Solvent effects on the antioxidant activity of vitamin E

Full text of DOI:10.1021/jo982360z

J . Org. Chem. 1999, 64, 3381-3383
3381
Solven t Effects on th e An tioxid a n t Activity
of Vita m in E1
and HBA, it was predicted that the magnitude of a KSE
would be independent of the reactivity of the attacking
radical.7 This prediction was confirmed for hydrogen
atom abstraction from TOH (and phenol) by highly
reactive tert-alkoxyl radicals (reaction 4) and by the very
Luca Valgimigli,*,2 J . T. Banks,*,3 J . Lusztyk, and
K. U. Ingold*
•
•
Steacie Institute for Molecular Sciences,
National Research Council of Canada, 100 Sussex Drive,
Ottawa, Ontario, Canada K1A 0R6
RO + TOH f ROH + TO
(4)
unreactive 1,1-diphenyl-2-picrylhydrazyl radical (DPPH•,
reaction 5) in about a dozen solvents chosen so that the
Received December 2, 1998
•
•
Vitamin E (R-tocopherol, TOH) is the major radical-
DPPH + TOH f DPPH + TO
(5)
2
4
,5
trapping, lipid-soluble antioxidant in mammals.
It
protects biological membranes from damage by trapping
the lipid peroxyl radicals, ROO , which would otherwise
rate constants for each radical plus substrate reaction
•
8,12
changed by about 2 orders of magnitude.
Only tert-
destroy the membrane lipids, RH, in a two-step chain
reaction.
butyl alcohol gave anomalous results in that the rate con-
•
stants for hydrogen atom abstraction by DPPH (but not
•
by RO ) were ca. 5 times greater than expected. Subse-
•
•
ROO + RH f ROOH + R
(1)
(2)
quently, rate constants for hydrogen atom abstraction
•
S
from 1,4-cyclohexadiene (CHD) by DPPH , k
,
DPPH/CHD
•
•
•
R + O f ROO
provided further evidence that the reactivity of DPPH
is enhanced by a factor of about 3 in tert-butyl alcohol
relative to its reactivity in many other solvents.
2
1
3
This inhibition of lipid peroxidation is effected via a very
facile transfer of the phenolic hydrogen atom of TOH to
4
,6
peroxyl radicals.
•
•
ROO + TOH f ROOH + TO
(3)
In the present work, we set out to determine whether
solvent effects on the ROO /TOH reaction mirrored those
of reaction 4 or reaction 5.
•
To understand and, hopefully, to model the ability of a
biological membrane to withstand oxidative stress re-
quires knowledge of the absolute rate constant for
S
reaction 3, k
, in biologically relevant systems.
Resu lts
ROO/TOH
However, this rate constant is very likely to be highly
dependent on the solvent, S, because we have found that
Peroxyl radicals were generated “instantaneously” by
UV (300-400 nm) flash photolysis of dicumyl ketone in
2
oxygen-saturated (ca. 5 mM [O ]) solutions at 298 K in
the presence of various concentrations of TOH (see
Scheme 1). Dicumyl ketone was chosen as the peroxyl
rate constants for abstraction of hydroxylic hydrogen
atoms show dramatic solvent effects.7
-11
These kinetic
solvent effects (KSEs) can be accounted for quantitatively
on the basis of hydrogen bonding between the hydrogen
bond donor (HBD) substrate (or antioxidant) and hydro-
Sch em e 1
gen bond acceptor (HBA) solvent.7
,8,11
Because the extent
of the HBD/HBA interaction depends only on the HBD
(
1) Issued as NRCC No. 40919.
(2) NRCC Visiting Scientist 1995. Current address: Department of
Organic Chemistry, University of Bologna, Via S. Donato 15, I-40127
Bologna, Italy.
(
3) NRCC Research Associate 1994-1996. Current address: De-
partment of Chemistry, Lakehead University, Thunder Bay, Ontario
P7B 5E1, Canada.
