A carbon-centered radical unreactive toward oxygen: Unusual radical stabilization by a lactone ring

Article abstract of DOI:10.1021/ol9913946

(equation presented) A lactone ring confers unusual stability to a diphenylmethyl-like radical that is virtually unreactive toward oxygen. Thus, the radical derived from HP-136 is about 10 000 times less reactive than typical carbon-centered radicals. A r

Full text of DOI:10.1021/ol9913946

ORGANIC
LETTERS
2
000
Vol. 2, No. 7
99-901
A Carbon-Centered Radical Unreactive
Toward Oxygen: Unusual Radical
Stabilization by a Lactone Ring
8
,†
†
†
‡
J. C. Scaiano,* Alan Martin, Glenn P. A. Yap, and K. U. Ingold
Department of Chemistry, UniVersity of Ottawa, 10 Marie Curie,
Ottawa, Ontario, K1N 6N5 Canada, and Steacie Institute for Molecular Sciences,
National Research Council, Ottawa, Ontario, K1A 0R6 Canada
tito@photo.chem.uottawa.ca
Received December 23, 1999
ABSTRACT
A lactone ring confers unusual stability to a diphenylmethyl-like radical that is virtually unreactive toward oxygen. Thus, the radical derived
from HP-136 is about 10 000 times less reactive than typical carbon-centered radicals. A reversible reaction with oxygen is proposed by
analogy with triphenylmethyl; however, the association constant is about 1000 times smaller for HP-136 than for triphenylmethyl. While the
lactone ring greatly influences the reactivity, the spectroscopy of the HP-136-derived radical is in line with that expected for a substituted
diphenylmethyl radical.
Antioxidants rarely involve the cleavage of C-H bonds. This
the benzylic C-H bond as the only anticipated good
hydrogen donor site. Abstraction, as illustrated in reaction
1, by tert-butoxyl radicals should lead to a carbon-centered
is due to the fact that the vast majority of carbon-centered
1
radicals react with dioxygen readily, leading to peroxyl
radicals that then propagate the oxidative chain. Thus, it is
not sufficient for a molecule to be a good hydrogen donor
for it to be an antioxidant. In addition, the radical produced
must be unable to propagate the autoxidation chain. Some
antioxidants, notably phenols, can produce radicals that not
only do not propagate the autoxidation chain but can
themselves trap a second radical, thus leading to a 2-to-1
2
stoichiometry for the radical-mediated antioxidant steps.
We were surprised by CIBA’s recent claims that HP-136
was an excellent antioxidant for high-temperature polymer
processing. Examination of the structure of HP-136 reveals
3
radical, virtually identical to the diphenylmethyl radical.
Indeed, photolysis of di-tert-butyl peroxide in the presence
†
University of Ottawa.
Steacie Institute for Molecular Sciences.
‡
(
1) Maillard, B.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc. 1983,
(5) Bensasson, R.; Land, E. J. Trans. Faraday Soc. 1971, 67, 1904.
(6) Land, E. J. Proc. R. Soc. London A 1968, 305, 457.
(7) Avila, D. V.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1992,
114, 6576.
(8) Howard, J. A.; Ingold, K. U. Can. J. Chem. 1968, 46, 2655.
(9) Avers, C. L.; Janzen, E. G.; Johnston, F. J. J. Am. Chem. Soc. 1966,
88, 2610.
1
1
05, 5095.
(
(
2) Ingold, K. U. Acc. Chem. Res. 1969, 2, 1.
3) Paul, H.; Small, R. D., Jr.; Scaiano, J. C. J. Am. Chem. Soc. 1978,
00, 4520.
4) Chatgilialoglu, C. In Handbook of Organic Photochemistry; Scaiano,
J. C., Ed.; CRC Press: Boca Raton, FL, 1989; Vol. II, p 3.
(
1
0.1021/ol9913946 CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/14/2000

