Kinetic study of the hydrogen abstraction reaction of the benzotriazole-N-oxyl radical (BTNO) with H-donor substrates

Article abstract of DOI:10.1021/jo051615n

The aminoxyl radical (>N-O^.) BTNO (benzotriazole-N-oxyl) has been generated by the oxidation of 1-hydroxybenzotriazole (HBT; >N-OH) with a Ce^IV salt in MeCN. BTNO presents a broad absorption band with λ_max 474 nm and e 184

Full text of DOI:10.1021/jo051615n

Kinetic Study of the Hydrogen Abstraction Reaction of the
Benzotriazole-N-oxyl Radical (BTNO) with H-Donor Substrates
Paolo Brandi, Carlo Galli,* and Patrizia Gentili*
Dipartimento di Chimica and IMC-CNR Sezione Meccanismi di Reazione,
Universita` ‘La Sapienza’, 00185 Roma, Italy
carlo.galli@uniroma1.it
Received August 2, 2005
The aminoxyl radical (>N-O•) BTNO (benzotriazole-N-oxyl) has been generated by the oxidation
of 1-hydroxybenzotriazole (HBT; >N-OH) with a CeIV salt in MeCN. BTNO presents a broad
absorption band with λmax 474 nm and ꢀ 1840 M-1 cm-1, and spontaneously decays with a first-
order rate constant of 6.3 × 10-3 s-1 in MeCN at 25 °C. Characterization of BTNO radical by EPR,
laser flash photolysis, and cyclic voltammetry is provided. The spontaneous decay of BTNO is
strongly accelerated in the presence of H-donor substrates such as alkylarenes, benzyl and allyl
alcohols, and alkanols, and rate constants of H-abstraction by BTNO from a number of substrates
have been spectroscopically investigated at 25 °C. The kinetic isotope effect confirms the
H-abstraction step as rate-determining. Activation parameters have been measured in the 15-40
°C range with selected substrates. A correlation between Ea and BDE(C-H) (C-H bond dissociation
energy) for a small series of H-donors has been obtained according to the Evans-Polanyi equation,
giving R ) 0.44. From this plot, the experimentally unavailable BDE(C-H) of benzyl alcohol can
be extrapolated, as ca. 79 kcal/mol. With respect to the H-abstraction step, peculiar differences in
the ∆Sq parameter emerge between an alkylarene, ArC(H)R2, and a benzyl alcohol, ArC(H)(OH)R.
The data acquired on the H-abstraction reactivity of BTNO are compared with those recently
reported for the aminoxyl radical PINO (phthalimide-N-oxyl), generated from N-hydroxyphthalimide
(HPI). The higher reactivity of radical PINO is explained on the basis of the higher energy of the
NO-H bond of HPI, as compared with that of HBT (88 vs ca. 85 kcal/mol, respectively), which is
formed on H-abstraction from the RH substrate.
Transfer of hydrogen atom from C-H bonds to free
radicals in general, and to oxygen-centered radicals in
particular,1 plays an important role in many processes,
including the degradation of chemicals in the tropo-
sphere,2 biochemical reactions,3 synthetic organic trans-
formations,4 but also the pathogenesis of some diseases.5
Removal of hydrogen is instrumental to enable the
reaction of a substrate with dioxygen.6 Aerobic oxidation
of organic compounds operates in many important in-
dustrial processes through the involvement of radical
active species from suitable catalysts.4,7 In this context,
hydroxylamines (>N-OH) as precursors of nitroxyl radi-
cal (>N-O•) catalysts have recently gained considerable
attention. For example, N-hydroxyphthalimide (HPI), in
combination with dioxygen and metal salt cocatalysts
such as Co(OAc)2 or Co(acac)2, provides a remarkable
catalytic system for the oxidation of suitable organic
compounds.4d,8,9 Alcohols, alkylarenes, amides, and even
* Corresponding author. Fax: +39 06 490421.
(1) Olivella, S.; Anglada, J. M.; Sole´, A.; Bofill, J. M. Chem. Eur. J.
2004, 10, 3404-3410.
(4) (a) Simandi, L. Catalytic Activation of Dioxygen by Metal
Complexes; Kluwer Academic Publisher: Boston, 1992. (b) Sibi, M.;
Renaud, P. Radicals in Organic Synthesis; Wiley-VCH: Weinheim,
2001. (c) Arends, I. W. C. E.; Sheldon, R. A. In Modern Oxidation
Methods; Ba¨ckvall, J.-E., Ed.; Wiley-VCH: Weinheim, 2004. (d) Adv.
Synth. Catal. 2004, 346 (Special Issue: Oxidations), 107-375.
(5) (a) Beal, M. F. Ann. Neurol. 1995, 38, 357-366. (b) Rietjens, I.
M. C. M.; den Besten, C.; Hanzlik, R. P.; van Bladeren, P. J. Chem.
Res. Toxicol. 1997, 10, 629-635. (c) McCall, M. R.; Frei, B. Free Radical
Biol. Med. 1999, 26, 1034-1053.
(2) Warnatz, J. In Combustion Chemistry; Gardiner, W. C. Ed.;
Springer: New York, 1984.
(3) (a) Cadenas, E. J.; Davies, K. J. A. Free Radical Biol. Med. 2000,
29, 222-230. (b) Sono, M.; Roach, M. P.; Coulter, E. D.; Dawson, J. H.
Chem. Rev. 1996, 96, 2841-2887. (c) Silverman, R. B. The Organic
Chemistry of Enzyme-Catalyzed Reactions; Academic Press: San Diego,
2000.
10.1021/jo051615n CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/14/2005
J. Org. Chem. 2005, 70, 9521-9528
9521

Brandi et al.
SCHEME 1
CHART 1
alkanes are accordingly oxidized under mild conditions.
