Solvent effects on the reactivity and free spin distribution of 2,2-diphenyl-1-picrylhydrazyl radicals

Full text of DOI:10.1021/jo9608578

J . Org. Chem. 1996, 61, 7947-7950
7947
this solvent the two DPPH• reactions were both ca. 5
times faster than would have been predicted from the
rates of the two corresponding alkoxyl radical reactions.
Herein, we report a study of KSE’s on C-H bond
cleavage induced by the DPPH• radical using 1,4-cyclo-
hexadiene as the substrate. These kinetic results proved
that, although the absolute reactivity of DPPH• was very
largely solvent independent, it was, indeed, larger in tert-
butyl alcohol than in any other solvent, including metha-
nol and ethanol in which its reactivity is “normal”. In
an attempt to understand the DPPH• reactivity-enhanc-
ing properties of tert-butyl alcohol9 we also measured
the effect of solvents on the distribution of unpaired
spin density between the two central nitrogen atoms of
DPPH• using EPR spectroscopy. Since these two nitrogen
atoms are spectroscopically almost equivalent, we syn-
thesized DPPH• labeled with 15N at the divalent nitrogen
(N1) which is the “formal” radical center (canonical
structure A).
Solven t Effects on th e Rea ctivity a n d
F r ee Sp in Distr ibu tion of
2,2-Dip h en yl-1-p icr ylh yd r a zyl Ra d ica ls1
Luca Valgimigli,*,2 K. U. Ingold,* and J . Lusztyk*
Steacie Institute for Molecular Sciences, National Research
Council of Canada, Ottawa, Ontario, Canada K1A 0R6
Received May 9, 1996
There is no kinetic solvent effect (KSE) on hydrogen
atom abstraction from a hydrocarbon (cyclohexane) by
the cumyloxyl radical.3 That is, the rate constants for
reaction 1 are equal within experimental error ((1.2 (
0.1) × 106 M-1 s-1 at 30 °C) in CCl4, C6H6, C6H5Cl,
•
PhCMe2O• + c-C6H12 f PhCMe2OH + c-C6H11
(1)
CH3C(O)OH, CH3CN, and (CH3)3COH. In contrast to
C-H bond cleavage, there is a large KSE on abstraction
of the hydroxylic hydrogen atom from tert-butyl hydro-
peroxide and from phenol by the cumyloxyl radical.4
Thus, for the latter reaction, 2:
PhCMe2O• + PhOH f PhCMe2OH + PhO• (2)
Resu lts
the rate constants at 25 °C in CCl4, C6H5Cl, C6H6, C6H5-
OCH3, CH3C(O)OH, CH3CN, and (CH3)3COH are, respec-
Kin etic Mea su r em en ts. The rate constants for reac-
tion of the DPPH• radical with 1,4-cyclohexadiene (CHD)
were measured at 30 ( 0.1 °C by spectrophotometry after
rapidly mixing concentrated, deoxygenated stock solu-
tions of the two reactants. Initial reagent concentrations
were chosen so as to have convenient reaction rates with
a large excess of CHD so that pseudo-first-order kinetics
obtained. Typically, [DPPH•] at 1.4 × 10-4 M was reacted
with five different concentrations of CHD in the range
(3-15) × 10-2 M. The decay of the DPPH• was monitored
simultaneously at 6 different wavelengths between 510
and 620 nm, including its band maximum at 520 nm. The
overall chemistry can be represented by reactions 4 and
5. Plots of the experimental first-order rate constant,
tively, 86, 48, 28, 5.6, 1.8, 0.58, and 0.36 × 107 M-1 s-1 5
.
