Persistent hydrogen-bonded and non-hydrogen-bonded phenoxyl radicals

Article abstract of DOI:10.1002/chem.201101509

The production of stable phenoxyl radicals is undoubtedly a synthetic chemical challenge. Yet it is a useful way to gain information on the properties of the biological tyrosyl radicals. Recently, several persistent phenoxyl radicals have been reported, but only limited synthetic variations could be achieved. Herein, we show that the amide-o-substituted phenoxyl radical (i.e. with a salicylamide backbone) can be synthesised in a stable manner, thereby permitting easy synthetic modifications to be made through the amide bond. To study the effect of H-bonding on the properties of the phenolate/phenoxyl radical redox couple, simple H-bonded and non-H-bonded o,p-tBu-protected salicylamidate compounds have been prepared. Their redox properties were examined by cyclic voltammetry and showed a fully reversible one-electron oxidation process to the corresponding phenoxyl radical species. Remarkably, the redox potential appears to be correlated, at least partially, with H-bond strength, as relatively large differences (ca. 300 mV) in the redox potential between H-bonded and non-H-bonded phenolate salts are observed. The corresponding phenoxyl radicals produced electrochemically are persistent at room temperature for at least an hour; their UV/Vis and EPR characterisation is consistent with that of phenoxyl radicals, which makes them excellent models of biological tyrosyl radicals. The analyses of the experimental data coupled with theoretical calculations indicate that both the deviation from planarity of the amide function and intramolecular H-bonding influence the oxidation potential of the phenolate. The latter H-bonding effect appears to be predominantly exerted on the phenolate and not (or only a little) on the phenoxyl radical. Thus, in these systems the H-bonding energy involved in the phenoxyl radical appears to be relatively small.

Full text of DOI:10.1002/chem.201101509

DOI: 10.1002/chem.201101509
Persistent Hydrogen-Bonded and Non-Hydrogen-Bonded Phenoxyl Radicals
Riccardo Wanke,[a] Laurent Benisvy,*[b] Maxim L. Kuznetsov,[a]
M. Fꢀtima C. Guedes da Silva,[a, c] and Armando J. L. Pombeiro*[a]
Abstract: The production of stable
phenoxyl radicals is undoubtedly a syn-
thetic chemical challenge. Yet it is a
useful way to gain information on the
properties of the biological tyrosyl rad-
icals. Recently, several persistent phe-
noxyl radicals have been reported, but
only limited synthetic variations could
be achieved. Herein, we show that the
amide–o-substituted phenoxyl radical
(i.e. with a salicylamide backbone) can
be synthesised in a stable manner,
thereby permitting easy synthetic
modifications to be made through the
amide bond. To study the effect of H-
bonding on the properties of the phe-
nolate/phenoxyl radical redox couple,
simple H-bonded and non-H-bonded
o,p-tBu-protected salicylamidate com-
pounds have been prepared. Their
redox properties were examined by
cyclic voltammetry and showed a fully
reversible one-electron oxidation pro-
cess to the corresponding phenoxyl
radical species. Remarkably, the redox
potential appears to be correlated, at
least partially, with H-bond strength, as
electrochemically are persistent at
room temperature for at least an hour;
their UV/Vis and EPR characterisation
is consistent with that of phenoxyl radi-
cals, which makes them excellent
models of biological tyrosyl radicals.
The analyses of the experimental data
coupled with theoretical calculations
indicate that both the deviation from
planarity of the amide function and in-
tramolecular H-bonding influence the
oxidation potential of the phenolate.
The latter H-bonding effect appears to
be predominantly exerted on the phe-
nolate and not (or only a little) on the
phenoxyl radical. Thus, in these sys-
tems the H-bonding energy involved in
the phenoxyl radical appears to be rel-
atively small.
relatively
large
differences
(ca.
300 mV) in the redox potential be-
tween H-bonded and non-H-bonded
phenolate salts are observed. The cor-
responding phenoxyl radicals produced
Keywords: electron transfer · hy-
drogen bonds · phenol · phenolate ·
phenoxyl radicals · redox chemistry ·
tyrosyl radical model
C
Introduction
iron core, whereas in PSII a Tyr H-bonded to an adjacent
histidine residue is formed. H-bonding is suspected to have
a strong influence on the redox properties of the tyrosine;
C
Tyrosyl radicals (Tyr ) play a crucial role in many biological
systems.[1] For instance, they are essential to the catalytic ac-
tivity of: 1) the class I RNR (iron-dependent ribonucleotide
reductase) enzymes that are relevant in DNA replication
and DNA repair, and 2) the photosystem II (PSII) that uses
sunlight to split water into O2 through photosynthesis. In
indeed, whilst the redox potential of Tyr/Tyr (versus the
C
normal hydrogen electrode) has a value of 1.00 V for the
non-hydrogen-bonded Tyr122 of RNR,[2] it decreases signifi-
cantly to 0.72–0.76 V for the H-bonded TyrD in PSII.[3]
These observations pose fundamental questions on the influ-
ence of H-bonding on the redox properties (and mechanism
C
class I RNR, a free Tyr is produced in the proximity of a di-
C
of formation) of the biological Tyr .
In less than a decade, remarkable synthetic progress has
been made in generating and characterising persistent H-
bonded phenoxyl radicals as chemical models for biological
Tyr radicals. For instance, o,p-tBu-protected phenol com-
pounds that are intramolecularly H-bonded to a nitrogen
base (from amino, imidazole or pyridine groups) have
proved to undergo a (quasi)reversible proton-coupled elec-
tron transfer (PCET) oxidation process to H-bonded phe-
[a] Dr. R. Wanke, Dr. M. L. Kuznetsov, Dr. M. F. C. Guedes da Silva,
Prof. A. J. L. Pombeiro
Centro de Quꢀmica Estrutural, Complexo I
Instituto Superior Tꢁcnico, TU, Lisbon
Avenida Rovisco Pais, 1049-001 Lisbon (Portugal)
Fax : (+351)21-846-4455
E-mail: pombeiro@ist.utl.pt
[b] Dr. L. Benisvy
noxyl radicals of the type R O ···H +N.[4–14] In these sys-
tems, the phenol oxidation process occurs at a much lower
potential than that of a non-H-bonded phenol, which could,
at first sight, be a consequence of H-bonding effects in the
reduced and oxidised forms. However, Mayer and co-work-
ers have elegantly demonstrated that this decrease in redox
potential results predominantly from the driving force for
proton movement in a concerted PCET (i.e. CPET) mecha-
ꢀ
ꢀ
C
Department of Chemistry, Laboratory of Biomimetic Chemistry
Bar-Ilan University, Ramat Gan 52900 (Israel)
Fax : (+972)3-7384053
E-mail: benisvl@mail.biu.ac.il
[c] Dr. M. F. C. Guedes da Silva
Universidade Lusꢂfona de Humanidades e Tecnologias, Lisbon
Av. Campo Grande 376, 1749-024 Lisbon (Portugal)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101509.
11882
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11882 – 11892

FULL PAPER
nism.[11,15] Thus, although these systems are excellent models
for mechanistic studies on PCET, the associated proton
movement during oxidation prevents direct insights into the
effects of H-bonding on the phenoxyl redox potential. With
the aim of studying the “real” effect of H-bonding on the
phenolate/phenoxyl radical redox couple, we have designed
new, simple, o,p-tBu-protected salicylamidate compounds
Solid, solution and calculated structures NRR’LH and their
corresponding phenolate salts [NRR’L]
ACHTNUTRGENNG[U NBu4]: It is well
known that 2-phenol amide compounds commonly exhibit a
ꢀ
planar conformation owing to strong intramolecular O
ꢀ
H···O H-bonding between the phenol O H and the adjacent
carbonyl O atom. Kanamori et al.[17] have elegantly shown
that deprotonation of the phenol group induces a conforma-
(
RR’Lꢀ; Scheme 1), which comprise (or not) a (phenola-
tion twist of the amide function, thereby yielding a planar
ꢀ
ꢀ
structure characterised by strong intramolecular O ···H N
H-bonding between the phenolate Oꢀ atom and the amide
ꢀ
N H, as illustrated in Scheme 2. Clearly for secondary
ꢀ
amines (that do not possess the N H amide function),
whilst intramolecular H-bonding can occur in the phenol
form, it cannot in the phenolate form. This constitutes the
basis of our strategy in building H-bonding versus non-H-
bonding phenolate compounds (Scheme 1).
Scheme 1.
ꢀ
ꢀ
te)O ···H N
N
show that both H-bonded and non-H-bonded phenolate
compounds undergo a straightforward one-electron reversi-
ble oxidation process (ET process) to the corresponding per-
sistent phenoxyl radical species, which indicates that the sali-
cylamide backbone can be used to produce stable phenoxyl
radicals. Thus, in these systems, the oxidation process does
not involve any proton transfer and yet allows H-bonding
(or not) at both the phenolate and phenoxyl radical. The
redox potential of this process is shown to be affected by
both the deviation of the amide backbone from planarity
and the presence of H-bonding, the effect of which is pre-
dominantly exerted on the phenolate.
Scheme 2. Conformational twist upon deprotonation of 2-phenol amide
compounds.