The key reactive intermediate in these procedures is the
phthalimide-N-oxyl radical (PINO), formed in the pre-
liminary interaction of HPI with O2 and a Co(II) salt, as
delineated in Scheme 1.8-10
SCHEME 2
The synthetic value of the procedure has prompted
studies on the reactivity features of PINO.10 In recent
investigations, PINO was generated from HPI by the
Pb(OAc)4 oxidant and the resulting >N-O• species
characterized by UV-vis spectrophotometry.11 Kinetic
studies of H-abstraction from suitable C-H- or O-H-
containing substrates have provided fundamental infor-
mation upon the reactivity of PINO toward organic
compounds.11,12 Knowledge of the energy of the NO-H
bond of HPI itself,9 as well as that of aryl-substituted
HPI’s, from EPR equilibration measurements, and of
their redox potentials from cyclic voltammetry12 has
provided additional information upon the reactivity
features of this remarkable oxidizing system.
Concerns about environmental pollution urge substitu-
tion of old synthetic procedures with “green” alterna-
tives.4c,d Catalytic systems enabling the activation of di-
oxygen for homogeneous oxidations represent one effort
in this direction.4 A natural process of fundamental im-
portance is the oxidative delignification of rotten wood
carried out by white-rot fungi.13 A similar process would
represent a dream for the paper-making industry that
conventionally performs the delignification of wood pulp
by utilizing chlorine-based polluting chemicals.14 Deligni-
fication with fungal enzymes might lead to novel, environ-
ment-respectful techniques.13a Attempts in this direction
have been reported,13,15 but the degradation of difficult
to oxidize residues of lignin, for example, the benzylic
alcohol groups, may require the widening of the sub-
strate range of some enzymes by adding mediator com-
pounds.13,15,16 Hydroxylamines, such as HPI or 1-hydroxy-
benzotriazole (HBT) or violuric acid (VLA) (Chart 1), have
recently shown their proficiency in mediating the phenol
oxidase laccase toward the aerobic oxidation of lignin
models having the structure of benzyl alcohols.13c,15a-c,16
Laccase, having a redox potential around 0.6-0.8 V
vs NHE,16c could never oxidize benzylic substrates en-
dowed with redox potentials g1.4 V vs NHE by electron
abstraction.17 In contrast, laccase/mediator systems have
shown promises even in the degradative oxidation of
wood pulp itself.18 The nitroxyl radical intermediate from
these >N-OH mediators, generated by laccase, performs
as the active species and promotes the radical cleavage
of benzylic C-H bonds, thereby extending the (indirect)
oxidation capability of laccase toward nonphenolic
substrates.17a This lead us back to the studies of the
H-abstraction reactivity by >N-O• species. We have
recently reported on the first generation of the aminoxyl
radical from HBT (dubbed BTNO) by monoelectronic
oxidation with cerium(IV) ammonium nitrate (i.e., CAN)
in MeCN solution (Scheme 2).19
(6) (a) Otamiri, T.; Sjodahl, R. Cancer 1989, 64, 422-425. (b)
Berliner, J. A.; Heinecke, J. W. Free Radical Biol. Med. 1996, 20, 707-
727. (c) Chatgilialoglu, C.; Ferreri, C. Acc. Chem. Res. 2005, 38, 441-
448.
(7) (a) Sheldon, R. A.; Kochi, J. K. Metal-Catalyzed Oxidations of
Organic Compounds; Academic Press: New York, 1981. (b) The
Activation of Dioxygen and Homogeneous Catalytic Oxidation; Barton,
D. H. R., Martell, A. E., Sawyer, D. T., Eds.; Plenum Press: New York,
1993. (c) Sheldon, R. A.; Arends, I. W. C. E.; ten Brink, G.-J.; Dijksman,
A. Acc. Chem. Res. 2002, 35, 774-781. (d) Whittaker, J. W. Chem. Rev.
2003, 103, 2347-2364.
(8) (a) Iwahama, T.; Yoshino, Y.; Keitoku, T.; Sakaguchi, S.; Ishii,
Y. J. Org. Chem. 2000, 65, 6502-6507. (b) Ishii, Y.; Sakaguchi, S.;
Iwahama, T. Adv. Synth. Catal. 2001, 343, 393-427.
(9) (a) Minisci, F.; Punta, C.; Recupero, F.; Fontana, F.; Pedulli, G.
F. J. Org. Chem. 2002, 67, 2671-2676. (b) Amorati, R.; Lucarini, M.;
Mugnaini, V.; Pedulli, G. F.; Minisci, F.; Recupero, F.; Fontana, F.;
Astolfi, P.; Greci, L. J. Org. Chem. 2003, 68, 1747-1754.
(10) Minisci, F.; Recupero, F.; Cecchetto, A.; Gambarotti, C.; Punta,
C.; Faletti, R.; Paganelli, R.; Pedulli, G. F. Eur. J. Org. Chem. 2004,
109-119.
(11) (a) Koshino, N.; Saha, B.; Espenson, J. H. J. Org. Chem. 2003,
68, 9364-9370. (b) Koshino, N.; Cai, Y.; Espenson, J. H. J. Phys. Chem.
A 2003, 107, 4262-4267. (c) Saha, B.; Koshino, N.; Espenson, J. H. J.
Phys. Chem. A 2004, 108, 425-431. (d) Cai, Y.; Koshino, N.; Saha, B.;
Espenson, J. H. J. Org. Chem. 2005, 70, 238-243.
(12) Annunziatini, C.; Gerini, M. F.; Lanzalunga, O.; Lucarini, M.
J. Org. Chem. 2004, 69, 3431-3438.
A full account of the generation and H-abstraction
reactivity of this radical intermediate toward substrates
containing C-H bonds of suitable energy is presented
here. Because HBT is one of the most efficient mediators
(15) (a) Call, H. P.; Mu¨cke, E. I. J. Biotechnol. 1997, 53, 163-202.
(b) Bourbonnais, R.; Paice, M. G. FEBS Lett. 1990, 267, 99-102. (c)
Bourbonnais, R.; Paice, M. G.; Freiermuth, B.; Bodie, E.; Borneman,
S. Appl. Environ. Microb. 1997, 63, 4627-4632. (d) Gamelas, J. A. F.;
Tavares, A. P. M.; Evtuguin, D. V.; Xavier, A. M. B. J. Mol. Catal. B:
Enzym. 2005, 33, 57-64.
(16) (a) ten Have, R.; Teunissen, P. J. M. Chem. Rev. 2001, 101,
3397-3413. (b) Li, K.; Xu, F.; Eriksson, K.-E. L. Appl. Environ.