The occurrence of large KSE’s on O-H bond cleavage was
attributed to hydrogen bond formation between the
substrate as the donor and hydrogen bond-accepting
(HBA) solvents with abstraction of a hydrogen atom
occurring only (or mainly) from substrate molecules
which were not hydrogen bonded.4 Since this is a solvent
effect on one of the reactants, i.e., it is a ground state
effect, it was predicted that the magnitude of the KSE
would be independent of the nature and reactivity of the
radical which abstracts the hydroxylic hydrogen atom.4
That is, for the reaction,
XOH + Y• f XO• + YH
(3)
the ratio of the measured rate constants in solvents A
and B was predicted to be independent of the structure
of Y•, i.e., (kXAOH/Y)/(kXBOH/Y) ) constant (for the same
XOH). This prediction was largely confirmed6 for XOH
) phenol with Y• ) cumyloxyl and 2,2-diphenyl-1-
picrylhydrazyl (DPPH•) and for XOH ) R-tocopherol
(vitamin E) and Y• ) tert-butoxyl and DPPH•.8 However,
there was one anomalous solvent, tert-butyl alcohol. In
ksexptl, versus the CHD concentration in each solvent,
S, were linear (r g 0.99) which proves that the reverse
of reaction 4 is unimportant even in the later stages of
the reaction. Absolute second-order rate constants, ks4,
were obtained from these plots by the method of least
squares:
(1) Issued at NRCC No. 39122.
(2) NRCC Visiting Scientist 1995. Permanent address: Universita´
di Bologna, Dipartimento di Chimica Organica “A. Mangini”, Via S.
Donato 15, 40127 Bologna, Italy.
(3) Avila, D. V.; Brown, C. E.; Ingold, K. U.; Lusztyk, J . J . Am. Chem.
Soc. 1993, 115, 466-470.
(4) Avila, D. V.; Ingold, K. U.; Lusztyk, J .; Green, W. H.; Procopio,
D. R. J . Am. Chem. Soc. 1995, 117, 2929-2930.
(5) Additional KSE data on reaction 2 can be found in references 6
and 7.
kesxptl ) ks0 + 2ks4[CHD]
These data are reported in Table 1.
(6) Valgimigli, L.; Banks, J . T.; Ingold, K. U.; Lusztyk, J . J . Am.
Chem. Soc. 1995, 117, 9966-9971.
EP R Spectr oscopic Measu r em en ts. The 2,2-diphen-
yl-1-picrylhydrazyl radical, DPPH•, was first reported in
192210 and was the first free radical for which hyperfine
(7) MacFaul, P. A.; Ingold, K. U.; Lusztyk, J . J . Org. Chem. 1996,
61, 1316-1321.
(8) In the same solvent, kAPhOH/CumO/kPAhOH/DPPH ) 1.0 × 1010 and
kATOH/BO/kATOH/DPPH ) 1.6 × 106, where TOH is R-tocopherol and BO is
tert-butoxyl.6
(9) Shared to a lesser extent by n-butanol and 2-propanol, vide infra.
(10) Goldschmidt, S.; Renn, K. Berichte 1922, 55, 628-643.
S0022-3263(96)00857-2 CCC: $12.00 Published 1996 by the American Chemical Society

7948 J . Org. Chem., Vol. 61, No. 22, 1996
Notes
Ta ble 1. Absolu te Ra te Con sta n ts for Hyd r ogen Atom Abstr a ction fr om 1,4-Cycloh exa d ien e by DP P H• a t 30 °C in
Tw elve Solven ts a n d EP R P a r a m eter s for DP P H• in th e Sa m e Solven ts
s
a
no.