Planar H-bonded phenols NHOHLH and NHMeLH and their
corresponding H-bonded phenolate salts [NHOHL]
ACHTUNGTRENNUNG
[
NHMeL]
ACHTUNGTRENNUNG
compounds described herein have been studied in solution
by NMR and IR spectroscopy, in the gas phase by DFT cal-
culations, and in the solid state by X-ray crystallography for
NHOHLH*, [NHOHL][NBu4] and [NHMeL]
ACTHNUTRGENNUG HCAUTGNTREN[NUGN NBu4] (see Tables 1
and 2, Figures 1 and 2). The X-ray structure of phenol
NHOHLH* (analogue to NHOHLH but lacking tBu groups,
Results and Discussion
Table 1. Crystallographic data for NMe2LH, NHOHLH*, [NHOHL]
ACHTUNGTRENNUNG
NMe2LH
NHOHLH*
G
E
ACTHUNGTRNENUG[
NHMeL]ꢀ
Synthesis of NRR’LH and
[
NRR’L]
ACHTUNGTRENNUNG[NBu4]: The phenols
formula
M
crystal system
space group
a [ꢄ]
b [ꢄ]
c [ꢄ]
a [8]
b [8]
C17H27N1O2
277.40
C9 H11 N1O3
181.19
C16H24NO2
262.36
NRR’LH (Scheme 1) are readily
synthesised in high yield[16] by
the condensation of the corre-
sponding amine, HNRR’, onto
triclinic
monoclinic
P21/c (No. 14)
8.5549(8)
12.1590(12)
9.1001(10)
90.00
115.466(2)
90.00
854.61(15)
4
monoclinic
P21/c (No. 14)
11.268(4)
16.369(6)
19.073(7)
90.00
115.31(2)
90.00
3180 (2)
4
¯
¯
P1 (No. 2)
6.2379(6)
16.1287(16)
18.7725(19)
110.978(4)
99.397(6)
97.111(5)
1705.4(3)
4
succinimide-activated
3,5-di-
tert-butyl salicylic acid. The de-
protonation of the phenol
moiety by using an equimolar
g [8]
V [ꢄ3]
Z
amount of OH
corresponding phenolate salt,
NRR’L]
[NBu4].
ACHTUNGTRENNUNG[NBu4] yields the
1calcd [gcmꢀ3
]
1.080
0.070
22308
1.408
0.106
4306
0.548
0.036
19075
m [cmꢀ1
]
[
ACHTUNGTRENNUNG
reflections collected
unique reflections
6211
1872
6675
5799
Rint
0.0656
0.0469
0.0807
0.0640
observed reflections
3573
1429
3535
3004
R
0.0578
0.0404
0.0636
0.1043
Rw
0.117
0.0561
0.1352
0.1452
Chem. Eur. J. 2011, 17, 11882 – 11892
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11883

L. Benisvy, A. J. L. Pombeiro et al.
NRR’
Table 2. Experimental and calculated geometrical parameters for phenol NRR’LH, phenolate NRR’Lꢀ and phenoxyl radical
HMe, HOH).
C
L compounds (RR’=Me2,
NMe2
NHMe
NHOH
NMe2LH
calcd
30
NMe2Lꢀ
calcd
60
L
NHMeLH
calcd
NHMeLꢀ
L
NHOHLH*, NHOHLH
expt calcd expt
NHOHLꢀ
C
C
C
L
Compound
Method
Phenol–amide
twist angle [8]
Intramolecular
H-bonding
expt
ca. 71
calcd
80
expt
calcd calcd
calcd calcd
non-planar
OH···O none
<10 planar
Oꢀ···HN
<10 planar
Oꢀ···HN
C
C
O ···HN
none
none
OH···O
O ···HN
OH···O
ꢀ
C1 O1
ꢀ
C1 C2
1.374(3)
1.395(3)
1.392(3)
1.379(3)
1.400(4)
1.387(4)
1.401(3)
1.501(3)
1.243(3)
1.334(3)
1.468(3)
1.355
1.416
1.408
1.390
1.411
1.396
1.421
1.491
1.251
1.370
1.462
1.276
1.450
1.404
1.393
1.417
1.389
1.466
1.492
1.246
1.382
1.450
1.258
1.461
1.382
1.409
1.424
1.381
1.474
1.510
1.235
1.363
1.455
1.458
1.350
1.418
1.410
1.387
1.412
1.393
1.424
1.488
1.252
1.359
1.457
1.290(3) 1.291 1.260
1.437(4) 1.447 1.467
1.373(4) 1.404 1.381
1.381(4) 1.391 1.410
1.402(4) 1.407 1.421
1.374(4) 1.390 1.380
1.433(4) 1.459 1.478
1.511(4) 1.494 1.517
1.243(4) 1.250 1.237
1.317(4) 1.361 1.352
1.456(4) 1.443 1.452
1.3606(17) 1.349 1.303(3) 1.291 1.260
1.4092(19) 1.419 1.426(4) 1.448 1.467
1.4043(19) 1.411 1.399(4) 1.405 1.381
ꢀ
C2 C3
ꢀ
C3 C4
1.376(2)
1.390(2)
1.382(2)
1.391(2)
1.387 1.375(4) 1.390 1.411
1.412 1.406(4) 1.416 1.421
1.393 1.371(4) 1.389 1.380
1.424 1.444(4) 1.459 1.478
ꢀ
C4 C5
bond
length
[ꢄ]
ꢀ
C5 C6
ꢀ
C6 C1
ꢀ
C2 C7
1.4850(18) 1.389 1.482(4) 1.492 1.518
1.2628(16) 1.252 1.251(3) 1.250 1.236
1.3276(18) 1.361 1.331(3) 1.366 1.356
1.4590(18) 1.459 1.443(4) 1.445 1.455
ꢀ
C7 O2
ꢀ
C7 N1
ꢀ
C8 N1
ꢀ
(C9 N1)
(1.452(3))
ACHTUNGTREN(NUNG 1.462) 1.448
Figure 2. DFT-calculated structures of NHMeLH (a), NHOHLH (b), NHMeLꢀ
(c) and NHOHLꢀ (d).
Figure 1), as expected, reveals a planar conformation (dihe-
ꢀ
dral angle 14.028) owing to the O H···O H-bonding between
ꢀ
the phenol O H and the adjacent carbonyl O atom. The pa-
ꢀ
rameters (O···O distance, O H···O angle) are as expected
and are comparable to those of other 2-amide phenol com-
pounds.[18] The H NMR spectra of NHOHLH and NHMeLH, in
1
ꢀ
CD3CN, exhibit a phenolic O H proton resonance at d=
ꢀ
13.31 and 13.39 ppm, respectively, characteristic of an O
H···O=C intramolecular H bond of the phenol,[17,18] whereas
ꢀ
the amide N H proton resonances at d=7.49 and 7.37 ppm,
respectively, are typical for a non-hydrogen-bonded amide
N H. Moreover, the IR spectra of NHOHLH and NHMeLH, in
ꢀ
ꢀ
CH3CN solution, exhibit an N H stretching frequency at
Figure 1. ORTEP representations (shown with 50% thermal ellipsoids)
of NHOHLH* (a), NHOHLꢀ in NHOHL][NBu4]·2H2O (b) and NHMeLꢀ in
NHMeL]
[NBu4] (c). Only the H atoms of NH and OH groups are shown.
3411 and 3413 cmꢀ1, respectively, typical of a non-H-bonded
[
ACHTUNGTRENNUNG
ꢀ
[
N
amide N H. Thus, these data clearly demonstrate that the
11884
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11882 – 11892

Persistent H-Bonded and Non-H-Bonded Phenoxyl Radicals
FULL PAPER
phenol NHOHLH and NHMeLH possess a planar conformation
with OH···O H-bonding in the solid state as well as in solu-
tion. The molecular structures of the phenolate salts
Non-planar phenol NMe2LH and its corresponding phenolate
ACTHNUTRGNEUNG
salt [NMe2L][NBu4]: The structures of the phenol NMe2LH and
its corresponding phenolate [NMe2L]
ACTHNUTRGEN[NUG NBu4] salt have been
studied in solution by NMR and IR spectroscopy, in the gas
phase by DFT calculation, and in the solid state by X-ray
crystallography for NMe2LH (Figures 3 and 4, Tables 1 and 2).
[
NHOHL]
N
N
X-ray crystallography (Figure 1), are quasi-planar with twist
angles between the phenolate and the CONH moiety of
6.868 in
[ ACHTUNGTRENNUNG
NHOHL]
and 5.528 in [NHMeL]
ACHTUNGTRENNUNG
planarity of both phenolate
compounds results from rela-
tively strong intramolecular
ꢀ
[20]
ꢀ
O ···H N H-bonding
involv-
ing the phenolate O atom and
ꢀ
the adjacent N H amide group,
as evidenced by the O···N dis-
ꢀ
tance and the O···H N angle in
[
NHOHL]
N
and 145.598) and in [NHMeL]-
[NBu4] (2.573 ꢄ and 139.368),
ACHTUNGTRENNUNG
which are comparable to those
Figure 3. ORTEP representations (with 50% thermal ellipsoids) of NMe2LH showing the dimers formed
through intermolecular OH···O bonding. Only the H atoms of OH groups are shown.
of other Oꢀ
G
ꢀ
ꢀ
(amide)/H O(carboxylic) intra-
molecularly H-bonded salicyla-
midate[17,18,21]/salicylate[22] com-
pounds. The bond lengths and angles within both structures
are as expected for phenolate compounds,[18] with a C O
ꢀ
bond length of 1.303(3) ꢄ (in NHOHLꢀ) and 1.290(3) ꢄ (in
1
NHMeLꢀ; see Table 2). The H NMR spectra of the phenolate
salts [NHOHL][NBu4] and [NHMeL]
ACHTUNGTRENNUNG ACHTUNTGERN[NUGN NBu4] in CD3CN each dis-
ꢀ
play a N H resonance at a much lower field (i.e. at d=
13.90 and 13.20 ppm, respectively) than that of the corre-
sponding parent phenol (d=7.49 and 7.37 ppm, respective-
ꢀ
ly), clearly indicating that the N H proton is involved in an
ꢀ
ꢀ
intramolecular O ···H N hydrogen bond, as shown in
Scheme 2.[18] Moreover, the IR spectra of NHOHLꢀ and
NHMe
ꢀ
ꢀ
L , in CH3CN solution, exhibit an N H stretching at a
much lower frequency (3005 and 3009 cmꢀ1, respectively)
than that of the corresponding phenol (3411 and 3413 cmꢀ1),
which is typically indicative of a relatively strongly H-
[19]
ꢀ
bonded N H bond. Thus, these data clearly demonstrate
Figure 4. DFT-calculated structures of NMe2LH (a) and NMe2Lꢀ (b).