Microbiol. 1999, 65, 2654-2660. (c) Xu, F.; Kulys, J. J.; Duke, K.; Li,
K.; Krikstopaitis, K.; Deusse, H.-J.; Abbate, E.; Galinyte, V.; Schneider,
P. Appl. Environ. Microbiol. 2000, 66, 2052-2056. (d) Fabbrini, M.;
Galli, C.; Gentili, P. J. Mol. Catal. B: Enzym. 2002, 16, 231-240. (e)
Baiocco, P.; Barreca, A. M.; Fabbrini, M.; Galli, C.; Gentili, P. Org.
Biomol. Chem. 2003, 1, 191-197.
(17) (a) Galli, C.; Gentili, P. J. Phys. Org. Chem. 2004, 17, 973-
977. (b) Cantarella, G.; Galli, C.; Gentili, P. J. Mol. Catal. B: Enzym.
2003, 22, 135-144.
(18) (a) Barreca, A. M.; Fabbrini, M.; Galli, C.; Gentili, P.; Ljunggren,
S. J. Mol. Catal. B: Enzym. 2003, 26, 105-110. (b) Barreca, A. M.;
Sjo¨gren, B.; Fabbrini, M.; Galli, C.; Gentili, P. Biocatal. Biotrans. 2004,
22, 105-112. (c) Annunziatini, C.; Baiocco, P.; Gerini, M. F.; Lanza-
lunga, O.; Sjo¨gren, B. J. Mol. Catal. B: Enzym. 2005, 32, 89-96.
(19) Galli, C.; Gentili, P.; Lanzalunga, O.; Lucarini, M.; Pedulli, G.
F. Chem. Commun. 2004, 2356-2357.
(13) (a) Kuhad, R. C.; Eriksson, K.-E. L. Microorganisms and
Enzymes Involved in the Degradation of Plant Fiber Cell Wall. In
Biotechnology in the Pulp and Paper Industry (Advances in Biochemical
Engineering Biotechnology); Eriksson, K.-E. L., Ed.; Springer: Berlin,
1997; Chapter 2; Vol. 57. (b) Reid, I. D.; Paice, M. G. FEMS Microb.
Rev. 1994, 13, 369-376. (c) Messerschmidt, A. Multi-Copper Oxidases;
World Scientific: Singapore, 1997.
(14) Roberts, J, C. The Chemistry of Paper; The Royal Society of
Chemistry: Cambridge, 1996.
9522 J. Org. Chem., Vol. 70, No. 23, 2005

Hydrogen Abstraction by BTNO from H-Donor Substrates
TABLE 1. Normalized Second-Order Rate Constant of H-Abstraction from H-Donor Substrates by BTNO at 25 °Ca,b
substrate, RH
(no. of equivalent
removable H atoms)
kH (M-1 s-1
)
BDE(C-H)
of RH
kH (M-1 s-1) at
at 25 °C
25 °C in AcOH
in MeCNc,d
(kcal/mol)e
with PINO radicalf
C6H5CH2OH (2)
0.94
0.57k
3.1
83-85g
83-85g
82-84g
82-84g
82-84g
82-84g
93
-
C6H5CH2OH (2)
4-MeO-C6H4CH2OH (2)
4-MeO-C6H4CH2OH (2)
4-MeO-C6H4CH2OMe (2)
3,4-diMeO-C6H3CH2OH (2)
decanol (2)
5.7
-
3.4k
3.0
22
-
11
-
0.11
0.33
0.091
0.35
0.55
5.9
-
cycloesanol (1)
92
-
4-MeO-toluene (3)
4-MeO-ethylbenzene (2)
4-MeO-isopropylbenzene (1)
1-(4-MeO-phenyl)ethanol (1)
PhCHdCHCH2OH (2)
PhOH (1)
89-90
85-87
83-85
82-84
80-83
88h
3.4
-
-
-
18
-
66
330i
-
9-Ph-fluorene (1)
fluorene (2)
9,9-di-D-fluorene (2)
Ph3CH (1)
9-OH-fluorene (1)
Ph2CHOH (1)
Ph2CH2 (2)
p-xylene (6)
p-An2CHOH (1)
geraniol (2)
PhCH2NEt2 (2)
18
74
1.9
82
20
-
59
-
0.14j
2.3
-
81
10
3.2
73l
79
58
6.6
0.99
-
0.36
ca. 0.02
19
84
89-90
75
11
82
-
too fast
<0.001
ca. 0.01
ca. 0.8
85
-
PhCH2COCH3 (2)
PhCH2CN
PhCHO (1)
82
-
-
-
87
11
a The radical was generated from HBT with CAN in MeCN solution. Literature values of the normalized H-abstraction rate constants
by PINO radical, generated from HPI with Pb(OAc)4 in AcOH solution at 25 °C, are given. b Conditions: [BTNO], 0.5 mM; [RH], 5-25
mM in MeCN solution. Determinations in triplicate; typical error (3%. c Monitored spectroscopically at 474 nm. d Normalized for the
number of equivalent H-atoms in RH. e From ref 26. f Followed at 382 nm; from ref 11. g A range of values from different sources and
techniques is reported in ref 26. h BDE(O-H). i Our determination. j As kD. k In AcOH. l For 9-MeO-fluorene.
tion of the substrate was in the (0.5-2) × 10-2 M range, to
enable a pseudo-first-order treatment of the kinetic data. Plots
of (At - A∞) vs time were well-fitted by a first-order exponential
over more than three half-lives, and the rate constant (k′)
obtained. From a plot of three to five k′ vs [subst]0 data pairs,
the second-order rate constant of H-abstraction (kH) was
obtained. Determination of H-abstraction rate constants was
analogously carried out at different temperatures (three
temperatures in the 13-43 °C range) with selected substrates
and enabled calculation of the activation parameters from
Arrhenius and Eyring plots.
Laser Flash Photolysis Experiment. Flash photolysis
experiments were carried out with a laser spectrometer using
the fourth harmonic (266 nm) of a Q-switched Nd:YAG laser.