(
4) Burton, G. W.; Ingold, K. U. Acc. Chem. Res. 1986, 19, 194-
2
01.
radical precursor for two reasons. First, decarbonylation
of the 2,2-dimethyl-2-phenylacetyl radical is sufficiently
(
5) For example, see: Burton, G. W.; J oyce, A.; Ingold, K. U. Lancet
1
982, 2, 327. Burton, G. W.; J oyce, A.; Ingold, K. U. Arch. Biochem.
Biophys. 1983, 221, 281-290. Cheeseman, K. H.; Collins, M.; Proud-
foot, K.; Slater, T. F.; Burton, G. W.; Webb, A. C.; Ingold, K. U. Biochem.
J . 1986, 235, 507-514. Cheeseman, K. H.; Emery, S.; Maddix, S. P.;
Slater, T. F.; Burton, G. W.; Ingold, K. U. Biochem. J . 1988, 250, 247-
8
-1 14
rapid (k
8
) 1.5 × 10 s
)
that oxygen trapping to give
•
2
the acylperoxyl radical, PhCMe C(O)OO , can be ig-
nored.15 Second, the tertiary cumylperoxyl radicals un-
2
52.
6) Burton, G. W.; Doba, T.; Gabe, E. J .; Hughes, L.; Lee, F. L.;
Prasad, L.; Ingold, K. U. J . Am. Chem. Soc. 1985, 107, 7053-7065.
7) Avila, D. V.; Ingold, K. U.; Lusztyk, J .; Green, W. H.; Procopio,
D. R. J . Am. Chem. Soc. 1995, 117, 2929-2930.
8) Valgimigli, L.; Banks, J . T.; Ingold, K. U.; Lusztyk, J . J . Am.
Chem. Soc. 1995, 117, 9966-9971.
9) Valgimigli, L.; Ingold, K. U.; Lusztyk, J . J . Am. Chem. Soc. 1996,
18, 3545-3549.
10) MacFaul, P. A.; Ingold, K. U.; Lusztyk, J . J . Org. Chem. 1996,
1, 1316-1321.
11) Banks, J . T.; Ingold, K. U.; Lusztyk, J . J . Am. Chem. Soc. 1996,
18, 6790-6791. Correction: 12485.
dergo only a slow bimolecular self-reaction (2k11 ) 6 ×
(
(
(12) In any particular solvent, kRO•/kDPPH• was ca. 106 for TOH and
1
0
ca. 10 for phenol.
(
(13) Valgimigli, L.; Ingold, K. U.; Lusztyk, J . J . Org. Chem. 1996,
61, 7947-7950.
(14) Gould, I. R.; Baretz, B. H.; Turro, N. J . J . Phys. Chem. 1987,
91, 925-929.
(15) The elimination of acylperoxyl radicals is important, because
they are very much more reactive than alkylperoxyl radicals, see:
Zaikov, G. E.; Howard, J . A.; Ingold, K. U. Can. J . Chem. 1969, 47,
3017-3029.
(
1
6
1
(
(
1
0.1021/jo982360z CCC: $18.00 © 1999 American Chemical Society
Published on Web 04/16/1999

3
382 J . Org. Chem., Vol. 64, No. 9, 1999
Notes
Ta ble 1. Absolu te Ra te Con sta n ts (M- s-1 a t 298 ( 2 K)
1
for Hyd r ogen Atom Abstr a ction fr om r-Tocop h er ol by
S
Cu m ylp er oxyl Ra d ica ls (k
), ter t-Bu toxyl Ra d ica ls
ROO/TOH
S
(
k
), a n d 1,1-Dip h en yl-2-p icr ylh yd r a zyl Ra d ica ls
RO/TOH
S
(
k
DPPH/TOH)
0-
5
10-8
10-2
k
DPPH/TOH
1
S
S
a
S
a
solvent
hexane
carbon tetrachloride
benzene
k
k
ROO/TOH
RO/TOH
1
200b
140
99c
42
31
74c
36
18
2
3
4
5
6
7
8
9
0
d
34
d
toluene
30
chlorobenzene
anisole
acetic acid
tert-butyl alcohol
acetonitrile
ethyl acetate
27
15
8.8
5.6f
3.8
2.3
36
20
7.7
1.8
9.4
2.9
27
14
6.2
5.7
4.9
1.6
1
a
b
F igu r e 1. Comparison of kinetic solvent effects for H-atom
abstraction from TOH by peroxyl and tert-alkoxyl radicals. The
straight line in the figure has a slope of 1.0 and R ) 0.84.
