dichloromethane (20 mL) at 0 °C and the solution was
warmed to 20 °C overnight. Water (10 mL) was added and
the product was extracted with 3 × 30 mL of CH2Cl2), washed
with 10% sodium hydroxide (15 mL) then with water (15 mL),
and dried (MgSO4). Removal of the solvent gave 6.5 g of an oil
(6.5 g). This was used without further purification as described
below. n-Butyllithium (19.4 mL, 1.7 M, 33 mmol) was added
slowly to a solution of (5.0 g, 27 mmol) dry THF (20 mL) under
nitrogen at -78 °C. The solution was maintained for 30 min.
Methyl iodide (3.4 mL, 55 mmol) was added slowly at -78 °C
and the mixture was allowed to warm to 0 °C over 2 h. The
product was extracted with ethyl acetate (3 × 30 mL), washed
with water (20 mL), and dried (MgSO4) and the solvent was
evaporated to give a brown oil (7.6 g). This oil was dissolved
in methanol (100 mL) and treated with concentrated hydro-
chloric acid (0.5 mL). The solution was heated to 62 °C for 1 h
and washed with saturated sodium bicarbonate. The product
was extracted with ethyl acetate, washed with water, and dried
(MgSO4) and the solvent was evaporated to give a solid. Silica
gel chromatography with ethyl acetate as eluent gave 25 as a
pink solid (3.5 g, 83%): mp 95-96 °C. 1H NMR ((CD3)2CO, 50
MHz) δ 2.05 (s, 3H), 5.86 (s, 2H), 6.26 (d, J ) 8 Hz, 1H), 6.46
(d, J ) 8 Hz, 1H), 7.94 (s, 1H); 13C NMR ((CD3)2CO, 50 MHz)
δ 8.9, 101.4, 105.6, 106.5, 108.3, 140.8, 147.5, 151.5; HRMS
calcd for C8H8O3 152.047345, found 152.04880.
4-Meth ylben zo[1,3]d ioxle-5,6-d iol (24). The oxidation of
25 with Fremy’s salt and subsequent reduction with sodium
dithionite was carried out as with the formation of 22. The
1
yield of 24, a light pink solid (mp 100-101 °C) was 78%. H
NMR ((CD3)2CO, 200 MHz) δ 2.08 (s, 3H), 5.78 (s, 2H), 6.32
(s,1H), 6.85 (s, 1H), 7.76 (s, 1H); 13C NMR ((CD3)2CO, 50 MHz)
δ 9.2, 95.9, 100.9, 108.7, 138.4, 139.1, 139.8, 139.9; HRMS calcd
for C8H8O4 168.04226, found 168.04109.
Ack n ow led gm en t. We thank Dr. K. U. Ingold for
helpful discussions, Mr. Gregory Litwinienko for as-
sistance with the stopped-flow apparatus, Dr. Manfred
Dunker (Henkel Corporation) for a very generous gift
of γ-tocopherol, and NSERC (Canada) for financial
support through a Strategic Project Grant to J .S.W.
Ben zo[1,3]d ioxole-5,6-d iol (22). Sesamol (3, 1.52 g, 24
mmol) in 30 mL of methanol was added to a rapidly stirred
solution of Fremy’s salt (7.96 g, 30 mmol) and 5.49 g (40 mmol)
of potassium dihydrogen orthophosphate in 400 mL of water
(400 mL) at 5 °C. The color of the solution changed from light
brown to bright yellow within 5 min. After 30 min the quinone
23 was extracted with 4 × 40 mL of ethyl acetate. The ethyl
acetate solution was treated with a solution of sodium dithion-
ite (9.0 g, 52 mmol) in water (30 mL) and after 5 min the yellow
color changed to a colorless solution. The organic layer was
acidified with hydrochloric acid (1 N), extracted with ethyl
Su p p or t in g In for m a t ion Ava ila b le: 13C NMR spectra
and assignments. This material is available free of charge via
the Internet at http://pubs.acs.org.
J O0301090
7032 J . Org. Chem., Vol. 68, No. 18, 2003