A 3 mL Suprasil quartz cell (10 mm × 10 mm) was used for
all the experiments. Argon-saturated MeCN solutions contain-
ing di-tert-butyl peroxide (1.0 M) and HBT (10 mM) were used.
The experiments were carried out at 25 °C under magnetic
stirring.
Isotope Effect Determinations. As described pre-
viously,16e,19 R-monodeuterated benzyl alcohol, PhCH(D)OH,
was oxidized with BTNO under the following experimental
conditions: 60 mmol alcohol, 20 mmol HBT, and 20 mmol
CAN, in 3 mL of MeCN at room temperature. After a 5 h
reaction time and conventional workup with ethyl acetate, the
relative amount of the PhCHO and PhCDO oxidation products
was determined by GC-MS analyses in the SIM mode,
enabling reckoning of the intramolecular kH/kD ratio. The
intermolecular kinetic isotope effect for fluorene vs 9,9-
dideuteriofluorene was calculated from the ratio of the kH and
kD rate constants with BTNO (see Table 1).
of laccase,17 our study may have relevance for the field
of biodelignification as well.13,15
Experimental Section
Materials. HBT (Aldrich) was used as received; all the
substrates were obtained commercially and used without
further purification. Cerium(IV) ammonium nitrate (CAN;
Erba RPE) was oven-dried before use. Reagent grade glacial
acetic acid and acetonitrile (Erba RPE) were used as solvents.
Mono-R-deuterated benzyl and p-MeO-benzyl alcohols were
already available in the laboratory from previous studies.16e
9,9-Dideuterated fluorene was obtained from fluorene by
means of hydrogen-deuterium exchange with t-BuOK in
(CD3)2SO-D2O, and its electron impact mass spectrum (m/e
168) was consistent with the replacement of two H-atoms with
two D-atoms.
Instrumentation. The kinetic study was carried out with
a Hi-Tech SFA-12 stopped flow instrument interfaced to a HP
8453 diode array spectrophotometer and having a thermo-
stated cell holder. A conventional UV-vis spectrophotometer
(Perkin-Elmer Lambda 18) was alternatively used. A thermo-
couple was employed for reading the temperature in the
cuvette.
Kinetic Procedure. The BTNO species was generated in
MeCN in a thermostated quartz cuvette, 1 cm optical path,
by adding a 0.5 mM solution of CAN to a 0.5 mM solution of
HBT. A broad absorption band developed almost immediately
(15 ms) in the 350-600 nm region (λmax at 474 nm, ꢀ 1840 M-1
cm-1).19 The A474 reading was not stable, but decayed with a
half-life of 110 s. The rate of decay was unaffected by the use
of deareated MeCN. Rate constants of H-abstraction from
H-donor substrates were determined at 25 °C in MeCN, by
following the decrease of A474 of BTNO. The initial concentra-
Electrochemical Determinations. Cyclic voltammetry at
the steady disk electrode (glassy carbon disk, 1.5 mm in
diameter)16d was carried out at 25°C in MeCN containing 0.1
J. Org. Chem, Vol. 70, No. 23, 2005 9523

Brandi et al.
FIGURE 2. Self-decomposition of BTNO, generated from
FIGURE 1. UV-vis spectrum of HBT (0.5 mM) in MeCN in
the absence of CAN (b) and in the presence of CAN (0.5 mM)
(1) 15 ms after mixing and (9) 90 s after mixing.
[HBT] ) [CAN] ) 0.5 mM in MeCN. Readings were at A474
;
the half-life is 110 s.
M Bu4NBF4. A three-electrode circuit was used with either an
electrochemical computerized system (Amel System 5000,
version 2.1) or a potentiostat with positive feedback ohmic-
drop compensation and MPLI hardware (Vernier Software &
Technology) controlled by a program written in C++ language
for Windows 95/98. The auxiliary electrode was a platinum
wire (surface 1 cm2), and a saturated calomel electrode (SCE)
was used as reference; the redox potential values are then
referred to NHE. The sweep scan ranged from 10 mV/s to 10
V/s; the concentration of HBT was 1.7 mM, and a 2-fold excess
of pyridine was used to deprotonate the >N-OH mediator into
the more easily oxidizable >N-O- form.
SCHEME 3
As a further characterization, laser flash photolysis of
dicumyl peroxide at 355 nm in MeCN solution generates
the cumyloxyl radical,19 which abstracts a H-atom from
HBT (Scheme 3), giving rise to the absorption spectrum
of Figure 3:
An absorption profile with the same λmax at 474 nm as
that reported in Figure 1 can be recognized. In a new
experiment, we generate the t-BuO. radical by laser flash
photolysis (LFP) of (t-BuO)2 at 266 nm in MeCN at 25 °C
and, in the presence of HBT, an absorption spectrum per-
fectly consistent with that of Figure 3 is obtained. There-
fore, the species resulting from H-abstraction from HBT
(Figure 3) by either PhCMe2O• or t-BuO• and that result-
ing from monoelectronic oxidation of HBT (Figure 1) with
CAN do have the same spectral features, thus supporting
the nature of an aminoxyl radical for both of them
(Scheme 4). We have dubbed BTNO (i.e., benzotriazole
nitroxyl radical) the radical species arising from HBT.19
The redox potential of HBT had been previously
determined by cyclic voltammetry (CV) in bufferd water
(pH 5) as E° 1.08 V vs NHE.16d In MeCN solution, a
reversible monoelectronic peak is obtained at a sweep
Results and Discussion
Characterization of the BTNO Radical. In keeping
with the characterization of the PINO radical from HPI,11
we provide here a full characterization of the aminoxyl
radical from HBT. The absorption spectrum of a 0.5 mM
solution of HBT in MeCN is reported in Figure 1 (trace
b).
Following the addition of a stoichiometric amount of
the monoelectronic oxidant cerium(IV) ammonium ni-
trate (i.e.,. CAN) in MeCN at 25 °C, a new and broad
absorption band in the 350-600 nm range develops (trace
1), having λmax at 474 nm and ꢀ 1840 M-1 cm-1 19
.