solvent
103 × 2k4 (M-1 s-1
)
a(N1)b (G)
a(N2) (G)
a(N1)/a(N2)
a(N1) + a(N2)
g
1
2
3
4
5
6
7
8
9
CCl4
1.3
1.4
1. 4
1.3
1.4
1.1
1.1
1.2
1.4
1.7
2.4
3.4
9.77 ( 0.02
9.77 ( 0.02
9.77 ( 0.02
9.78 ( 0.02
9.8 ( 0.1
7.94 ( 0.02
7.94 ( 0.02
7.94 ( 0.02
7.93 ( 0.02
7.9 ( 0.1
1.23
1.23
1.23
1.23
1.2
1.15
1.17
1.18
1.18
1.18
1.18
1.17
17.71
17.71
17.71
17.71
17.7
2.00365
2.00364
2.00363
2.00363
2.0036
c
benzene
ethyl acetate
γ-valerolactone
acetonitrile
DMSO
acetic acid
methanol
ethanol
n-butanol
2-propanol
tert-butyl alcohol
9.3 ( 0.2
8.1 ( 0.2
17.4
9.60 ( 0.02
9.54 ( 0.02
9.55 ( 0.02
9.58 ( 0.02
9.60 ( 0.02
9.62 ( 0.02
8.22 ( 0.02
8.10 ( 0.02
8.11 ( 0.02
8.15 ( 0.02
8.17 ( 0.02
8.20 ( 0.02
17.82
17.64
17.66
17.73
17.77
17.82
2.00356
2.00356
2.00356
2.00356
2.00358
2.00359
10
11
12
Errors ) (10-15%. Calculated from the measured 15N hfcc by multiplication with the magnetogyric ratio, γ(14N) / γ(15N) ) 0.7129.
a
b
c Reference for instrument calibration.
splitting of the EPR signal was observed.11 Its spectral
parameters in single crystals, powders, and glasses12 and
in solution13 have been carefully evaluated in several
investigations by EPR, NMR, ELDOR, ENDOR, and
TRIPLE techniques. Solvent effects on the spin distribu-
tion and, as a consequence, on the electron-nuclear
hyperfine coupling constants (hfcc) of DPPH• were cur-
sorily examined many years ago,14 and limited evidence
was obtained that the ratio of the hfcc for the two central
nitrogen atoms was solvent dependent.
To investigate this solvent effect in greater detail,
DPPH-15N1 was synthesized from diphenylamine by di-
azotization with Na15NO2 to give N-[15N]-nitrosodiphen-
ylamine which was then reduced to [15N2]-1,1-diphenyl-
hydrazine. The hydrazine was isolated as the p-toluene-
sulfonate which was treated with picryl chloride in the
presence of sodium carbonate to give 2,2-diphenyl-1-
[15N1]-picrylhydrazine. This was subsequently oxidized
to give the desired stable free radical.
EPR spectra were recorded at 25 °C using DPPH•-15N1
at concentrations in the range (8-20) × 10-5 M in
deoxygenated solvents. Spectra were simulated until
they were in satisfactory agreement with the experimen-
tal spectra in order to obtain the hyperfine coupling
constants of the two central nitrogen atoms. The mea-
sured g-values were corrected with respect to the known
value in benzene solution and were confirmed by com-
parison with the measured value for unlabeled DPPH•
powder. These results are given in Table 1.
methanol (8) and ethanol (9), have no significant effect
on the intrinsic reactivity of DPPH•.15 That is, for these
nine solvents 2ks4 ) (1.25 ( 0.15) × 10-3 M-1 s-1
.
However, and as predicted,6 the reactivity of DPPH• is
enhanced in tert-butyl alcohol (12), there being a three-
Me3COH
fold increase in 2k4
with respect to the nine
“normal” solvents. 2-Propanol (11) enhances the reactiv-
Me2CHOH
ity of DPPH• by a factor of two (2k4
) 2.4 × 10-3
M-1 s-1) and there may be a small enhancement of
reactivity in n-butanol (10) (2kn4-BuOH ) 1.7 × 10-3 M-1
s-1).
When the enhanced reactivity of DPPH• in tert-butyl
alcohol was first reported,6 we hypothesized that it was
due to the unpaired electron becoming more localized at
its formal site on N1. The idea was that tert-butyl alcohol
formed a hydrogen bond to N2 (structure C) and that this
reduced conjugative electron delocalization from N1 to N2,
i.e., structure A, would be of greater importance relative
to structure B in tert-butyl alcohol relative to other
solvents. This would imply that the N1 hfcc, a(N1), would
be larger and the N2 hfcc would be smaller in tert-butyl
alcohol than in other solvents. An examination of Table
1 reveals that this is not the case.