of HP-136 reveals the formation of the radical of Figure 1,
where the UV band is virtually identical to that for diphen-
ylmethyl. The visible band (see 500-600 nm region) is more
would lead to an upper limit for its reacton with oxygen of
5
-1 -1
e10 M s , i.e., over 10 000 times less reactive than
4
1
typical free radicals. Careful comparison of nitrogen- and
oxygen-saturated samples, at either the 340 or 540 nm
maxima, shows that while neither the decay nor the growth
kinetics are significantly affected by oxygen (see Figure 2),
Figure 1. Transient spectra recorded following 308 nm laser
excitation of a sample containing 1% (w/v) HP-136 and 5% (v/v)
di-tert-butyl peroxide in acetonitrile under nitrogen. Monitored at
different times following the laser pulse, where the signal grows
and decays as indicated. The inset shows an expansion of the visible
region and facilitates visualizing the order of the traces.
Figure 2. Transient kinetic traces recorded at 340 nm following
3
1
08 nm laser excitation of a sample containing 2.5% (w/v) HP-
36 and 5% (v/v) di-tert-butyl peroxide in benzene under nitrogen
and oxygen.
•
intense than that in the case of Ph
2
C H, but enhancement of
the signal amplitude was 18% smaller in the presence of
oxygen. The true decrease is smaller than suggested by this
figure; it can be attributed to two origins, which we label
trivial and nontrivial. A trivial decrease of 11% is due to an
enhancement of the solution absorption at 308 nm (laser
excitation wavelength) in the presence of oxygen. Such
charge-transfer-mediated enhancements are common, al-
though they are frequently overlooked or believed to be
significantly smaller. The remaining 7% decrease in signal
intensity (thus contributing to the observed 18% decrease)
must have a different origin. We propose that this “nontrivial”
decrease should be attributed to a reversible reaction of the
HP-136 radical with oxygen, as illustrated in reaction 2. This
this band is common when heteroatoms are involved; the
•
best documented example is Ph
2
C OH, the ketyl radical from
4
-6
benzophenone.
presence of HP-136 shows that the formation of the HP-
36 radical (the same as in Figure 1) occurs concurrently
Photolysis of dicumyl peroxide in the
1
with the decay of cumyloxyl. We note that cumyloxyl
radicals are far more easily detectable than tert-butoxyl and
allow the easy corroboration of the precursor-derivative
relationship between cumyloxyl and the HP-136 radical.
C H was produced by laser irradiation
of di-tert-butyl peroxide in the presence of diphenylmethane.
7
•
In an experiment, Ph
2
When the sample was saturated with oxygen, the signal from
•
Ph
2
C H was totally quenched, with barely a fast spike
indicating its presence in the early stages following the laser
pulse. Remarkably, when the same experiment was repeated
with HP-136, oxygen did not cause any appreciable changes
in the rate of decay of the radical. Taking this lack of
quenching to reflect radical unreactivity toward oxygen
(
10) Janzen, E. G.; Johnston, F. J.; Avers, C. L. J. Am. Chem. Soc. 1967,
9, 1176.
11) Data were collected on a Bruker AX SMART 1K CCD diffracto-
8
(
type of reversibility is unprecedented for a diphenylmethyl-
type structure but is documented for triphenylmethyl radicals,
which are known to react with oxygen reversibly;
meter, Mo KR (λ ) 0.71073 Å), in the range 4.8° e 2θ e 45° using 0.3°
ω scans at 0, 90, and 180° in φ. No absorption corrections were required
-
1
8-10
(
µ ) 0.71 cm ). Systematic absences in the diffraction data [15943
collected, 2654 observed, unique (I > 2σI)] and unit-cell parameters [a )
however, in this case the equilibrium is extensively displaced
1
2
1
3.791(3) Å, b ) 10.797(2) Å, c ) 14.412(3) Å, â ) 108.174(3)°, V )
038.9(6), Z ) 4] were uniquely consistent with monoclinic, P21/c (No.
4). A tert-butyl group was found rotationally disordered in two positions
8
toward the peroxyl radical side. Howard and Ingold studied
this equilibrium as early as 1968 and determined an equi-
with a refined site occupancy distribution of 70/30. All non-hydrogen atoms
were refined with anisotropic displacement parameters. All hydrogen atoms
3
-1
librium constant of 8 × 10 M , or more conveniently
2
were treated as idealized contributions. RF(wRF ) ) 0.0492 (0.1287), GOF
expressed in terms of oxygen pressure above the solution,
)
1.040. All scattering factors and anomalous dispersion factors are
-
1
K ) 60 atm at 30 °C in hydrocarbon solvents. A similar
contained in the SHEXTL 5.10 program library (Sheldrick, G. M. Bruker
AXS, Madison, WI, 1997).
analysis for HP-136 would suggest an equilibrium constant
900
Org. Lett., Vol. 2, No. 7, 2000

-
1
for reaction 2 of ca. 0.08 atm . This means that even
accepting the hypothesis of reversibility, the HP-136 radical
is about 1000 times less reactiVe than triphenylmethyl toward
oxygen. While this is consistent with the manufacturer’s
claims, it also suggests an unprecedented lack of reactivity
toward oxygen for a relatively simple carbon-centered
radical.
decreases the reactivity toward electrophillic dioxygen. Steric
effects are unlikely to provide much hindrance at the radical
site given the rather remote location of the substituents. Thus,
we suggest that the antioxidant effectiveness of lactone
antioxidants is largely due to the forced conformation of these
systems and to electronic effects reflecting the electron-
withdrawing nature of the lactone ring.
In an attempt to rationalize this unusual result, we obtained
11
an X-ray structure for HP-136 (see Figure 3 and Supporting
Thus, HP-136 is indeed an antioxidant due to the low
reactivity toward oxygen of its diphenylmethyl-like radical.
It is quite possible that this radical itself can act as a radical
scavenger by trapping carbon-centered, or more likely,
peroxyl radicals. Further, it is likely that at the operating
temperatures of polymer processing the equilibrium of
reaction 2 is even further displaced to the left, coincidentally
optimizing the operating conditions for this antioxidant.
Whether this is a unique example or whether the principles
discovered in this work can be further extended in the rational
design of antioxidants remains to be seen.
Figure 3. ORTEP diagram showing the structure of HP-136.
Acknowledgment. J.C.S. thanks the Natural Sciences and
Engineering Research Council of Canada for support and
the Instituto de Tecnolog ´ı a Qu ´ı mica (Valencia) where this
article was completed while J.C.S was a guest under the
auspices of the Ministerio de Educaci o´ n y Ciencia (Spain).
We are grateful to Drs. Rick King and Ray Seltzer from
CIBA Specialty Chemicals for a generous gift of HP-136.
Information). The lactone ring clearly forces the coplanarity
of the attached aromatic ring, thus maximizing delocalization
of the unpaired election onto this ring; the mean deviation
from planarity of the lactone ring is only 0.0041 Å. This
effect seems to make the two rings in HP-136 more effective
3
than the three rings in Ph C in providing resonance stabiliza-
Supporting Information Available: Crystal structure of
HP-136. This material is available free of charge via the
Internet at http://pubs.acs.org.
tion. In the triphenylmethyl radical, steric interactions force
the rings into a propeller-like structure, none of them ideally
oriented for electron delocalization. Further, it is likely that
the electron acceptor nature of the lactone ring further
OL9913946
Org. Lett., Vol. 2, No. 7, 2000
901