ꢀ
ꢀ
that the O ···H N H-bonding identified in the X-ray struc-
ture of NHOHLꢀ and NHMeLꢀ is preserved in solution.
DFT calculations have been performed on the phenol and
corresponding phenolate compounds. The calculated opti-
mised structures are in good agreement with the corre-
sponding structures determined experimentally. As shown in
Table 2, the maximum difference between the calculated
and experimental bond lengths is 0.03 ꢄ and often falls
within a 3s interval of the experimental data. Additionally,
the optimised structures of phenols NHOHLH and NHMeLH are
planar and present intra-OH···O H-bonding; in addition, the
optimised structures of phenolates NHOHLꢀ and NHMeLꢀ are
planar with intra-Oꢀ···HN H-bonding (see Figure 2,
Table 2). Thus, these calculations reproduce well the solid-
state and solution structures determined experimentally.
The crystal structure of NMe2LH is in great contrast to that of
NHOHLH*. Two molecules were found in the asymmetric
unit, and the phenol ring and the amide backbone in each
ꢀ
are no longer coplanar; instead, the O=C N(Me)2 planes
are quasi-perpendicular to the phenol planes, with dihedral
angles between the two planes of 71.18 and 71.778
ꢀ
(Figure 3). Thus, the intramolecular O H···O=C H-bonding
ꢀ
is prevented. Instead, both the phenol O H and the C=O
group act as H-bond donor and acceptor to another mole-
cule, thus yielding dimer assemblies (Figure 3). The twist of
the amide backbone is most likely to have occurred as a
result of a steric clash between one of the two methyl
groups and the nearby meta proton, which prevents the mol-
ecule from being planar. Thus, in NMe2LH, not only is the in-
Chem. Eur. J. 2011, 17, 11882 – 11892
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11885

L. Benisvy, A. J. L. Pombeiro et al.
ꢀ
tramolecular O H···O=C bonding lost, but also the delocali-
sation of the p system is disrupted. However, this may be
the result of crystal packing and the preferred arrangement
of the molecules in crystals.
1
In solution, the H NMR spectrum of NMe2LH, in CD3CN,
ꢀ
exhibits a phenolic O H proton resonance at d=10.33 ppm,
which is shifted upfield relative to the H-bonded NHOHLH
and NHMeLH (13.31 and 13.39 ppm, respectively). Neverthe-
less, the resonance at 10.33 ppm is far too high to be consid-
ered as the result of a proton resonance from a non-hydro-
gen-bonded phenol OH group. Moreover, the strong possi-
bility of intermolecular H-bonding has been discarded, be-
cause this resonance has been found to be independent of
sample dilution (with dilution of 1:100). Therefore, despite
the steric hindrance imposed by the methyl groups, it ap-
pears that in solution intramolecular hydrogen bonding still
occurs, to a weaker extent than for NHOHLH and NHMeLH.
The DFT-calculated optimised structure appears to match
the NMR data. Indeed, the optimised structure shows a
phenol–amide twist angle of only about 308 and still permits
Figure 5. Cyclic voltammograms of [NRR’L]
containing 0.2m [NBu4] .
[BF4] at 298 K recorded at a scan rate of 0.1 Vsꢀ1
ACHTUNGTREN[NUNG NBu4] (ca. 1 mm) in CH3CN
G
relatively strong O H···O=C intramolecular H-bonding (see
(E1/2 =ꢀ0.423 V vs. Fc/Fc+) occurs at a much lower poten-
tial than those for NHOHLꢀ and NHMeLꢀ, but at a higher poten-
tial than that of the non-hydrogen-bonded 2,4,6-tri-tert-bu-
tylphenoxide ArOꢀ (E1/2 =ꢀ0.572 V vs. Fc/Fc+ in
CH3CN).[25] This can be explained by combining 1) the elec-
tron-donating effect of the additional o-tBu group in ArOꢀ,
which makes it easier to oxidise, and 2) the electron-with-
drawing effect of the amide functionality in NMe2Lꢀ, which
decreases the electron density at the O atom in NMe2Lꢀ thus
making it harder to oxidise. Both these effects contribute to
increasing the oxidation potential of the phenolate in NMe2Lꢀ
as compared to ArOꢀ. Most importantly, relatively large dif-
ferences (262 and 304 mV, respectively) in the redox poten-
tials are observed between planar H-bonded (NHMeLꢀ and
NHOHLꢀ) and twisted non-H-bonded (NMe2Lꢀ) salts. This
redox shift may result from both combining effects acting in
concert: 1) an H-bonding effect that reduces the electron
density at the phenolate O atom, thus making the oxidation
potential increase for NHMeLꢀ and NHOHLꢀ relative to NMe2Lꢀ;
and 2) the disruption of the extension of the p system to the
amide backbone in NMe2Lꢀ, which increases the electron den-
sity at the phenolate O atom, thereby making it easier to ox-
idise relative to NHMeLꢀ and NHOHLꢀ. One should note that
this contrasts, at least in part, with the early assumptions
that H-bonding does not induce a significant redox shift.[11]
ꢀ
Figure 4, Table 2), as well as minimising steric clashes be-
tween a methyl group and a meta proton. Thus, whilst the
introduction of a supplementary methyl group at the amide
function induces a twist and to some degree disrupts the ex-
tended p system, intramolecular H-bonding is evidenced in
solution and confirmed by DFT calculation. The crystal
structure presumably represents an extreme case in which
optimisation of crystal packing and intermolecular H-bond-
ing may have introduced further twisting.
The structure of the corresponding phenolate anion
NMe2Lꢀ was estimated by DFT calculation. The optimised
calculated structure reveals a large deviation from planarity,
with a twist angle between the phenolate and the amide
planes of approximately 608 (see Figure 4). This confirms
that, as compared to [NHMeL][NBu4] and [NHOHL]
ACTHNUTRGNENUG ACHTUNGTREN[NUGN NBu4], the
loss of H-bonding by the introduction of a methyl group is
also accompanied by a partial disruption of the delocalisa-
tion of the p system by the loss of planarity, and both effects
should be taken into account (see below).
Electrochemical studies of [NRR’L]
mogram of each phenolate salt [NRR’L]
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
298 K, exhibits a fully reversible one-electron oxidation pro-
cess attributed to the formation of the corresponding phe-
noxyl radical species (Scheme 1, Figure 5). This oxidation
process occurs at E1/2 =ꢀ0.119, ꢀ0.161 and ꢀ0.423 V versus
ferrocene/ferrocenium (Fc/Fc+) for NHOHLꢀ, NHMeLꢀ and
NMe2Lꢀ, respectively. The full reversibility of the oxidation
process of NHOHLꢀ and NHMeLꢀ most likely indicates the re-
Generation and characterisation of the neutral phenoxyl
NRR’
C
radicals
L : The electrochemical one-electron oxidation
of each NRR’Lꢀ anion, in CH3CN at room temperature, yields
the corresponding stable (for several hours under inert at-
ꢀ C
ꢀ
tention of the N H···
U
mosphere) bright-green-coloured phenoxyl radical species
NRR’
C
sumably very little conformational change between reduced
and oxidised species. The oxidation potentials of NHOHLꢀ
and NHMeLꢀ are comparable to that of the previously report-
ed 2-amido-4,6-di-tert-butylphenol recorded in water/
CH3CN under basic conditions (pH>12).[24] In contrast, the
oxidation of the non-hydrogen-bonded phenolate NMe2Lꢀ
L , as indicated by their UV/Vis and EPR spectra (see
below). The cyclic voltammogram of each of the radicals
produced is identical to that of its parent salt, thus confirm-
ing the chemical reversibility of the oxidation process.
The X-band EPR spectrum of each of the electrogenerat-
NRR’
C
L species, both in fluid and frozen CH3CN solutions,
ed
11886
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11882 – 11892

Persistent H-Bonded and Non-H-Bonded Phenoxyl Radicals
FULL PAPER
NMe2
C
L , the spectrum
exhibits an isotropic radical signal centred at giso =2.0046
the phenoxyl ring. In contrast to that of
NMe2
NHOH
NHMe
NHMe
C
C
C
C
(
L ), 2.0043 (
L ) or 2.0042 (
L ) with a peak-to-
of
L is unresolved and lacks hyperfine features. Howev-
NHOH
C
L exhibits a barely observable dou-
peak line width of 3–4 G (Figure 6). The relatively small line
width of each signal most likely indicates that there is no
significant coupling of the radical with the N atom from the
amide function, thus implying that the unpaired electron is
er, the spectrum of
blet. The latter pattern is reminiscent of that obtained in a
peptide-linked phenoxyl radical, invoking the coupling with
only one Hmeta atom (aH =1.7 G).[24] These observations most
likely indicate that H-bonding induces a redistribution of
the electron spin density, as demonstrated earlier by Lucari-
ni et al.[26]
not (or to a very little extent) delocalised on the amide
NMe2
C
L is well resolved (at
group. The X-band spectrum of
271 K) and exhibits a three-line pattern, which has been suc-
cessfully simulated with the inclusion of a hyperfine cou-
pling of the electron spin with both meta protons of the phe-
noxyl ring (aHmeta1 =1.60 G; aHmeta2 =1.45 G; Figure 6). As
expected for o,p-protected phenoxyl radicals, the hyperfine
coupling with the meta protons is relatively small (i.e.