C
Discu ssion
The measured rate constants for hydrogen atom ab-
straction from 1,4-cyclohexadiene by the DPPH• radical
have a probable experimental error of ca. ( 10 to 15%.
Thus, the first nine solvents listed in Table 1, including
Solvent effects on the hfcc’s of the two central nitrogen
atoms and on the g-value of DPPH are small but
significant.17 The solvent effect on the hfcc of the
trivalent nitrogen atom, a(N2), is similar to, though
smaller than, the very well studied solvent effect on the
(trivalent) nitrogen atoms’ hfcc’s of nitroxide radicals19
(see Figure 1). This is reasonable since hydrazyl radicals
and nitroxides are electronically related. In a valence
(11) Hutchinson, C. A., J r.; Pastor, R. C.; Kowalsky, A. G. J . Chem.
Phys., 1952, 20, 534-535. In this study only the averaged hfcc of the
two central nitrogen atoms was observed. The first report of the
inequality of these two nitrogen hfcc’s is: Deal, R. M.; Koski, W. S. J .
Chem. Phys. 1959, 31, 1138-1139.
(12) Holmberg, R. W.; Livingston, R.; Smith, W. T., J r. J . Chem.
Phys. 1960, 33, 541-546. Gamo, K.; Masuda, K.; Yamaguchi, J .;
Kakitani, T. J . Phys. Soc. J pn. 1965, 20, 1730. Gubanov, V. A.;
Koriakov, V. I.; Chirkov, A. K. J . Magn. Reson. 1973, 9, 263-274.
(13) Chen, M. M.; Sane, K. V.; Walter, R. I.; Weil, J . A. J . Phys.
Chem. 1961, 65, 713-717. Hyde, J . S.; Sneed, R. C., J r.; Rist, G. H. J .
Chem. Phys. 1969, 51, 1404-1416. Gubanov, V. A.; Koryakov, V. I.;
Chirkov, A. K. J . Magn. Reson. 1973, 11, 326-334. Dalal, N. S.;
Kennedy, D. E.; McDowell, C. A. J . Chem. Phys. 1973, 59, 3403-3410.
Biehl, R.; Mo¨bius, K.; O’Connor, S. E.; Walter, R. I.; Zimmermann, H.
J . Phys. Chem. 1979, 83, 3449-3456.
(15) The only related work of which we are aware and which appears
to have followed “clean” and sensible kinetics involved hydrogen atom
abstraction from 9,10-dihydroanthracene by DPPH.16 However, these
rather old results16 are not congruent with our own in that a solvent
effect was observed where we see none (CCl4 vs PhH) or would expect
none (dioxane). Thus, the rate constants reported in CCl4 are 0.26 and
0.68 × 10-3 M-1 s-1 at 30 and 50 °C, respectively, in benzene 0.12 and
0.37 × 10-3 M-1 s-1 at the same two temperatures, and in dioxane
0.22 × 10-3 M-1 s-1 at 50 °C.
(14) Garif’yanov, N. S.; Il’yasov, A. V.; Yablokov, Yu. V. Dokl. Akad.
Nauk SSSR 1963, 149, 876-879. English translation, pp 280-283.
(16) Hogg, J . S.; Lohmann, D. H.; Russell, K. E. Can. J . Chem. 1961,
39, 1394-1396.

Notes
J . Org. Chem., Vol. 61, No. 22, 1996 7949
structure C).6 It is now clear that there is no simple
correlation between reactivity, 2ks4, and a(N2)s (see
Table 1). However, if we are willing to ignore the data
in DMSO (6) (in part because of the large errors in the
magnitudes of the two nitrogen hfcc’s)20 and in acetic acid
(7) (for unknown reasons), there may be a weak correla-
tion between the reactivity of DPPH• and the total spin
density on its two central nitrogen atoms, a(N1) + a(N2).