Data from ref 8. In cyclohexane at 293 K, a value of 68 ×
05 M
-1
-1
18 c
1
s
was obtained by pulse radiolysis.
Value in pentane.
d
In styrene at 303 K, a value of 32 × 10 M-1
5
-1
s
was obtained
6
e
from the rate of the TOH-inhibited oxidation of styrene.
Mea-
sured at 303 K. f Measurements of the rate of the TOH inhibited
oxidation of methyl linoleate at 310 K in tert-butyl alcohol gave
5
-1 -1 19
5
.1 × 10 M
s
and of the TOH-inhibited oxidation of linoleic
5
-1 -1 20
acid at 303 K in tert-butyl alcohol gave 2.3 × 10
M
s .
03 M-1 s-1
)
16
so they can be exclusively trapped by the
1
1
7
added TOH.
•
2
PhCMe OO f nonradical products
(11)
2
Hydrogen atom abstraction from TOH by the cumylper-
oxyl radicals was monitored by following the pseudo-first-
order growth of the 420 nm absorption of the tocopheroxyl
radicals, kobs. At least three separate measurements of
k
obs were made at each concentration of TOH employed,
and at least five different concentrations of TOH were
F igu r e 2. Comparison of kinetic solvent effects for H-atom
abstraction from TOH by peroxyl and DPPH radicals. The
straight line in the figure has a slope of 1.0 and R ) 0.90.
employed in each solvent. Absolute second-order rate
•
S
constants, k
, in each solvent, S, were obtained by
ROO/TOH
least-squares fitting of plots of kobs vs [TOH] according
to
determined by measurement of the TOH-inhibited au-
S
5
-1 -1 6
toxidation of styrene (32 × 10 M
s ) is quite remark-
able and speaks well for both techniques.
kobs ) k + k
[TOH]
o
ROO/TOH
S
Excellent straight lines were obtained in all solvents (R
g 0.98). Values of k
with our earlier kinetic data for reaction 4 (k
reaction 5 (k
As expected, values of k
are strongly solvent
ROO/TOH
S
are given in Table 1 together
dependent and the magnitude of the KSEs on reactions
ROO/TOH
S
3
-5 are essentially identical. That is, Figure 1 shows a
) and
RO/TOH
S
S
S
plot of log k
vs log k
, and it can be seen
).
RO/TOH
ROO/TOH
DPPH/TOH
that all of the points lie close to the straight line shown
that has been drawn with the theoretical slope of 1.0.
Discu ssion
Where comparison is possible, the present values of
S
Figure 2 shows a similar plot of log k
vs log
DPPH/TOH
S
k
. These two figures add further support to our
ROO/TOH
S
k
are in satisfactory agreement with earlier di-
ROO/TOH
view that KSEs for abstraction of hydroxylic hydrogen
atoms are, to a first approximation, independent of the
attacking radical, depending instead on the HBD ability
rect18 and indirect
6,19,20
rate constant determinations in
the same (or similar) solvents (see footnotes b, d, and f
in Table 1). The agreement between the values of
7,8,11,21
of the substrate and the HBA ability of the solvent.
S
5
k
determined in this work in benzene (34 × 10
ROO/TOH
A closer examination of the two figures reveals that
-
1
-1
5 -1 -1
M
s
) and in toluene (30 × 10
M
s
) and that
•
the kinetic data for the ROO /TOH reaction correlates
•
better with that for the DPPH /TOH reaction (R ) 0.90,
(
16) Howard, J . A.; Ingold, K. U. Can. J . Chem. 1968, 46, 2655-
660.