<2 G),[23] in agreement with the odd-alternant pattern of
phenoxyl radicals in which the electron density is mainly
spread on the O atom, and both ortho and para positions of
Interestingly, the relatively large Dgiso shift of 0.0003–
0.0004 between the signal of the non-hydrogen-bonded
NMe2
NHOH
NHMe
C
C
C
L
L (g=2.0046) and that of
L (g=2.0043) or
(g=2.0042) is consistent with the latter two radicals being
H-bonded phenoxyl radicals. Indeed, theoretical[27–30] and ex-
perimental studies[6,7,31–33] have indicated that hydrogen
bonding to a phenoxyl (or Tyr) radical results in a consider-
able lowering of the gx value but leaves gy and gz essentially
unaffected. Thus, assuming that gy and gz are the same for
RR’
C
L , a Dgx shift of 0.0012–0.0016 (3Dgiso) is obtained.
all
Such a Dgx shift is in good agreement with the gx shift ob-
served between an H-bonded phenoxyl (Tyr) radical
(2.0061<gx <2.0068 (phenoxyl)[6,7,33,34] and 2.0075<gx <
2.0076[35,36] (biological Tyr radicals)) and non-hydrogen-
bonding phenoxyl radicals such as the 2,4,6-tri-tert-phenoxyl
radical (gx =2.00735)[33] or “free” Tyr (gx =2.0087–
C
2.0089).[31,37]
The UV/Vis spectrum of each electrogenerated oxidised
NRR’
C
L exhibits two intense bands at about 400 nm
species
(e>1000mꢀ1 cmꢀ1) and a weak near-IR (NIR) broad band at
600–700 nm (e ca. 250mꢀ1 cmꢀ1), which are characteristic of
free
or
H-bonded
phenoxyl
radical
transitions
(Figure 7).[4,6,9,12,38] The ratio e400/e700 between the intensities
of the approximately 400 nm (p–p* transitions) and the
circa 700 nm bands appeared to be indicative of the delocal-
isation of the phenoxyl unpaired electron; that is, for a
“pure” phenoxyl radical, e400/e700 is about 5, whereas for a
highly delocalised radical e400/e700 can even be <1.[9,38,39]
Thus, the e400/e700 ratio of about 4 observed for each of the
NRR’
C
L
Figure 6. Top: X-band EPR spectra of electrogenerated radicals
(ca. 1 mm) in CH3CN containing [NBu4]
263 K. Bottom: X-band EPR spectrum (c) of the electrogenerated
AHCTUNRTGEG[NUNN BF4] (0.2m) recorded at 253–
NMe2
C
L (ca. 1 mm) in CH3CN containing [NBu4]ACHTUNTGRNUE[NG BF4] (0.2m) re-
radical
corded at 271 K; centre field: 3353.38; modulation frequency: 100 kHz;
modulation amplitude: 0.3 Gpp; receiver gain: 4.48ꢅ10ꢀ4; conversion
time: 40.96 ms; microwave power: 0.4 mW; time constant: 81.92 ms;
sweep time: 167.77 s. Simulated spectrum (g) obtained using Bruker
SimFonia software, with the following fitting parameters: gx =gy =gz =
2.0046; aH1meta =1.60 G; aH2meta =1.45 G.
Figure 7. Room-temperature UV/Vis spectra of [NRR’L]
ACTHNUGTRENNG[U NBu4] (ca. 1 mm,
NRR’
C
c) and their corresponding electrogenerated radicals
L (g), in
CH3CN containing [NBu4]ACTHUNGRTNE[NUG BF4] (0.2m).
Chem. Eur. J. 2011, 17, 11882 – 11892
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11887

L. Benisvy, A. J. L. Pombeiro et al.
radicals studied strongly indicates that there is no or very
count as well as the H-bonding energies in the reduced and
oxidised forms (i.e. DGHBred and DGHBox, respectively). This
leads to the thermochemical cycle depicted in Scheme 3. Be-
NRR
C
L onto
little delocalisation of the unpaired electron in
the amide backbone, and that the electronic structure of
NRR’
C
L resembles that of a true phenoxyl radical.
each
DFT calculations were performed on the radical species
to gain further information on their electronic structure. The
NHOH
NHMe
C
C
indicates a
geometry optimisation of
L
and
L
ꢀ
C
planar conformation allowing O ···H N H-bonding between
ꢀ
the phenoxyl radical and the N H amide, which confirms
the reversible behaviour of the oxidation process from
NHR
NMe2
NHRLꢀ to
L . In contrast, the calculated structure of
C
C
L
shows a twist of the amide function with respect to the phe-
noxyl plan. The twist angle is approximately 808, essentially
similar to its corresponding phenolate (twist angle 608; see
Table 2). Thus, it is likely that there is no (or only a little)
NMe2
conformation change between NMe2Lꢀ and
C
L . In all three
radical species, the corresponding calculated bond lengths
are considered to be essentially identical (see Table 2) and
the calculated SOMO is characterised by a typical odd-alter-
nant pattern of phenoxyl radicals. As a result, the electron
spin density is quasi-equally distributed over both ortho and
para C atoms and the phenoxyl O atom (see Figure 8),[27]
and, as expected, very little electron spin density (<2%) is
located on the amide N atom.
NHOH
NHMe
Scheme 3.
C
C
L
Figure 8. Calculated spin density distribution in (left)
L and
NMe2
C
(numbers in parentheses), and in (right)
L .
NMe2
cause the twist angles of NMe2Lꢀ and
C
L are sensibly the
H-bonding influence on the phenolate/phenoxyl radical
redox couple: From the results described above, it is clear
that both H-bonded phenolate compounds NHMeLꢀ and
NHOHLꢀ, and twisted non-H-bonded phenolate NMe2Lꢀ under-
go a one-electron oxidation process to their corresponding
persistent phenoxyl radical without any significant confor-
same, one can crudely assume that DGtwistred =DGtwistox.
Then, the difference in ionisation energy DE (EHBꢀENHB),
C
which can be approximate to E
N
NHRLꢀ/ )ꢀE
NMe2Lꢀ/ ) experi-
C
ACHUTNGRNENUG(
mentally determined, is only related to the difference DDG
(i.e. DGHBredꢀDGHBox) between the H-bonding strength of
the reduced form (DGHBred) and that of the oxidised form
(DGHBox). Most hydrogen bonds are 3–8 kcalmolꢀ1, so the
observed difference DDG of about 6–7 kcalmolꢀ1 would in-
dicate that the H-bonding interaction in the phenolate form
is much stronger that of the corresponding phenoxyl radical
form, and thus imply that the H bond in the phenoxyl radi-
cal would be relatively weak.
mational changes. Yet, large differences (DE=EHBꢀENHB
)
in redox potential (0.262 and 0.304 V) are observed between
H-bonded (HB) species (E
NMe2Lꢀ/ )).
C
tively) and the non-H-bonded (NHB) one (EACHTNUGTRENNN(UG
NHMeLꢀ/ ) or E NHOHLꢀ/ ), respec-
C C
ACHTNUTGRENUNG( ACHTGNUTRENNUG(
These differences correspond to 25.3 and 29.3 kJmolꢀ1 (ca. 6
and 7 kcalmolꢀ1), respectively, and correlate well with the
computed differences in the ionisation potentials (IPs) for
CH3CN solutions between H-bonded phenolate (NHMeLꢀ and
NHOHLꢀ) and non-H-bonded phenolate NMe2Lꢀ of 26.0 and
The hydrogen-bonding strength is related to the differ-
ence in pKa values between the donor and acceptor; the
higher this difference, the weaker the H-bond energy. Upon
oxidation from phenolate to the phenoxyl radical, the pKa
decreases from about 10 to about 0, whilst the pKa of the
27.0 kJmolꢀ1, respectively. Because the non-H-bonded
NMe2
NMe2Lꢀ and
C
L are significantly twisted from planarity, ro-
ꢀ
tational energy in the reduced and oxidised forms (i.e.
DGtwistred and DGtwistox, respectively) should be taken into ac-
N H amide bond is assumed to remain the same (ca. 17).
Thus, upon oxidation, the H-bond strength is expected to
11888
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11882 – 11892

Persistent H-Bonded and Non-H-Bonded Phenoxyl Radicals
FULL PAPER
decrease significantly, which is consistent with our experi-
mental results.
tential is predominantly exerted on the phenolate and not
(or only a little) on the phenoxyl radical, in contrast to earli-
er common belief. As a whole, these compounds constitute
valuable chemical models of biological Tyr radicals, and this
study also represents a major synthetic development in the
chemistry of phenoxyl radicals, as it will enable extensive
and versatile synthetic variations or attachments through the
amide bond. Further studies towards these goals are in prog-
ress in our laboratories.
Biological relevance in enzymatic mechanism: Overall, our
findings indicate that there is a large difference in redox po-
tential between H-bonded and non-H-bonded phenolate
compounds. This is valuable information in the context of
understanding the local environmental factors that modulate
C
the redox potential of the Tyr/Tyr redox couple in biological
systems, and can help to comprehend aspects of enzymatic
mechanistic processes. As shown in Scheme 4, we proposed
Experimental Section
General: All syntheses were carried
out under an atmosphere of dinitro-
gen, by using standard Schlenk tech-
niques. All solvents were dried, de-
gassed and distilled prior to use. The
reagents N-hydroxysuccinimide and
dicyclohexylcarbodiimide
(Aldrich)
Scheme 4.
were purchased and used without fur-
ther purification.