This possible correlation runs from methanol (2k4 ) 1.2
× 10-3 M-1 s-1, a(N1) + a(N2) ) 17.64 G) to tert-butyl
alcohol (2k4 ) 3.4 × 10-3 M-1 s-1, a(N1) + a(N2) ) 17.82
G) with all the other solvents (except 6 and 7) lying in-
between. A correlation between reactivity and the total
spin density on N1 and N2 would appear reasonable.
We initially assumed that enhanced values of a(N1) +
a(N2) were associated with a solvent-induced reduction
of electron delocalization into the aromatic rings of
DPPH• due to a solvent-induced twisting of one or more
aromatic rings out of conjugation with the SOMO (which
is located principally on the two central nitrogen atoms).
However, a reviewer pointed out that such a solvent-
induced twisting of the aromatic rings could cause a
measurable effect on the electronic spectra of DPPH,
especially in tert-butyl alcohol compared to methanol.
Careful measurements revealed no obvious solvent-
induced spectral effects.22 As an alternative to twisting
of the aromatic rings this reviewer raised the possibility
that tert-butyl alcohol provides a unique solvation shell
around DPPH because of steric crowding between solvent
molecules competing for sites on the DPPH, the idea
being that this unique solvation shell would enhance the
reactivity of DPPH and increase the total spin density
on N1 and N2. In the absence of other explanations we
gratefully accept this proposal.
F igu r e 1. Correlation of values of a(N1) and a(N2) for DPPH•
in various solvents with reported19a values of a(N) for 4-amino-
2,2,6,6-tetramethylpiperidin-N-oxyl in the same solvents (which
are numbered as in Table 1).
bond representation of the nitroxide’s three-electron
bond, i.e., the two canonical forms, D and E, make about
equal contributions to the overall unpaired electron’s
distribution, as is also true for DPPH•, structures A and
B. Polar, polarizable, and hydrogen bonding solvents
stabilize the dipolar form of nitroxides, E, thereby
increasing the spin density on nitrogen. The same is true
for the N2 nitrogen of DPPH• as is clearly brought out by
a plot of the a(N2) values from Table 1 against the a(N)
values measured by Knauer and Napier19a for the 4-amino-
2,2,6,6-tetramethylpiperidin-N-oxyl radical in the same
solvents (see Figure 1).20 Of course, a solvent-induced
increased contribution from the dipolar canonical form,
B, of DPPH• necessarily implies a decreased spin density
on N1. As can be seen in Figure 1, the plot of a(N1) vs
a(N) for Knauer and Napier’s nitroxide has the expected
negative slope.
In conclusion, we have confirmed our earlier observa-
tion6 that the reactivity of DPPH in hydrogen atom
abstractions is significantly enhanced in tert-butyl alco-
hol. With two phenols this enhancement of reactivity
amounts to a factor of about 56 but with 1,4-cyclohexa-
diene the enhancement amounts only to a factor of about
3. We have shown that this reactivity enhancement is
not a general property of alcohols or other hydroxylic
solvents (acetic acid) but seems to be confined to the
sterically more demanding alcohols (tert-butyl alcohol >
2-propanol > n-butanol, with no enhancement in ethanol
and methanol).
The enhanced reactivity of DPPH• in tert-butyl alcohol
was originally suggested to be due to increased spin
density at the divalent nitrogen atom induced by hydro-
gen bonding between tert-butyl alcohol and DPPH• (e.g.,
Exp er im en ta l Section
Ma ter ia ls. Except for DMSO,20 solvents were of the purest
grade commercially available and were used without further
purification. 1,4-Cyclohexadiene (Aldrich 97%) was percolated
twice through activated basic alumina immediately prior to use
to remove the stabilizer (the absence of which was confirmed
by HPLC).
(17) Our results are in satisfactory agreement with the early report
of Garif’yanov et al.14 that a(N1)/a(N2) ) 1.20 and a (N1) + a (N2) )
17.6 ( 0.2 G in benzene (and in toluene and chloroform) and 1.16 and
17.8 ( 0.2 G in methanol. Later, Ryzhmanov and Egorova18 suggested
that solvent effects on a(N1) and a (N2) were due to the formation of a
charge transfer complex between the DPPH• and a solvent molecule,
and these effects were correlated with the ionization potential (IP) of
the solvent. Unfortunately, there are no tabulated data in this
publication, and the correlation is only shown graphically with the
solvents not identified.