17) Primary and secondary alkylperoxyl radicals undergo much
faster bimolecular self-reactions (Ingold, K. U. Acc. Chem. Res. 1969,
, 1-9), and exclusive trapping by the added TOH could not be
guaranteed.
•
Figure 2) than with that for the RO /TOH reaction (R )
2
0
.84, Figure 1). The better correlation seen in Figure 2
(
S
S
arises largely because both k
and k
are
ROO/TOH
DPPH/TOH
2
larger in tert-butyl alcohol (8) than in acetonitrile (9) and
S
(18) J ovanovic, S. V.; J ankovic, I.; J osimovic, L. J . Am. Chem. Soc.
ethyl acetate (10). In contrast, k
is lowest in tert-
RO/TOH
1
1
992, 114, 9018-9021.
butyl alcohol and decreases along the solvent series,
(19) Niki, E.; Saito, T.; Kawakami, A.; Kamiya, Y. J . Biol. Chem.
984, 259, 4177-4182.
20) Barclay, L. R. C.; Baskin, K. A.; Locke, S. J .; Schaefer, T. D.
Can. J . Chem. 1987, 65, 2529-2540.
(
(21) Lucarini, M.; Pedulli, G. F.; Valgimigli, L. J . Org. Chem. 1998,
63, 4497-4499.

Notes
J . Org. Chem., Vol. 64, No. 9, 1999 3383
acetonitrile > ethyl acetate > tert-butyl alcohol, an order
kinetics of a few hydrogen atom abstraction reactions
remains to be determined.
which is consistent with the relative HBA abilities of
2
2
these three solvents. Our present results indicate,
•
Exp er im en ta l Section
therefore, that ROO radicals have an enhanced reactivity
in tert-butyl alcohol comparable in magnitude to the
Ma ter ia ls. Solvents were of the purest grade commercially
available and were used as received. 2R,4′R,8′R-R-tocopherol
(natural vitamin E) was purified chromatographically on silica
gel as described previously.8 Dicumyl ketone was synthesized
from dibenzyl ketone (Aldrich, purified by crystallization from
hexane to >99% by GC-MS) “piecemeal” in three steps with
purification after each step, more or less following a literature
procedure.27 Briefly, dibenzyl ketone was slowly added to 2.4
equiv NaH in dry THF under argon at 0 °C, the mixture was
allowed to warm to room temperature with stirring (30 min),
•
previously discovered enhanced reactivity of DPPH
8
,13
radicals in this solvent
(but not in methanol or
ethanol).13 These two radicals are isoelectronic and both
are stabilized by conjugative electron delocalization:
2
.4 equiv of methyl iodide was added, and the mixture was
refluxed for 2 h. After the addition of cold (0 °C) water, the crude
dimethylated derivative was extracted with ethyl ether, the
ether was removed under vacuum, and the derivative was
purified by crystallization from hexane. This process was
repeated twice more using half the quantities of NaH and MeI,
with purification by recrystallization of the intermediate tri-
methylated derivative. The dicumyl ketone was obtained in pure
form (>99%, GC-MS) in an overall yield of 40%: mp 111-113
therefore, a similar tert-butyl alcohol-induced increase in
their reactivities, whatever its cause, is not entirely
unreasonable. However, the situation must be somewhat
more complex, because one of us has previously demon-
strated that there is no KSE in a reaction in which a C-H
bond is broken by an attacking peroxyl radical (reaction
27
28 1
°
C, lit. 111-112 °C, 110-115 °C (dec); H NMR (200 MHz,
CDCl ) δ: 1.35 (s, 12 H), 7.2 (m, 2 H), 7.3 (m, 8 H), lit.27 (CDCl3)
3
•
21
1
, with RH ) cumene and ROO ) cumylperoxyl).
δ: 1.26 (s, 12 H), 7.1 (bs, 10 H).