Abstraction of an electron from HBT (E° 1.08 V vs
NHE)16d by CAN (E° 1.3 V vs NHE)20 is exoergonic and
occurs quantitatively. A practically superimposable UV-
vis spectrum is obtained when using Pb(OAc)4 as the
oxidant of HBT, in either AcOH or MeCN solutions.19 In
comparison, PINO radical has λmax at 382 nm with ꢀ 1360
M-1 cm-1 in AcOH.11
The spectrum of the oxidized species from HBT, both
in MeCN and AcOH solution, is not stable (Figure 2), and
the reading of the λ474 absorption decays according to an
exponential curve that is almost clean first-order (kdecay
) 6.3 × 10-3 s-1 in MeCN at 25 °C) with a half-life of
110 s (20 s in AcOH solution). Espenson11 and Masui21
had advanced hypotheses about the self-decomposition
of PINO radical, which is second-order and has the first
half-life of 7900 s at 25 °C, at [HPI]0 ) 0.2 mM in AcOH.
The nature of the self-decomposition of PINO is not
precisely known yet, and we have not investigated the
self-decomposition of the oxidized species from HBT any
further.
FIGURE 3. Absorption spectrum obtained after 355-nm laser
flash photolysis of an Ar-saturated MeCN solution containing
1.0 M dicumyl peroxide and 0.03 M HBT, recorded 1.4 ms after
the laser flash (from ref 19).
(20) Prabhakar Rao, G.; Vasudeva Murthy, A. R. J. Phys. Chem.
1964, 68, 1573-1576.
(21) Ueda, C.; Noyama, M.; Ohmori, H.; Masui, M. Chem. Pharm.
Bull. 1987, 35, 1372-1377.
9524 J. Org. Chem., Vol. 70, No. 23, 2005

Hydrogen Abstraction by BTNO from H-Donor Substrates
SCHEME 5
protons (aH ) 0.037, 0.203, and 0.458 mT), consistently
with the structure of the BTNO radical, where two out
of the four hydrogen atoms have almost the same
hyperfine coupling constant.
Unfortunately, due to the short lifetime of BTNO, EPR
equilibration experiments with species such as phenols
having removable H-atoms and known bond dissociation
energy BDE(O-H) values9b,23 was unfeasible. Determi-
nation of the BDE(NO-H) of HBT is thus lacking yet,
whereas the corresponding value of BDE(NO-H) for HPI
was acquired as 88.1 kcal/mol by EPR equilibration ex-
periments.9b,12 Semiempirical calculations had previously
provided values around 75-80 kcal/mol for the BDE-
(NO-H) of HBT,18b,24 whereas we have recently attained
a higher value (i.e., 85 kcal/mol) from theoretical calcula-
tions (DFT) and also from a thermochemical cycle.25
Hydrogen-Abstraction Reactivity of BTNO Radi-
cal. The spontaneous decay of BTNO, followed at λ474, is
strongly accelerated in the presence of H-donor sub-
strates such as alkylarenes, benzyl or allyl alcohols, and
alkanols, and the H-abstraction reactivity of BTNO from
a number of substrates has been investigated here at 25
°C (Scheme 5)
FIGURE 4. Cyclic voltammetry in MeCN solution; [HBT] )
1.65 mM, [Py] ) 3.3 mM.
BTNO was generated in the spectrophotometric cu-
vette at an initial 0.5 mM concentration in MeCN, by
quickly mixing stoichiometric amounts of HBT and CAN
solutions at 25 °C. The initial reading of A474 confirmed
quantitative formation of BTNO. Fast addition by syringe
of a solution of the C-H bearing substrate, at an initial
concentration 10-50 times higher than that of BTNO,
marked the beginning of the kinetic experiment. The
progress of the reaction was monitored at 474 nm by
stopped-flow or conventional spectrophotometers, the
half-lives of the substrates ranging from 15 ms to tenths
of seconds, respectively. Under these pseudo-first-order
conditions, the H-abstraction rate was faster than the
spontaneous decay of BTNO in all cases reported in Table
1, while product analysis had already confirmed the
formation of the carbonyl derivative from selected pre-
cursors.17,19 No investigation was possible with more
reactive H-donor substrates, as the reaction would be too
fast to follow with our experimental technique, nor with
less reactive substrates, whose “apparent” first-order rate
constant matched the k(decay) of BTNO.
FIGURE 5. Experimental EPR spectrum of BTNO, obtained
at room temperature by reaction of CAN (20 mM) with HBT
(20 mM) in MeCN (a), and its computer simulation (b) (from
ref 19).
SCHEME 4
(a) Rate Constants of H-Abstraction at 25 °C. The
rate constants k′, determined at 25 °C under pseudo-first-
order conditions (excess of substrate, at three or four
different initial concentrations) from at least duplicated
experiments, were converted into second-order H-ab-
straction rate constants (kH) by determining the slope of
a k′ vs [subst]0 plot, as shown in Figure 6 for 4-MeO-
benzyl alcohol. The typical uncertainty of the kinetic
determinations ranged from 2 to 5%.
scan of 0.5 V/s (E° 0.97 V vs NHE; Figure 4) in the
presence of a 2-fold molar amount of pyridine. This
indicates one-electron oxidation of the conjugate base of
HBT (pKa 4.6)22 to BTNO (cf. Scheme 4).
Finally, an EPR spectrum of BTNO was obtained by
mixing stoichiometric amounts of CAN and HBT in
MeCN at 25 °C.19 This spectrum (Figure 5) was inter-
preted on the basis of the coupling of the unpaired
electron with three different nitrogen nuclei (aN ) 0.056,
0.152, and 0.475 mT), and with three unequivalent
Intercept values in these plots confirm the order of
magnitude of the self-decay rate constant of BTNO. The
kH values reported in Table 1 for all the substrates
(23) (a) Guerra, M.; Amorati, R.; Pedulli, G. F. J. Org. Chem. 2004,
69, 5460-5467. (b) Pratt. D. A.; Dilabio, G. A.; Mulder, P.; Ingold, K.
U. Acc. Chem. Res. 2004, 37, 334-340.
(22) Koppel, I.; Koppel, J.; Leito, I.; Pihl, V.; Grehn, L.; Ragnarsson,
U. J. Chem. Res. (S) 1993, 446-447.
J. Org. Chem, Vol. 70, No. 23, 2005 9525

Brandi et al.