(18) Ryzhmanov, Yu. M.; Egorova, A. A. Dokl. Akad. Nauk SSSR
1970, 191, 148-150. English translation, pp 227-229.
(19) (a) Knauer, B. R.; Napier, J . J . J . Am. Chem. Soc. 1976, 98,
4395-4400. (b) Reddock, A. H.; Konishi, S. J . Chem. Phys. 1979, 70,
2121-2130.
(20) The DMSO employed was, unfortunately, an old sample avail-
able from previous work21 which contained some water. Both our EPR
data and kinetics in DMSO were probably influenced by this fact. Thus,
the point for DMSO (6) is not included in Figure 1 because it is much
too imprecise. The point for γ-valerolactone (4) is not included because
we could find no literature value for the nitroxide’s hfcc in this solvent.
The hfcc of the nitroxide in ethyl acetate (3) was not measured by
Knauer and Napier17a but was estimated from data for di-tert-butyl
nitroxide, see footnote f to Table IV of reference 21.
N-[15N]-Nitr osod ip h en yla m in e23 was prepared by reacting
diphenylamine (2.2 g; 13.0 mmol) dissolved in 20 mL of ethanol
with 1.6 mL of concd HCl, immediately followed by Na15NO2
(21) Beckwith, A. L. J .; Bowry, V. W.; Ingold, K. U. J . Am. Chem.
Soc. 1992, 114, 4983-4992.
(22) Band maxima (nm) and absorbance (in parentheses) for 0.25
mM DPPH in various solvents at 20.0 °C. 1, CCl4, 518 (2.30), 330 (3.00);
2, C6H6, 520 (2.45), 328 (3.20); 5, CH3CN, 516 (2.75), 328 (3.75); 6,
DMSO, 524 (2.90), 324 (3.85); 7, AcOH, 516 (2.20), 324 (3.85); 8, MeOH,
516 (3.10), 330 (4.00); 12, t-BuOH, 518 (2.75), 328 (3.75). The only
noticeable solvent effect is in acetic acid in which the ratio of the
UV/visible absorbances is 1.75 vs a range from 1.29 (MeOH) to 1.36
(CH3CN and t-BuOH) in the other solvents.
(23) Chen, M. M.; D’Adamo, A. F. J r.; Walter, R. I. J . Org. Chem.
1961, 26, 2721-2727.

7950 J . Org. Chem., Vol. 61, No. 22, 1996
Notes
99.8% (1.0 g; 14.3 mmol; 1.1 equiv) in 2 mL of water. The
mixture was stirred for 30 min at 0 °C, and the yellow precipitate
was collected, washed with water and (briefly) with cooled (-5
°C) ethanol, and dried by suction at room temperature, yield
81%, mp 68 °C.
EP R Mea su r em en ts. Spectra were recorded at 25 °C on a
Brucker ESR 300 spectrometer equipped with a Varian NMR
gaussmeter and a Systron Donner 6016 frequency counter.
Solutions of DPPH-15N1 in the desired solvent were filtered
directly into EPR tubes and deoxygenated by bubbling with
argon. Measured coupling constants were used to simulate a
low resolution experimental spectrum, and these coupling
constants were then adjusted until satisfactory agreement with
the experimental spectrum was achieved. Measured g-values
were corrected with respect of the known value for DPPH• in
benzene and were confirmed by comparison with the measured
g-value for DPPH• powder.