S
Kin etic Mea su r em en ts. Conventional flash photolysis work
was performed in a commercial PRA FP-1000 flash system. The
system was operated at 6.0 kV, and the exciting light was filtered
through Pyrex. Transient absorptions were monitored at 420 nm.
The signals were captured by a Tektronix 24.32 digital scope
interfaced to a PowerMac 7100 computer that provided suitable
processing facilities. Experiments were performed in oxygen-
saturated (760 Torr) solvents, using a 10 cm optical path quartz
Specifically, k
is the same within experimental
ROO/RH
error in tert-butyl alcohol, isooctane, benzene, acetonitrile
and pyridine.21 This result was not expected because the
•
rate constant for C-H bond breaking by DPPH ,
S
k
(reaction 6) was 3 times greater in tert-butyl
DPPH/CHD
alcohol than in other solvents.
1
3
To summarize, the peroxyl radical/R-tocopherol reac-
tion exhibits the expected large kinetic solvent effects.
However, our work reveals that there are still some
relatively small (ea factor of ca. 5) KSEs on hydrogen
atom abstraction reactions which are not understood at
present. To be specific, tert-butyl alcohol enhances the
2
9
cuvette. The solvents contained ca. 50 mM dicumyl ketone and
0
.5-2.5 mM R-tocopherol. At least five different concentrations
of TOH were employed for each solvent, and every measurement
was repeated at least three times. The pseudo-first-order growth
of the TO absorbance was monitored at 420 nm. Absolute
•
S
second-order rate constants, kROO/TOH, were obtained by least-
•
13
squares fitting of the observed pseudo-first-order rate constant,
rate constants for (i) C-H bond breaking by DPPH ,
•
8
k
obs, vs [TOH].
(
ii) O-H bond breaking by DPPH , and (iii) O-H bond
•
breaking by ROO (this work) but has no special effect
Ack n ow led gm en t. Our sincere thanks go to Profes-
•
21
on (iv) C-H bond breaking by ROO , (v) C-H bond
sor J . C. Scaiano not only for allowing us to use his
equipment for all experiments described herein but also
for his continued interest, advice, and encouragement.
We also thank the National Foundation for Cancer
Research for partial support of this work, and one of us
(L.V.) thanks the University of Bologna for a grant.
•
25
• 8,10
breaking by RO , (vi) O-H bond breaking by RO ,
or
•
26
(
vii) O-H bond breaking by R . The reason for the
deviant behavior of tert-butyl alcohol solvent on the
H
2
(
22) For acetonitrile, ethyl acetate, and tert-butyl alcohol, â
)
2
3
0
.44, 0.45, and 0.49, respectively, and â ) 0.31, 0.45, and g0.6,
24
respectively.
(
23) Abraham, M. H.; Grellier, P. L.; Prior, D. V.; Morris, J . J .;
Taylor, P. J . J . Chem. Soc., Perkin Trans. 2 1990, 521-529.
24) Kamlet, M. J .; Abboud, J .-L. M.; Abraham, M. H.; Taft, R. W.
J . Org. Chem. 1983, 48, 2877-2887.
25) Avila, D. V.; Brown, C. E.; Ingold, K. U.; Lusztyk, J . J . Am.
Chem. Soc. 1993, 115, 466-470.
26) Franchi, P.; Lucarini, M.; Pedulli, G. F.; Valgimigli, L.; Lunelli,
B. J . Am. Chem. Soc. 1999, 121, 507-514.
J O982360Z
(
(27) Dyllick-Brenzinger, R. A.; Patel, V.; Rampersad, M. B.; Stothers,
J . B.; Thomas, S. E. Can. J . Chem. 1990, 68, 1106-1115.
(28) Gould, I. R.; Zimmt, M. B.; Turro, N. J .; Baretz, B. H.; Lehr, G.
F. J . Am. Chem. Soc. 1985, 107, 4607-4612.
(
(
(29) The dicumyl ketone was added to the solvent at such
a
concentration that the OD at 320 nm in a 7 mm cuvette was ca. 1.0.