C, H and N analyses were carried out
by the Microanalytical Service of the
a putative mechanism to illustrate this fact. For instance,
one could envisage that within an enzyme-active site, the
amino acids tyrosine and lysine are located in close proximi-
ty. The lysine amine group is more basic than the tyrosine
phenol residue, so it is reasonable to propose that the tyro-
sine is deprotonated (phenolate form) and the lysine is pro-
Instituto Superior Tꢁcnico. Infrared spectra (4000–400 cmꢀ1) were record-
ed on a BIO-RAD FTS 3000MX instrument in KBr pellets. Frequencies
are expressed in cmꢀ1. UV/Vis–NIR spectra (1600–200 nm) were record-
ed on a Shimadzu UV-3101PC UV/Vis NIR spectrophotometer. 1H,
13C NMR spectra were measured on Bruker 300 and 400 UltraShield
spectrometers. H and 13C chemical shifts d are expressed in ppm relative
to Si(Me)4. Coupling constants are in Hz; abbreviations: s, singlet; d,
doublet; m, complex multiplet.
1
tonated (ammonium form). The two amino acids could es-
ꢀ
ꢀ
tablish a H-bonding interaction TyrO ···H NLys of the
Electrochemistry: The electrochemical experiments were carried out on
an EG&G PAR 273A potentiostat/galvanostat connected to a personal
computer through a general-purpose interface bus (GPIB). Cyclic vol-
tammetry studies were undertaken in a two-compartment three-electrode
cell, with platinum disk working (d=0.5 mm) and counter electrodes. A
Luggin capillary connected to a silver-wire pseudo-reference electrode
was used to control the working electrode potential. The solutions were
saturated with N2 by bubbling this gas before each run and were 10ꢀ3 m in
same type as our model compounds NHMeLꢀ and NHOHLꢀ.
One could envisage then that conformational changes of the
protein under external factors, or more likely substrate bind-
ing at the active site, would induce steric constraints and
eventually H-bond breaking between the Tyr–Lys conjugate.
This in turn would decrease the oxidation potential of the
C
tyrosine, thereby making the oxidation to Tyr easily achiev-
the test compound and 0.2m in [NBu4]ACHTNUGRTENUNG[BF4] as supporting electrolyte.
able through electron transfer to a natural redox cofactor
Controlled potential electrolysis (CPE) was carried out in a two-compart-
ment three-electrode cell with platinum-gauze working and counter elec-
trodes in compartments separated by a glass frit; a Luggin capillary,
probing the working electrode, was connected to a silver-wire pseudo-ref-
erence electrode. The solutions were saturated with N2 by bubbling this
(e.g. FeS clusters, NAD+). Thus, H-bond breaking would
C
trigger the formation of the Tyr , thereby triggering the oxi-
dation catalysis commonly achieved by H atom abstraction
from the substrate by the Tyr radical.
gas before each run and were 0.2m in [NnBu4]
ACHTUNGTREN[NUGN BF4] as supporting elec-
trolyte and approximately 10ꢀ3 m in the test compound.
Reversibility tests and CPE (in CH3CN at RT): [NHOHL]
ACHTNUGETNR[UNGN NBu4]: E1/2 =
ꢀ0.119 V versus Fc/Fc+, Epa =ꢀ89 mV, DE=61 mV at 120 mVsꢀ1. DE=
Conclusion
c
58–71 mV (scan rate=20–900 mVsꢀ1); ip /ipa =1ꢁ0.1 at scan rate
c
a
ꢂ100 mVsꢀ1; at 20, 50 and 80 mVsꢀ1, ip /ip was 1.18, 1.34 and 1.29, re-
We have shown that simple o,p-tBu-protected salicylamidate
compounds can be reversibly oxidised to produce persistent
phenoxyl radicals, thus demonstrating that the presence of
an amide function at one ortho position of the di-tBu-pro-
tected phenol ring does not significantly disturb either the
stability or the electronic structure of the generated phenox-
yl radical, as compared to the 2,4,6-tri-tert-butyl phenoxyl
radical. The redox potential of these phenolate compounds
is strongly influenced by H-bonding. The analyses of the ex-
perimental data coupled with theoretical calculations strong-
ly indicate that the effect of the H-bonding on the redox po-
c
a
spectively; ip (and ip ) / (scan rate)1/2. For Fc/Fc+: DE=60–88 mV (scan
rate=20–900 mVsꢀ1). Thus, the process is considered to be fully reversi-
ble. CPE of [NHOHL]
ACTHNUGTRENNG[U NBu4] performed in acetonitrile at room tempera-
ture at 0.360 V versus Ag/AgCl indicated the transfer of 0.97 electrons
per molecule. Thus, this is a one-electron oxidation process.
A
NMe2L]
E
E
1/2 =ꢀ0.423 V versus Fc/Fc+, Epa =ꢀ385 mV, DE=
c
a
77 mV at 120 mVsꢀ1. DE=69–89 mV (scan rate=20–900 mVsꢀ1); ip /ip
=
1ꢁ0.1 at scan rate ꢂ100 mVsꢀ1; at 20, 50 and 80 mVsꢀ1, ip /ip was 1.21,
c
a
1.21 and 1.33, respectively; ip (and ipa) / (scan rate)1/2. For Fc/Fc+: DE=
c
72–90 mV (scan rate=20–900 mVsꢀ1). Thus, the process is considered to
be fully reversible. CPE of [NMe2L]
ACHTUNGTRENNUNG
Chem. Eur. J. 2011, 17, 11882 – 11892
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11889

L. Benisvy, A. J. L. Pombeiro et al.
0.98 electrons per molecule. Thus, this is a one-electron oxidation pro-
cess.
Crystal structure analysis: Single crystals of NHOHLH, NMe2LH, [NHOHL]-
AHCTUNGTRENN[GNU NBu4]·2H2O and [ ACHTUGNETRN[NUGN NBu4] were obtained as indicated above.
NHMeL]
CCDC-782425, 782426, 823791 and 823792 contain the supplementary
crystallographic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[
NHMeL]
[NBu4]:
E
1/2 =ꢀ0.161 V versus Fc/Fc+, Epa =ꢀ129 mV, DE=
c
a
65 mV at 120 mVsꢀ1. DE=65–99 mV (scan rate=20–900 mVsꢀ1); ip /ip
=
c
a
c
1ꢁ0.1 at scan rate ꢂ50 mVsꢀ1; at 20, ip /ip was 1.27; ip (and ipa) / (scan
rate)1/2. For Fc/Fc+: DE=69–89 mV (scan rate=20–900 mVsꢀ1). Thus,
the process is considered to be fully reversible. CPE of [NHMeL]
ACTHNUTRGEN[UNG NBu4]
Intensity data were collected at 150 K, by using a Bruker AXS-KAPPA
APEX II diffractometer with graphite-monochromated MoKa (l=
0.71069 nm) radiation. Data were collected using omega scans of 0.58 per
frame and full spheres of data were obtained. Cell parameters were re-
trieved by using Bruker SMART software and refined with Bruker
SAINT[47] on all the observed reflections. Absorption corrections were
applied by using SADABS.[47] Structures were solved by direct methods
by using the SHELXS-97 package[48] and refined with SHELXL-97.[49]
Calculations were performed with the WinGX System,Version 1.80.03.[50]
performed in acetonitrile at room temperature at 0.320 V versus Ag/
AgCl indicated the transfer of 0.98 electrons per molecule. Thus, this is a
one-electron oxidation process.
EPR spectroscopy: EPR spectra were recorded on a Bruker ESP 300E
X-band spectrometer equipped with an ER 4111 VT variable-tempera-
ture unit; g values were calculated from the formula n=1.39962gB’, in
which n is the frequency measured (GHz) and B’ is the corrected value
of field (kGauss). The field correction (DB) was calculated from the dif-
ference between the theoretical and experimental values of field
(B’ꢀBexp) of our reference compound (i.e. perylene radical in concd. sul-
+
The NBu4 counterion of NHMeLꢀ is highly disordered and could not be
modelled. PLATON/SQUEEZE[51] was used to correct the data. A po-
tential volume of 2090 ꢄ3 was found with 587 electrons per unit cell
worth of scattering, which fits to the cation. All hydrogen atoms were in-
serted in calculated positions. Least-squares refinements with anisotropic
thermal motion parameters for all the non-hydrogen atoms and isotropic
parameters for the remaining atoms were employed.
furic acid) with known g value (i.e. 2.002569): B’=nexp/(1.39962g).
NHOH
C
L] (in CH3CN at 253–270 K): centre field: 3354.3358; modulation
[
frequency: 100 kHz; modulation amplitude: 1.0–0.1 Gpp; receiver gain:
4.48ꢅ10ꢀ4; conversion time: 40.96 ms; time constant: 81.92 ms; sweep
time: 167.77 s. Calculation for 263 K: g values were calculated from the
formula n=1.39962gB’, in which n is the frequency measured (GHz) and
B’ is the corrected value of field (kGauss). The field correction (DB) was
calculated from the difference of theoretical and experimental values of
field (B’ꢀBexp) of our reference compound (i.e. perylene radical in concd.
sulfuric acid) with known g value (i.e. 2.002569): B’=nexp/(1.39962g). Per-
N-3,5-Di-tert-butylsalicyloxysuccinimide: The first step consisted in the
elimination of water from commercial 3,5-di-tert-butylsalicylic acid mono-
hydrate, by dissolving it in diethyl ether and leaving to dry for 1 day
over Na2SO4. The solution was then filtered and evaporated to give the
dried compound. The latter (10 g, 72.4 mmol) and N-hydroxysuccinimide
(9.2 g, 79.6 mmol, 1.1 equiv) were dissolved in dioxane (250 mL). The so-
lution was cooled to 08C and a solution of dicyclohexylcarbodiimide
(16.4 g, 79.6 mmol, 1.1 equiv) in dioxane (200 mL) was added dropwise.