If we compare the different kH’s obtained with BTNO
and PINO, the substrate being equal, the rate values
with the latter radical are always higher (Table 1). The
higher reactivity of PINO ought to reflect the higher BDE
of the NO-H bond in precursor HPI with respect to HBT
(88 vs ca. 85 kcal/mol, respectively).9b,25 This difference
is significant for a radical H-abstraction process by a
>N-O• species: the stronger the O-H bond being
formed, the faster the H-abstraction step. With other
hydroxylamines, we plan to further investigate this
comparison between H-abstraction reactivity of the >N-
O• intermediate and the BDE(NO-H) value of the >N-
OH species. Finally, in a literature report on the elec-
trochemical generation of BTNO from HBT, the kH of
reaction with 3,4-diMeO-C6H3CH2OH (i.e., veratryl al-
cohol) is given as 2.5 M-1 s-1 in buffered water solution.27
This value, when compared with our own determination
(i.e., 11 M-1 s-1, in Table 1) in MeCN, supports a similar
reactivity of the BTNO species once generated by two
independent methods in two different media. Future
work on the comparison of reactivity between the elec-
trochemical vs chemical (by CAN or other oxidants)
method of generation of the >N-O• species is planned.
(b) Mechanistic Insight. Additional clues confirm the
radical nature of the reaction of BTNO with the RH
substrates of Table 1. The Hammett correlation for
reaction of BTNO with p-substituted benzyl alcohols gave
a F of -0.55 in MeCN,19 a value reasonably small for a
radical process.28 It compares with a F value of -0.41 for
PINO vs the same substrates, when PINO was generated
from HPI with Pb(OAc)4 in AcOH solution.11a Additional
F values in the -0.5 to -0.7 range were obtained with
the aminoxyl radicals generated from X-aryl-substituted
HPIs according to the Ishii procedure (i.e., X-HPIs/Co(II)/
MCBA/O2)8 in MeCN.12 In all these cases, better Ham-
mett correlations were obtained vs σ+ values, in agree-
ment with the electrophilic character of the aminoxyl
radicals.9a,10,28 Thus, a uniform pattern of selectivity
among these structurally similar >N-O• species emerges.
Lastly, firm evidence in favor of the intervention of a
benzyl radical during the oxidation of benzyl alcohols
with BTNO has been recently acquired through the
interception of the radical intermediate by the use of a
cyclizable probe substrate.25
FIGURE 6. Pseudo-first-order (k′) rate constants of H-
abstraction from 4-MeO-benzyl alcohol by BTNO, in MeCN
at 25 °C. The slope of the plot gives the kH value.
investigated are normalized for the number of equivalent
hydrogens.11 Available values of the BDE(C-H) of the
RH substrates are given.11,26 For a couple of substrates,
the kH rate constant was determined also in AcOH
solution, CAN being the oxidant, to enable a better
comparison with the reactivity data11 of PINO presented
in the table.
Some comments on the rate constants of Table 1 are
possible. First of all, the matching value of kH for 4-MeO-
C6H4CH2OH and 4-MeO-C6H4CH2OMe confirms beyond
any doubt that BTNO removes the H-atom from the
benzylic C-H position and not from the O-H bond of
the alcohol, in keeping with the much higher BDE (ca.
104 kcal/mol) of the latter.26 Second, H-abstraction by
BTNO becomes increasingly faster on going from an
alkanol (decanol, 0.11 M-1 s-1) to benzyl alcohol (0.94 M-1
s-1) and further on to an allyl alcohol (geraniol, 11 M-1
s-1; cinnamyl alcohol, 18 M-1 s-1), thereby reflecting the
BDE(C-H) of 93 kcal/mol, ca. 84 and ca. 82 M-1 s-1
,
respectively,26 approximate as these values are. Analo-
gously, increasing the number of Ph-substituents of the
C-H group undertaking cleavage causes an increase of
kH value from ca. 0.02 M-1 s-1 in p-xylene (ArCH3) to 0.33
M-1 s-1 in Ph2CH2 up to 2.2 M-1 s-1 in Ph3CH; this
reflects the trend of BDE(C-H) values of ca. 90, 84, and
81 kcal/mol, respectively.26 Similarly, PhCH2OH (one Ph)
vs Ph2CHOH (two Ph) gives kH values (in M-1 s-1) of 0.94
vs 3.2, and fluorene vs 9-Ph-fluorene analogously gives
1.7 vs 18, in line with the corresponding BDE(C-H)
data.26 Although it is reasonable to expect that rate
constants and thermochemical data do correlate in a
radical process for stucturally similar compounds, it is
appropriate to point out that this comparison is dimen-
sionally not homogeneous, because the reactivity data are
free energies (∆Gq), whereas the BDE(C-H) data are
enthalpic values. A more homogeneous correlation has
been sought, and found, through the determination of the
activation parameters for the H-abstraction process of
BTNO with selected substrates (vide infra). We simply
observe at this stage that RH substrates endowed with
BDE higher than 92-94 kcal/mol are almost unreactive
with BTNO. Finally, the solvent (MeCN vs AcOH) does
not affect the value of kH substantially (Table 1), and this
is conceivable for a radical process.
In the reaction of BTNO with p-MeOC6H4CH(D)OH in
MeCN at room temperature, a kH/kD ratio (intramolecu-
lar) of 5.6 from product analysis was obtained.19 An
additional experiment carried out here gives an intramo-
lecular kH/kD value of 7 from product analysis for the
reaction of BTNO with PhCH(D)OH. Moreover, an in-
termolecular kH/kD value of 13 from the ratio of the
kinetic rate constants of fluorene vs 9,9-dideuteriofluo-
rene with BTNO (Table 1) is consistent with a rate-
determining H-abstraction step. The latter kH/kD value
of 13 supports the contribution of tunelling, as already
commented on for the reactions of PINO with various
H-donors (kH/kD values in the 11-27 range).11a,12,21 Fi-
nally, the relative reactivity of abstraction of a primary
vs secondary vs tertiary benzylic hydrogen by BTNO,
reckoned from comparing the rate constants of MeO-
(24) Sealey, J.; Ragauskas, A. J.; Elder, T. J. Holzforschung 1998,
53, 498-502.