[
15N2]-1,1-Dip h en ylh yd r a zin e.23,24 N-[15N]-nitrosodiphen-
ylamine (2.1 g; 10.5 mmol) was reduced with 1.5 equiv of LiAlH4,
using the method of inverse addition,25 in anhydrous ether for
1 h at 0-5 °C. After decomposition of the reduced intermediate
(EtOAc 20 mL, wet ether 20 mL, 20% sodium potassium tartrate
70 mL), the organic layer was separated and then combined
with ether extracts of the aqueous phase (3 × 50 mL), dried
(Na2SO4), and concentrated to 100 mL under vacuum. To this
solution was added 1.2 equiv of p-toluenesulfonic acid dissolved
in 30 mL of tert-amyl alcohol. The white precipitate was
collected, washed well with dry ether, and dried at room
temperature, yield 75%, mp 189-190 °C dec. 1H NMR (d6-
DMSO): δ 10.40 (s, NH3+, broad), 2.28 (s, CH3), 7.1 to 7.5
(multiple lines, 14H aromatic). Purity > 97% (by HPLC),
isotopic labeling > 99% (by GC/MS of the recovered free base).
2,2-Dip h en yl-[15N1]-1-p icr ylh yd r a zin e.24 The tosylate salt
(7.9 mmol) and picryl chloride (8.3 mmol; 1.05 equiv) were
dissolved in 50 mL of methanol to which was added 1.0 g of
sodium carbonate in 20 mL of water. The mixture was stirred
for 2 h at room temperature under argon, diluted with water
(10 mL), and cooled (-5 °C). The brick-red precipitate was
collected, washed with methanol/water (5:2), with water, and
dried under vacuum (1 mmHg) at room temperature for 10 h.
The yield of almost pure product was 99%, mp 172-173 °C dec.
1-Ch lor o-2,4,6-tr in itr oben zen e (picryl chloride) was syn-
thesized from 1-hydroxy-2,4,6-trinitrobenzene by standard meth-
ods (POCl3/N,N-diethylamine, 0 °C, 45 min), yield 92%.
Kin etic Mea su r em en ts. A concentrated, deoxygenated
stock solution of 1,4-cyclohexadiene (CHD) was rapidly injected
into a thermostated, deoxygenated solution of DPPH• in the
same solvent in a 10 × 10 mm2 quartz cuvette sealed with a
rubber septum. The cuvette sat within a Hewlett Packard
8425A diode array spectrophotometer and was thermostated at
30 ( 0.1 °C. The decay of DPPH• was followed at 6 different
wavelengths from 510 to 600 nm, including the band maximum
(520 nm). Initial concentrations were typically 1.4 × 10-4
M
DPPH• with (3-15) × 10-2 M CHD. For each solvent, five
measurements were made of ksexptl with different [CHD] and
the absolute second-order rate constants 2ks4 were calculated
by least squares fitting of plot of ksexptl vs [CHD] (r > 0.99).
Ack n ow led gm en t. One of us (L.V.) would like to
thank the “Dottorato di Ricerca in Scienze Farmaceu-
tiche” of the University of Bologna (Italy) for a grant.
We thank an unknown reviewer for his helpful com-
ments, Drs. J . A. Howard and J . R. Morton for access
to the EPR spectrometer and for kind assistance with
computer simulations, Drs. J . T. Banks and M. J onsson
for helpful discussions, and Leanne Bedard for the data
recorded in footnote 22. This work was partially sup-
ported by the Association for International Cancer
Research and the National Foundation for Cancer
Research.
DP P H-15N1 was prepared by stirring the 15N-labeled hy-
23
drazine (7.6 mmol) with PbO2 (35 g) and Na2SO4 (3.0 g) in
chloroform for 2 h at room temperature. The solution was
filtered, collected with several washings of the residue, and
concentrated to 30 mL. The product was completely precipitated
by diluting with hexane (100 mL), collected, and dried (1 mmHg,
40 °C) for 5 h, yield 90%, mp 127-130 °C dec. A sample for the
EPR work was recrystallized from petroleum ether.
(24) O’Connor, S. E.; Walter, R. I. J . Org. Chem. 1977, 42, 577-578.
(25) Poirier, R. H.; Benington, F. J . Am. Chem. Soc. 1952, 74, 3192.
J O9608578