The reaction mixture was left stirring at room temperature under an
inert atmosphere for 24 h, with formation of a white precipitate (urea de-
rivative). The mixture was filtered using a Buchner funnel and the filtrate
ylene radical in concd. sulfuric acid: g=2.002569; Bexp =3345.5890; nexp
=
9.409219 GHz; B’=3357.0356; DB=11.4466. [NHOHL] : Bexp =3341.3852;
C
C
n
exp =9.405720 GHz; B’=3352.8318; g=2.00433. [NHMeL] (in CH3CN at
253–270 K): centre field: 3354.6030; modulation frequency: 100 kHz;
modulation amplitude: 1.0–0.1 Gpp; receiver gain: 4.48ꢅ10ꢀ4; conversion
time: 40.96 ms; time constant: 81.92 ms; sweep time: 167.77 s. Calculation
was evaporated to dryness to give
a
white powder in 85% yield.
(phenol)), 7.82 (d, J4 =
for 263 K: perylene radical in concd. sulfuric acid: g=2.002569; Bexp
=
C
C
1H NMR (400 MHz, CDCl3): 10.11 (s, 1H; OH
AHCTUNGTRENNUNG
3347.2468; nexp =9.411541 GHz; B’=3357.8641; DB=10.6172. [NHMeL] :
2.5 Hz, 1H; ArH), 7.64 (d, J4 =2.5 Hz, 1H; ArH), 2.92 (s, 4H; -CH2-
CH2-), 1.41 (s, 9H; tBu), 1.30 ppm (s, 9H; tBu); 13C{1H} NMR (100 MHz,
CDCl3): 169.10 (CO), 159.75 (C-, Ar), 141.52 (Ar), 137.82 (Ar), 133.00
(HC(Ar)), 123.40 (HC(Ar)), 107.07 (Ar), 35.23 (C-, tBu), 34.37 (C-, tBu),
31.23 (CH3, tBu), 29.28 (CH3, tBu), 25.66 ppm (-CH2CH2-).
Bexp =3343.9866; nexp =9.410058 GHz; B’=3354.6038; g=2.00420. [NMe2L]
(in CH3CN at 253–270 K): centre field: 3353.3816; modulation frequency:
100 kHz; modulation amplitude: 1.0–0.1 Gpp; receiver gain: 4.48ꢅ10ꢀ4
;
conversion time: 40.96 ms; time constant: 81.92 ms; sweep time: 167.77 s.
Calculation for 263 K: perylene radical in concd. sulfuric acid: g=
3,5-Di-tert-butyl-2-hydroxy-N-(2-hydroxyethyl)benzamide, NHOHLH:
A
2.002569; Bexp =3348.1435;
n
exp =9.414326 GHz; B’=3358.8577; DB=
solution of ethanolamine (0.4 g, 6.7 mmol) in DMF (5 mL) was added to
a solution of N-3,5-di-tert-butylsalicyloxysuccinimide (2.0 g, 5.7 mmol) in
DMF (10 mL). Then, triethylamine (6 mL, pre-dried over NaOH) was
added to the reaction mixture. After a few minutes a precipitate formed
and the reaction mixture was left stirring for 24 h. The mixture was then
poured into ice/water 10% HCl (20 mL) and the white precipitate was
isolated by filtration through a Buchner funnel, collected and dried
under vacuum to give NHOHLH in 86% yield. 1H NMR (CD3CN): 13.31
10.7142. [NMe2L] : Bexp =3343.2112; nexp =9.410179 GHz; B’=3353.9254;
g=2.00463. In all cases the microwave power supplied to the resonator
was varied in the range 0.2–2 mW.
C
DFT calculations: The full geometry optimisation of all structures was
carried out at the DFT/Hartree–Fock (HF) hybrid level of theory using
Beckeꢆs three-parameter hybrid exchange functional in combination with
the gradient-corrected correlation functional of Lee, Yang and Parr
(B3LYP)[40,41] with the help of the Gaussian 98[42] program package. Re-
stricted approximations for the structures with closed electron shells and
unrestricted methods for the structures with open electron shells were
employed. No symmetry operations were applied. The standard 6-31+
G(d) basis set was used for all atoms. The Hessian matrix was calculated
analytically for the optimised structures to prove the location of correct
minima (no imaginary frequencies) and to estimate the thermodynamic
parameters, the latter being calculated at 258C. Several possible confor-
mations were calculated and only the most stable ones are discussed.
Total energies corrected for solvent effects (Es) were estimated at the
single-point calculations on the basis of gas-phase geometries using the
polarisable continuum model in the CPCM version[43,44] with CH3CN as
solvent. The entropic term in CH3CN solution (Ss) was calculated accord-
ing to the procedure described by Wertz[45] and Cooper and Ziegler.[46]
Adiabatic ionisation potentials were calculated as the difference of the
Gibbs free energies in solution (Gs,oxꢀGs,nox) or of the total energies in
the gas phase (Eg,oxꢀEg,nox), in which the index “nox” corresponds to the
non-oxidised species and the index “ox” corresponds to the oxidised spe-
cies with an unrelaxed geometry.
(s, 1H; OHACTHNUTRGNEUNG
(phenol)), 7.49–7.48 (m, 1H+1H; ArH+NH), 7.41 (d, J4 =
2.5 Hz, 1H; ArH), 3.86 (broad, 2H; CH2-OH), 3.46 (q, 2H; CH2-
NHCO), 2.97 (broad, 1H; OH), 1.40 (s, 9H; tBu), 1.31 ppm (s, 9H; tBu);
13C{1H} NMR (100 MHz, CDCl3): 172.01 (CO), 158.67 (C-OH, Ar),
139.90 (C-tBu, Ar), 137.94 (C-tBu, Ar), 128.81 (CH, Ar), 119.69 (CH,
Ar), 113.10 (C-CO, Ar), 61.86 (CH2-OH), 42.37 (CH2-NH), 35.16 (C-
(CH3)3), 34.31 (C-(CH3)3), 31.48 (CH3, tBu), 29.35 ppm (CH3, tBu); MS
(EI (+)): m/z (%): 294 [M+H]+ , 316 [M+Na]+; IR (KBr pellets): 3486
(N H), 3313 (PhO H), 1624 (C=O), 1586, 1554 cmꢀ1; (5 mm sol in
ꢀ
ꢀ
CH3CN): 3411 (N H), 1634 (C=O) cmꢀ1; elemental analysis calcd (%)
ꢀ
for C17H27NO3 (293.19): C 69.59, H 9.28, N 4.77; found: C 69.26, H 8.82,
N 4.93.
3,5-Di-tert-butyl-2-hydroxy-N-dimethylbenzamide, NMe2LH: A 2.0m solu-
tion in THF of dimethylamine (3.45 mL, 6.9 mmol) diluted with DMF
(3 mL) was added to a solution of N-3,5-di-tert-butylsalicyloxysuccinimide
(2.0 g, 5.7 mmol) in DMF (13 mL). Then, triethylamine (3 mL, pre-dried
over NaOH) was added to the reaction mixture. After few minutes the
mixture became milky and a solid started to form. The suspension was
left stirring for 24 h, then poured into ice/water 10% HCl (20 mL) and
11890
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11882 – 11892

Persistent H-Bonded and Non-H-Bonded Phenoxyl Radicals
FULL PAPER
the milky final mixture was stirred at 08C for 10 min. The white precipi-
tate was removed by filtration through a Buchner funnel, collected and
dried under vacuum to give NMe2LH in 80% yield. 1H NMR (CD3CN):
hydroxide (7.2 mL, 7.2 mmol) was added to a solution of 3,5-di-tert-butyl-
2-hydroxy-N-dimethylbenzamide (2 g, 7.2 mmol) in dry methanol
(10 mL). The colourless solution turned pale yellow and was stirred
under N2 atmosphere for 5 h. Then the solvent was removed under
vacuum leaving a yellow oil. This crude oil was suspended in Et2O, the
mixture was stirred for 2 h and the solvent was gently decanted off. The
residue was dried under vacuum to leave the corresponding salt [NMe2L]-
10.33 (s, 1H; OHACHTUNGTRENNUNG
(phenol)), 7.41 (d, J4 =2.4 Hz, 1H; ArH), 7.26 (d, J4 =
2.4 Hz, 1H; ArH), 3.09 (s, 6H; (CH3)2N), 1.40 (s, 9H; tBu), 1.29 ppm (s,
9H; tBu); 13C{1H} NMR (100 MHz, CDCl3): 173.44 (CO), 159.82 (C-OH,
Ar), 139.61 (C-tBu, Ar), 137.52 (C-tBu, Ar), 126.84 (CH, Ar), 123.16
(CH, Ar), 116.42 (C-CO, Ar), 36.49 ((CH3)2N), 35.24 (C-(CH3)3), 34.32
(C-(CH3)3), 31.57 (CH3, tBu), 29.58 ppm (CH3, tBu); 13C{1H} NMR
(CD3CN): 173.51 (CO), 156.28 (C-OH, Ar), 140.78 (C-tBu, Ar), 137.74
(C-tBu, Ar), 127.52 (CH, Ar), 124.47 (CH, Ar), 117.84 (C-CO, Ar), 38.73
(broad, (CH3)2N), 35.74 (C-(CH3)3), 34.93 (C-(CH3)3), 31.62 (CH3, tBu),
AHCTUNGTRENNUNG
[NBu4] in 89% yield. 1H NMR (400 MHz, CD3CN): 7.69 (br, 1H; ArH),
7.23 (br, 1H; ArH), 3.11–3.05 (m, 8H; -CH2-N, [NBu4]), 2.98 (s, 6H,;
(CH3)2-N), 1.64–1.53 (m, 8H; -CH2-, [NBu4]), 1.39–1.29 (m, 9H+8H;
tBu+[NBu4]), 1.24 (s, 9H; tBu), 1.00–0.94 ppm (m, 12H; CH3-, [NBu4]);
13C{1H} NMR (100 MHz, CDCl3: 173.10 (CO), 159.01 (C-Oꢀ, Ar), 138.26
(C-tBu, Ar), 136.76 (C-tBu, Ar), 126.11 (CH, Ar), 124.39 (CH, Ar),
117.24 (C-CO, Ar), 59.12 (-CH2-N, [NBu4]), 36.21 ((CH3)2N), 35.87 (C-
(CH3)3), 34.98 (C-(CH3)3), 31.23 (CH3, tBu), 29.65 (CH3, tBu), 23.71
(-CH2-, [NBu4]), 19.64 (-CH2-, [NBu4]), 13.48 ppm (-CH3, [NBu4]); MS
(EI (ꢀ)): m/z (%): 262 [NMe2L]ꢀ; MS (EI (+)): m/z (%): 242 [NBu4]+; IR
(nujol): 1638 (C=O) cmꢀ1; (5 mm sol in CH3CN): 1633 (C=O) cmꢀ1; ele-
mental analysis calcd (%) for C33H62N2O2 (518.47): C 76.40, H 12.04, N
5.40; found.: C 76.75, H 12.32, N 5.21.