(25) Astolfi, P.; Brandi, P.; Galli, C.; Gentili, P.; Gerini, M. F.; Greci,
L.; Lanzalunga, O. New J. Chem. 2005, 29, 1308-1317.
(26) Luo, Y.-R. Handbook of Bond Dissociation Energies in Organic
Compounds; CRC Press: Boca Raton, FL, 2003.
(27) Bourbonnais, R.; Leech, D.; Paice, M. G. Biochim. Biophys. Acta
1998, 1379, 381-390.
(28) d’Acunzo, F.; Baiocco, P.; Fabbrini, M.; Galli, C.; Gentili, P. New
J. Chem. 2002, 26, 1791-1794.
9526 J. Org. Chem., Vol. 70, No. 23, 2005

Hydrogen Abstraction by BTNO from H-Donor Substrates
TABLE 2. Rate Constants and Activation Parameters for the H-Abstraction Reaction by BTNO from Selected RH
Substrates at Various Temperaturesa
temp
(°C)
((0.2)
Ea
log A
(M-1 s-1
((0.5)c
BDE(C-H)
of RH
kH
(kcal/mol)
)
∆Hq(kcal/mol)
((0.15)d
∆Sq(eu)
((1)d
b
substrate (RH)
(M-1 s-1
)
((0.15)c
(kcal/mol)e
4-MeO-C6H4CH2OH
18.5
29.8
40.8
18.3
29.7
41.5
20.7
30.1
41.3
18.5
30.0
41.2
18.1
29.5
40.9
13.1
22.7
30.9
18.2
29.5
40.6
18.8
29.3
40.3
18.2
29.5
40.8
2.48
3.36
5.75
2.86
3.78
5.66
0.52
0.64
0.88
2.05
3.24
4.93
2.47
3.71
5.70
12.5
16.0
21.1
0.40
1.01
1.46
1.63
2.52
4.75
1.15
2.35
3.48
6.8
5.3
4.8
7.0
6.7
5.1
10.6
8.8
8.9
13
10
7.5
13
12
11
17
16
16
6.2
4.7
4.2
6.4
6.0
4.5
10.0
8.2
8.3
-36
-40
-46
-35
-36
-38
-26
-29
-29
-
4-MeO-C6H4CH2OH (in AcOH)
PhCH2OH (in AcOH)
4-MeO-C6H4CH2OMe
Ph2CHOH
79
9-OH-FLf
ca. 73
84
Ph2CH2
Ph3CH
81
FLf
82
a Initial conditions as in Table 1. b Normalized values. Runs in triplicate; typical error (3%. c From the Arrhenius equation. d From
the Eyring equation. e From ref 26. f FL stands for fluorene.
C6H4CH3 (taken as 1.0 M-1 s-1) vs MeO-C6H4CH2Me (4.2
M-1 s-1) vs MeO-C6H4CHMe2 (6.3 M-1 s-1) in Table 1, is
in keeping with the BDE(C-H) values (88, 85, and 84
kcal/mol, respectively)26 and perfectly in line with the
relative rates (M-1 s-1) of 1:3.2:6.8 reported for the
analogous reactions with t-BuO..29 The trend is confirmed
by the relative rates (M-1 s-1) of 1:4.4:9.7 similarly
obtained with the phenyl radical.29 All this evidence
supports a rate-determining H-abstraction process by
BTNO toward H-donor substrates.
(c) Activation Parameters. Rate constants of H-
abstraction from a few RH substrates by BTNO have
been measured at various temperatures. The substrates
were selected according to two requisites: (i) the BDE-
(C-H) of RH had to be known with reasonable accuracy,
and (ii) the kH at 25 °C had to be not so fast to let a too
fast rate constant be foreseen at higher temperature(s)
for our stopped flow device (t1/2 not < 15 ms). The
thermostated cell holder of the stopped flow allowed
investigating the rate constants in the 10-50 °C range.
At least three temperatures, spaced approximately 12-
15 °C between them, were tested for each substrate. As
explained before for the kinetic determinations at 25 °C
(Table 1 and Figure 6), each second-order kH value
derives from averaging three or four pseudo-first-order
rate constants determined at different initial concentra-
tions of RH. The initial concentration of BTNO was 0.2-
0.5 mM, and that of RH ranged between 5 and 20 mM.
Once again BTNO was generated by quickly mixing HBT
and CAN solutions in stoichiometric amounts, and the
progress of the reaction was monitored at 474 nm.
The second-order rate constants (kH), normalized for
the number of equivalent hydrogen atoms in RH, are
reported in Table 2 for each RH at the various temper-
atures investigated. From the Arrhenius equation (log
kH ) log A - Ea/RT), the log A and Ea parameters were
obtained, whereas from the Eyring equation (log kH/T )
log kB/h - ∆Hq/RT + ∆Sq/R), the ∆Hq and ∆Sq were
obtained.
A proper correlation between reactivity of H-abstrac-
tion and thermochemical data can now be attempted
according to the Evans-Polanyi equation (Ea ) R BDE-
(C-H) + cost)30 (Figure 7), because both the attained Ea
parameters and the BDE(C-H) data of the RH sub-
strates have the dimensions of enthalpy. Within the
limited number of substrates that lend themselves to this
analysis under our conditions, it can be quantitatively
verified that the reactivity of H-abstraction follows the
C-H bond energies linearly. The reaction of BTNO is
increasingly slower (i.e., higher Ea value) toward increas-
ingly stronger C-H bonds, and the slope of the correla-
tion is R ) 0.44.
The R parameter indicates the selectivity of BTNO; a
value of 0.44 suggests that the H-atom being removed is
almost equally shared between the aminoxyl radical
abstracting it and the C-atom that is losing it. The NO-H
bond in formation and the C-H bond being cleaved must
have therefore a comparable energy (vide infra). The R
of PINO, reported by Espenson et al. as 0.38,11b is slightly
lower in value than that of BTNO; this is likely due to
the fact that the BDE(NO-H) of HPI is higher (88 kcal/
mol)9b,12 than that of HBT (85 kcal/mol),25 thereby
(29) Baciocchi, E.; d’Acunzo, F.; Galli, C.; Lanzalunga, O. J. Chem.