29.74 ppm (CH3, tBu) MS (EI (+)): m/z (%): 278 [M+H]+, 300
+
ꢀ
[M+Na] ; IR (KBr pellets): 3127 (PhO H), 1612 (C=O), 1586, 1478,
1406 cmꢀ1; (5 mm sol in CH3CN): 1622 (C=O) cmꢀ1; elemental analysis
calcd (%) for C17H27NO2 (277.20): C 73.60, H 9.81, N 5.05; found: C
73.51, H 10.36, N 5.00.
3,5-Di-tert-butyl-2-hydroxy-N-methylbenzamide, NHMeLH: A 2.0m solu-
tion in THF of methylamine (4.6 mL, 9.26 mmol) was added to a solution
of N-3,5-di-tert-butylsalicyloxysuccinimide (2.8 g, 8.1 mmol) in DMF
(15 mL). Then, triethylamine (8 mL, pre-dried over NaOH) was added to
the reaction mixture. After a few minutes the mixture became milky and
a solid started to form. The suspension was left stirring for 24 h. After
this time the pale yellow mixture was poured into ice/water 10% HCl
(30 mL) and the final mixture was stirred at 08C for 5 min and filtered
through a Buchner funnel. The solid was collected, washed with pentane
and dried under vacuum to give NHMeLH in 82% yield. 1H NMR
2,4-Di-tert-butyl-6-(2-methylcarbamoyl)phenolate tetrabutylammonium
salt, [NHMeL]
ACTHNUTRGEN[NUG NBu4]: A 1.0m solution in methanol of tetrabutylammonium
hydroxide (3.8 mL, 3.8 mmol) was added to a solution of 3,5-di-tert-butyl-
2-hydroxy-N-methylbenzamide (1 g, 3.8 mmol) in dry methanol (5 mL).
The colourless solution turned pale yellow and was stirred under N2 at-
mosphere for 5 h. Then the solvent was removed under vacuum leaving a
yellow oil. This crude oil was suspended in Et2O, the mixture was stirred
for 2 h and the solvent was gently decanted off. The sticky off-white solid
(CD3CN): 13.39 (s, 1H; OHACHTNUTGRNEUNG
(phenol)), 7.47 (d, J4 =2.4 Hz, 1H; ArH),
was dried under vacuum to leave the corresponding salt [NHMeL]
ACHTUNGTRENNUNG[NBu4] in
7.37 (m, 1H+1H; ArH+NH), 2.87 (d, J3 =4.8 Hz, 3H; H3C-NH), 1.40
(s, 9H; tBu), 1.30 (s, 9H; tBu); 13C{1H} NMR (100 MHz, CDCl3): 172.01
(CO), 158.71 (C-OH, Ar), 139.93 (C-tBu, Ar), 138.26 (C-tBu, Ar), 128.80
(CH, Ar), 119.15 (CH, Ar), 113.42 (C-CO, Ar), 35.32 (C-(CH3)3), 34.41
(C-(CH3)3), 31.62 (CH3, tBu), 29.50 (CH3, tBu), 26.65 ppm (H3C-N);
13C{1H} NMR (CD3CN): 172.70 (CO), 159.40 (C-OH, Ar), 140.89 (C-tBu,
Ar), 138.11 (C-tBu, Ar), 129.24 (CH, Ar), 121.40 (CH, Ar), 114.34 (C-
CO, Ar), 35.71 (C-(CH3)3), 35.04 (C-(CH3)3), 31.67 (CH3, tBu), 29.61
quantitative yield. 1H NMR (400 MHz, CD3CN): 13.20 (br, 1H; NH),
7.63–7.08 (br, 2H; ArH), 3.08–3.04 (m, 8H; -CH2-N, [NBu4]), 2.80 (s,
3H; CH3-NH), 1.61–1.54 (m, 8H; -CH2-, [NBu4]), 1.38–1.29 (m, 9H+
8H; tBu+[NBu4]), 1.23 (s, 9H; tBu), 0.97 ppm (t, J3 =7.0 Hz, 12H; CH3-,
[NBu4]); 13C{1H}-NMR (100 MHz, CD3CN): 172.34 (CO), n.d. (C-Oꢀ,
Ar), 139.52 (C-tBu, Ar), 128.43 (C-tBu, Ar), 125.86 (HC(Ar)), 124.04
(HC(Ar)), 59.19 (-CH2-N, [NBu4]), 35.85 (C-, tBu), 34.23 (C-, tBu), 32.32
(CH3, tBu), 29.93 (CH3, tBu), 25.32 (CH3-NH), 24.23 (-CH2-, [NBu4]),
20.24 (-CH2-, [NBu4]), 13.74 ppm (-CH3, [NBu4]); MS (EI (ꢀ)): m/z (%):
(CH3, tBu) 26.32 ppm (H3C-N); MS (EI (+)): m/z (%): 264 [M+H]+
,
+
ꢀ
ꢀ
286 [M+Na] ; IR (KBr pellets): 3419 (N H), 3328 (PhO H), 1620 (C=
262 [NHMeL]ꢀ; MS (EI (+)): m/z (%): 242 [NBu4]+; IR (KBr pellets):
O), 1582, 1544 cmꢀ1; (5 mm sol in CH3CN): 3413 (N H), 1639 (C=O)
ꢀ
ꢀ1
ꢀ
ꢀ
3274 (N H), 1623 (C=O), 1584 cm ; (5 mm sol in CH3CN): 3009 (N H),
1626 (C=O) cmꢀ1; elemental analysis calcd (%) for C32H60N2O2 (504.47):
C 76.13, H 11.97, N 5.55; found: C 75.63, H 10.20, N 5.90. X-ray-quality
single crystals were grown by slow diffusion of dry diethyl ether in a con-
centrated solution of the title compound in dichloromethane, under a di-
nitrogen atmosphere.
cmꢀ1; elemental analysis calcd (%) for C16H25NO2 (263.18): C 72.96, H
9.57, N 5.32; found: C 72.89, H 10.09, N 5.29.
2,4-Di-tert-butyl-6-(2-hydroxyethylcarbamoyl)phenolate tetrabutylammo-
nium salt, [NHOHL]
ACHTUNGTRENNUNG[NBu4]: A 1.0m solution in methanol of tetrabutylam-
monium hydroxide (5.1 mL, 5.1 mmol) was added to a solution of 3,5-di-
tert-butyl-2-hydroxy-N-(2-hydroxyethyl)benzamide (1.5 g, 5.1 mmol) in
dry methanol (10 mL). The transparent solution was stirred under N2 at-
mosphere for 5 h. Then the solvent was removed under vacuum leaving a
white power that was crystallised from CH2Cl2/Et2O to give the corre-
ACHTUNGTRENNUNG
sponding salt [NHOHL][NBu4] in 72% yield. 1H/15N-HSQC (CD3CN) and
1H NMR (CD3CN): 13.90 (br, 1H; NH), 7.63 (br, 1H; ArH), 7.13 (br,
1H; ArH), 5.65 (br, 1H; OH), 3.58 (m, 2H; CH2-OH), 3.42 (m, 2H;
CH2-NHCO), 3.10–3.04 (m, 8H; -CH2-N, [NBu4]), 1.65–1.54 (m, 8H;
-CH2-, [NBu4]), 1.39–1.30 (m, 9H+8H; tBu+[NBu4]), 1.25 (s, 9H; tBu),
0.99–0.93 ppm (m, 12H; CH3-, [NBu4]); 13C{1H} NMR (100 MHz,
CDCl3): 172.18 (CO), 157.77 (C-Oꢀ, Ar), 137.72 (C-tBu, Ar), 136.10 (C-
tBu, Ar), 127.42 (HC(Ar)), 122.19 (HC(Ar)), 114.75 (C-CO, Ar), 61.76
(CH2-OH), 58.53 (-CH2-N, [NBu4]), 42.77 (CH2-NH), 35.03 (C-, tBu),
34.12 (C-, tBu), 31.64 (CH3, tBu), 29.54 (CH3, tBu), 23.80 (-CH2-,
[NBu4]), 19.54 (-CH2-, [NBu4]), 13.53 ppm (-CH3, [NBu4]); MS (EI (ꢀ)):
m/z (%): 292 [NHOHL]ꢀ; MS (EI (+)): m/z (%): 242 [NBu4]+; IR (KBr
Acknowledgements
This work was partially supported by the Foundation for Science and
Technology (FCT) for the grants SFRH/BD/23187/2005 and SFRH/BPD/
70953/2010 (R.W.). Dr. J. P. Telo and Dr. M. C. Oliviera are also acknowl-
edged for assistance with EPR and MS measurements.
[1] J. Stubbe, W. A. van der Donk, Chem. Rev. 1998, 98, 705.
[2] K. E. Silva, T. E. Elgren, L. Que, M. T. Stankovich, Biochemistry
1995, 34, 14093.
[3] I. Vass, S. Styring, Biochemistry 1991, 30, 830.