Soc., Perkin Trans. 2 1996, 133-140.
(30) Evans, M. G.; Polanyi, M. Trans. Faraday Soc. 1938, 34, 11-
24.
J. Org. Chem, Vol. 70, No. 23, 2005 9527

Brandi et al.
substrate is responsible for that depression of ∆Sq with
respect to the RH counterpart. We plan to investigate
this comparison of C-H vs C-OH reactivity with other
>N-O• species in the future.
From the linear Evans-Polanyi plot obtained (Figure
7), and from the experimental Ea value determined for
4-MeO-C6H4CH2OH (i.e., 6.8 kcal/mol), we extrapolate a
BDE value of 77 ( 1 kcal/mol for the benzylic C-H bond
of this specific substrate undergoing cleavage. This
number, which is not experimentally available or re-
ported with reasonable confidence for benzyl alcohols,26
compares well with a BDE(C-H) of 81 ( 1 kcal/mol that
we extrapolate for PhCH2OH from the Evans-Polanyi
plot of Espenson et al.11b Because the BDE(C-H) of
toluene is 88.5 kcal/mol,26 our extrapolated values for
ArCH2OH (averaged as 79(2 kcal/mol) point out to a
significant weakening (of about 9 kcal/mol) of the benzylic
C-H bond due to the presence of a geminal OH group.
Finally, we have recently calculated a BDE of 85 kcal/
mol for the NO-H bond of HBT,25 by making use of a
thermochemical cycle and theoretical calculations. This
calculated BDE(NO-H) value is consistent with the
present finding that BTNO abstracts H-atom efficiently
from substrates endowed with C-H bonds in the 80-90
kcal/mol energy range (cf. Tables 1 and 2). In particular,
it is consistent with the experimental finding of a value
for the R of Evans-Polanyi close to 0.5. Lower values of
BDE(NO-H)18b,24 for HBT would be inconsistent with the
observed H-abstraction reactivity of BTNO.
FIGURE 7. Evans-Polanyi plot for the H-abstraction reac-
tion by BTNO with selected substrates (FL stands for fluo-
rene).
endorsing a slightly earlier transition state of H-abstrac-
tion with PINO. In that investigation,11b however, the
Evans-Polanyi correlation was obtained starting from
an Arrhenius treatment where it was assumed that the
preexponential factor A was constant within the series
of substrates investigated.11b As our analysis shows,
instead, this assumption is not warranted, with A (M-1
s-1) values ranging from 8.9 × 104 (with 9-hydroxyfluo-
rene) to 1.9 × 107 (with Ph2CH2). Analogously, the ∆Sq
values range from -38 to -26 eu for the same two
substrates, always being rather prominent in value as
expected for a bimolecular H-abstraction step. Looking
more closely at the entropic parameter for structurally
similar compounds, we verify that replacement of one H
atom with one OH group (e.g., Ph2CH2 vs Ph2CHOH)
geminal to the C-H bond undergoing cleavage causes a
decrease of ∆Sq of 9 eu, i.e., from -27 to -36 eu, as if the
transition state with the alcohol substrate would become
more ordered. Consistently, ∆Sq decreases again 9 eu on
passing from fluorene (-29 eu) to 9-hydroxyfluorene (-38
eu). Unfortunately, we cannot extend this C-H vs C-OH
comparison any further, because of the lack of suitable
substrates to test. For example, the rate constant of
H-abstraction by BTNO from PhCH3 (or from p-MeO-
toluene), taken as a C-H substrate, is too slow for
yielding reliable activation parameters to compare with
that of PhCH2OH (or of p-MeO-benzyl alcohol), as the
corresponding C-OH substrate. More on this point, we
wondered if the “spectator” OH group in either Ph2CHOH
or 9-hydroxyfluorene could be responsible for any peculiar
H-bonding interaction with the approaching BTNO in the
transition state, making up for the above-reported de-
pression of ∆Sq with respect to the corresponding C-H
substrates. A test of this hypothesis was done with the
ether 4-MeO-C6H4CH2OMe, where no H-bonding interac-
tion toward BTNO would be structurally possible. Be-
cause the ∆Sq value comes out identical (-35 eu) within
experimental errors to that of the corresponding alcohol
4-MeO-C6H4CH2OH (-36 eu), we conclude that no H-
bonding interaction between BTNO and the R-OH
Conclusions
The aminoxyl radical BTNO (>N-O•) has been gener-
ated from HBT by oxidation with a CeIV salt and
characterized by spectrophotometry, EPR, cyclic volta-
mmetry, and laser flash photolysis. Rate constants of
H-abstraction (kH) from a number of H-donor substrates
by BTNO have been determined spectroscopically and
range in value from 10-3 to 102 M-1 s-1 at 25 °C in MeCN
solution. The kinetic isotope effect confirms that the
H-abstraction step is rate-determining. For selected
substrates, the activation parameters have also been
measured. Correlation of the energy of activation of
H-abstraction by BTNO vs the energy of the C-H bond
undergoing cleavage in the investigated substrates, by
following the Evans-Polanyi equation, provides the
selectivity (R ) 0.44) of this radical process. The unavail-
able BDE(C-H) of benzyl alcohol can be extrapolated as
ca. 79 kcal/mol from the Evans-Polanyi plot. We compare
the reactivity data of BTNO with similar H-abstraction
results recently obtained with the aminoxyl radical PINO
generated from HPI.11 The higher reactivity of PINO is
understandable on the basis of the higher energy of the
NO-H bond of HPI that is formed on H-abstraction from
the substrate, if compared with the corresponding NO-H
bond energy of HBT (88 vs ca. 85 kcal/mol, respectively).
Acknowledgment. We thank the Italian MIUR for
financial support (COFIN and FIRB projects), Dr. Lo¨ıc
Mottier for writing the software for the electrochemical
experiments, and Prof. Osvaldo Lanzalunga for con-
ducting the laser flash photolysis experiment.
JO051615N
9528 J. Org. Chem., Vol. 70, No. 23, 2005