[4] T. Maki, Y. Araki, Y. Ishida, O. Onomura, Y. Matsumura, J. Am.
Chem. Soc. 2001, 123, 3371.
[5] I. J. Rhile, J. M. Mayer, J. Am. Chem. Soc. 2004, 126, 12718.
[6] F. Thomas, O. Jarjayes, M. Jamet, S. Hamman, E. Saint-Aman, C.
Duboc, J. L. Pierre, Angew. Chem. 2004, 116, 604; Angew. Chem.
Int. Ed. 2004, 43, 594.
pellets): 3400 (CH2O H), 3114 (N H), 1617 (C=O), 1584 cmꢀ1; (5 mm sol
ꢀ
ꢀ
in CH3CN): 3005 (N H), 2966, 1624 (C=O), 1572 cmꢀ1; elemental analy-
ꢀ
sis calcd (%) for C33H62N2O3 (534.47): C 74.11, H 11.68, N 5.24; found: C
74.10, H 12.48, N 5.33. X-ray-quality single crystals were grown by slow
diffusion of dry diethyl ether in a concentrated solution of the title com-
pound in dichloromethane, under a dinitrogen atmosphere.
2,4-Di-tert-butyl-6-(2-dimethylcarbamoyl)phenolate tetrabutylammonium
salt, [NMe2L]
ACHTUNGTRENNUNG[NBu4]: A 1.0m solution in methanol of tetrabutylammonium
Chem. Eur. J. 2011, 17, 11882 – 11892
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11891

L. Benisvy, A. J. L. Pombeiro et al.
[7] L. Benisvy, R. Bittl, E. Bothe, C. D. Garner, J. McMaster, S. Ross,
C. Teutloff, F. Neese, Angew. Chem. 2005, 117, 5448; Angew. Chem.
Int. Ed. 2005, 44, 5314.
[8] I. J. Rhile, J. M. Mayer, Angew. Chem. 2005, 117, 1624; Angew.
Chem. Int. Ed. 2005, 44, 1598.
[26] M. Lucarini, V. Mugnaini, G. F. Pedulli, J. Am. Chem. Soc. 2003,
125, 8318.
[27] P. J. OꢆMalley, J. Am. Chem. Soc. 1998, 120, 11732.
[28] M. Engstrçm, F. Himo, A. Graslund, B. Minaev, O. Vahtras, H.
Agren, J. Phys. Chem. A 2000, 104, 5149.
[9] L. Benisvy, E. Bill, A. J. Blake, D. Collison, E. S. Davies, C. D.
Garner, G. McArdle, E. J. L. McInnes, J. McMaster, S. H. K. Ross,
C. Wilson, Dalton Trans. 2006, 258.
[10] C. Costentin, M. Robert, J. M. Saveant, J. Am. Chem. Soc. 2006,
128, 4552.
[11] I. J. Rhile, T. F. Markle, H. Nagao, A. G. DiPasquale, O. P. Lam,
M. A. Lockwood, K. Rotter, J. M. Mayer, J. Am. Chem. Soc. 2006,
128, 6075.
[29] M. Kaupp, T. Gress, R. Reviakine, O. L. Malkina, V. G. Malkin, J.
Phys. Chem. B 2003, 107, 331.
[30] J. R. Asher, N. L. Doltsinis, M. Kaupp, J. Am. Chem. Soc. 2004, 126,
9854.
[31] S. Un, M. Atta, M. Fontecave, A. W. Rutherford, J. Am. Chem. Soc.
1995, 117, 10713.
[32] P. Faller, C. Goussias, A. W. Rutherford, S. Un, Proc. Natl. Acad.
Sci. USA 2003, 100, 8732.
[12] L. Benisvy, D. Hammond, D. J. Parker, E. S. Davies, C. D. Garner, J.
McMaster, C. Wilson, F. Neese, E. Bothe, R. Bittl, C. Teutloff, J.
Inorg. Biochem. 2007, 101, 1859.
[33] F. Lachaud, A. Quaranta, Y. Pellegrin, P. Dorlet, M.-F. Charlot, S.
Un, W. Leibl, A. Aukauloo, Angew. Chem. 2005, 117, 1560; Angew.
Chem. Int. Ed. 2005, 44, 1536.
[13] T. F. Markle, J. M. Mayer, Angew. Chem. 2008, 120, 750; Angew.
Chem. Int. Ed. 2008, 47, 738.
[34] A. L. Maniero, V. Chis, A. Zoleo, M. Brustolon, A. Mezzetti, J.
Phys. Chem. B 2008, 112, 3812.
[14] T. F. Markle, I. J. Rhile, A. G. DiPasquale, J. M. Mayer, Proc. Natl.
Acad. Sci. USA 2008, 105, 8185.
[35] a) P. J. van Dam, J. P. Willems, P. P. Schmidt, S. Potsch, A. L. Barra,
W. R. Hagen, B. M. Hoffman, K. K. Andersson, A. Graslund, J. Am.
Chem. Soc. 1998, 120, 5080; b) I. Garcia-Bosch, A. Company, C. W.
Cady, S. Styring, W. R. Browne, X. Ribas, M. Costas, Angew. Chem.
2011, 123, 5766–5771; Angew. Chem. Int. Ed. 2011, 50, 5648–5653.
[36] P. Dorlet, A. W. Rutherford, S. Un, Biochemistry 2000, 39, 7826.
[37] P. Allard, A. L. Barra, K. K. Andersson, P. P. Schmidt, M. Atta, A.
Graslund, J. Am. Chem. Soc. 1996, 118, 895.
[15] J. J. Warren, T. A. Tronic, J. M. Mayer, Chem. Rev. 2010, 110, 6961.
[16] C. A. Jimꢁnez, J. B. Belmar, Tetrahedron 2005, 61, 3933.
[17] D. Kanamori, T. A. Okamura, H. Yamamoto, N. Ueyama, Angew.
Chem. 2005, 117, 991; Angew. Chem. Int. Ed. 2005, 44, 969.
[18] D. Kanamori, A. Furukawa, T. Okamura, H. Yamamoto, N.
Ueyama, Org. Biomol. Chem. 2005, 3, 1453.
[19] One should note that, whilst crystals of [NHOHL]
N
[38] E. R. Altwicker, Chem. Rev. 1967, 67, 475.
water molecules, which establish an H-bonding network (see Fig-
ure SI2, Supporting Information), the bulk organic salt does not con-
tain water, as evidenced by the NMR, IR and micronalysis charac-
[39] E. Land, G. Porter, E. Strachan, Trans. Faraday Soc. 1961, 57, 1885.
[40] A. D. Becke, J. Chem. Phys. 1993, 98, 5648.
[41] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[42] Gaussian 98 (Revision A.9), M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski,
J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q.
Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P. M. W. Gill, B. G. Johnson, W.
Chen, M. W. Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle,
J. A. Pople, Gaussian, Inc., Pittsburgh, PA, 1998.
terisations. In [OHNHL]
ACHTNUTRGNEUGN[NBu4]·2H2O, quasi-planar dimers are formed
through mutual intermolecular H-bonding interaction between the
alcohol pendant arm O(3)–H(3) of one molecule and phenolate–
O(1) from another molecule. The latter dimers [NHOHL]2 are con-
nected through H-bonding with tetramer water clusters (Fig-
ure SI2a), thereby giving rise to a hydrophilic H-bonding chain sand-
wiched by a hydrocarbon shell generated by the tert-butyl groups.
The arrangement of cations and anions in these organic salts is dif-
ferent (Figure SI2b). In [NHOHL]
ACHTNUGTRNE[UNG NBu4]·2H2O the hydrophobic NBu4
cations are sandwiched between hydrophilic chains generated from
the NHOHL anions and the water molecules (Figure SI2b). The NHOHL
anions in [NHOHL]
ACHTUNGTRENNUNG[NBu4] can be envisaged as trapped in nutshells of
hydrophobic NBu4 cations (Figure SI3).
[20] G. A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford Uni-
versity Press, Oxford, 1997.
[21] D. Kanamori, T. Okamura, H. Yamamoto, S. Shimizu, Y. Tsujimoto,
N. Ueyama, Bull. Chem. Soc. Jap. 2004, 77, 2057.
[22] a) L. Benisvy, I. Mutikainen, M. Quesada, U. Turpeinen, P. Gamez,
J. Reedijk, Chem. Commun. 2006, 3723; b) L. Benisvy, P. Gamez,
W. T. Fu, H. Kooijman, A. L. Spek, A. Meijerink, J. Reedijk, Dalton
Trans. 2008, 3147.
[43] J. Tomasi, M. Persico, Chem. Rev. 1994, 94, 2027.
[44] V. Barone, M. Cossi, J. Phys. Chem. A 1998, 102, 1995.
[45] D. H. Wertz, J. Am. Chem. Soc. 1980, 102, 5316.
[46] J. Cooper, T. Ziegler, Inorg. Chem. 2002, 41, 6614.
[47] Bruker, APEX2 & SAINT, Madison, Wisconsin, 2004.
[48] G. M. Sheldrick, Acta Crystallogr. Sect. A 1990, 46, 467.
[49] G. M. Sheldrick, Acta Crystallogr. Sec. A 2008, 64, 112.
[50] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837.
[23] C. Xie, P. M. Lahti, C. George, Org. Lett. 2000, 2, 3417.
[24] M. A. Hossain, F. Thomas, S. Hamman, E. Saint-Aman, D. Boturyn,
P. Dumy, J. Pept. Sci. 2006, 12, 612.
[51] A. L. Spek, Acta Crystallogr. Sect. C 1990, 46, C34.
Received: May 16, 2011
[25] F. G. Bordwell, J. P. Cheng, J. Am. Chem. Soc. 1991, 113, 1736.
Published online: September 5, 2011
11892
www.chemeurj.org
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11882 – 11892