First characterisation of rotational conformers in a chiral nitroxide by EPR spectroscopy

Article abstract of DOI:10.1039/b200323f

The detection and characterisation in liquid solution by EPR spectroscopy of the rotational conformers of a nitroxide radical containing a chiral centre is reported for the first time.

Full text of DOI:10.1039/b200323f

First characterisation of rotational conformers in a chiral nitroxide by
EPR spectroscopy†
Paola Franchi,a Marco Lucarini,*a Gian Franco Pedullia and Elisa Bandinib
a Dipartimento di Chimica Organica ‘A. Mangini’, Università di Bologna, Via S. Donato 15, I-40127
Bologna, Italy. E-mail: lucarini@alma.unibo.it
b I.Co.C.E.A, Consiglio Nazionale delle Ricerche, Via P. Gobetti 101, I-40129 Bologna, Italy
Received (in Cambridge, UK) 8th January 2002, Accepted 31st January 2002
First published as an Advance Article on the web 15th February 2002
The detection and characterisation in liquid solution by EPR
spectroscopy of the rotational conformers of a nitroxide
radical containing a chiral centre is reported for the first
time.
The evaluation of the relative importance of these two terms
is not an easy task since the only reliable method for their
estimation would be to record spectra at temperatures low
enough to freeze the internal rotation around the N–Ca bond on
the EPR time scale. Under these conditions the spectra of each
radical rotamer would be observed and these will provide both
the methylene hydrogen splittings and the relative populations
of each conformation. To the best of our knowledge, no
example has yet been reported of chiral free radicals where
rotamers have been detected by EPR and therefore the relative
importance of these two terms could never be experimentally
determined.
We wish to report here an EPR study of an optical pure
nitroxide of formula XYZC–CH2–N(O·)–CMe3 in which the
rotation about the N–Ca bond could be completely frozen at
easily accessible temperatures (5 276 °C). The nitroxides 2b
and 3b were generated directly inside the EPR cavity by
oxidising the chiral amines 2a or 3a (1.0 3 1023 M), obtained
as shown in Scheme 2, with m-choloroperbenzoic acid (1.0 3
1023 M) in toluene. Three groups X, Y, and Z, other than
hydrogen, were chosen as substituents at the chiral carbon on
the basis of the observation reported by Gilbert and Trenwith1
that a strong broadening of the two inner lines of the doublet of
doublets due to the methylene protons, is observed when
substituents bulkier than hydrogen are present.4
Disappointingly, in the room temperature EPR spectrum of
nitroxide 2b (X = OH, Y = CF3, Z = Ph) only a weak line
broadening was observed, this indicating that the rotation about
the N–Ca bond is fast on the EPR time scale (aH1 = 1.058, aH2
= 1.158 mT, aN = 1.560 mT, aF = 0.090 mT, g = 2.0061).
A more interesting behaviour was instead observed with
nitroxide 3b (X = OH, Y = CH3, Z = Ph) produced by
oxidation of the amine 3a that represents a minor product of the
reduction of the amide 1 with LiAlH4 whose yield, however,
can be increased by increasing the reaction time and the moles
of LiAlH4 added (see ESI†). Actually, the EPR spectrum of 3b
was strongly dependent on temperature and, at 95 °C, consisted
of twelve lines (Fig. 1a) due to the coupling of the unpaired
electron with the nitrogen nucleus (aN = 1.540 mT) and with
The EPR spectroscopic observation of free radicals in solution
showing unequivalent hyperfine splittings at the methylene
protons adjacent to a chiral carbon atom was reported for the
first time by Gilbert et al. 30 years ago.1,2 They discovered that
racemic mixtures of radicals of formula XYZC–CH2–NO2·2
and XYZC–CH2–N(O·)–CMe3 showed magnetically non-
equivalent b-hydrogens. Similar behaviour has been found in
more recent EPR investigations on related chiral radicals.3
The observed non-equivalence of the diastereotopic hydro-
gens originates from two different contributions: an intrinsic
term and a term depending on the conformer populations. In a
nitroxide radical of the type XYZC–CH2–N(O·)–C(CH3)3
where a prochiral methylene group is bonded to a carbon
bearing three different substituents, the lower energy conforma-
tions are those where the chiral centre avoids the tert-butyl
group (see Scheme 1).4
Since the nitroxide contains a Cb chiral centre the spatial
orientation of the two b-hydrogens with respect to both the
radical centre and the Cb substituents will differ for the two
rotamers. Consequently, the methylene protons will be magnet-
ically non-equivalent and the energies associated with the two
rotamers will likewise be different. When the interconversion
between the two unequivalent conformers is fast on the EPR
time scale the observed b-hydrogen splittings will be time
averaged over the residence time in each conformation. Since
the two conformations have different energies the two splittings
will be different, their values being closer to those of the
favoured conformer.
It should be pointed out that even in the case that all
conformations were equally populated, magnetic nonequiva-
lence would still be observed since in each conformer the two
methylene splittings can be coincident only by accident. The
residual non-equivalence still present when all conformations
are energetically degenerate, is called intrinsic non-equiva-
lence.
Scheme 1
† Electronic supplementary data (ESI) available: preparation and spectro-
scopic characterization of 3a, Arrhenius plot for the rotation of the N–Ca
bond in 3b. See http://www.rsc.org/suppdata/cc/b2/b200323f/
Scheme 2
560
CHEM. COMMUN., 2002, 560–561
This journal is © The Royal Society of Chemistry 2002

two unequivalent protons (aH1 = 1.016, aH2 = 1.243 mT). At
this temperature the observed b-proton splittings are averaged
over the residence times in each conformation indicating that
the rotation about the N–Ca bond is fast on the EPR time scale,
although the broadening of the MI(2H) = 0 lines suggests that
the rate of conformational interconversion is not fast enough to
reach the motional narrowing limit. By lowering the tem-
perature to 20 °C, the latter lines became too broad to be
detected (Fig. 1b) and only the sum of the b-H splittings could
be measured. (aN = 1.539, aH1+aH2 = 2.251 mT, g = 2.0060).
At 276 °C new lines, corresponding to two distinct para-
magnetic rotamers, appeared in the spectrum (Fig. 1c), this
indicating that the rotation about the N–Ca bond is slow on the
EPR time scale at this temperature. Simulation of the latter
spectrum (Fig. 1d) provided the values of both the hyperfine
splitting constants and relative amounts of the frozen rotamers
(see Scheme 1).5
The measured data, reported in Scheme 1 show that in both
rotamers the b-proton splittings are almost identical, one being
large and the other close to zero. On the basis of the b-proton
splitting dependence on the dihedral angle q between the
symmetry axes of the 2pz orbital containing the unpaired
electron and the N–C–H plane, predicted by the well-known
Heller-McConnell equation,6 it can be inferred that the
minimum energy conformations of the radical are those where
one of the methylene C–H bond makes an angle of ca. 30° with
the 2pz orbital on nitrogen and the other is almost eclipsed by
the tert-butyl group (see Scheme 1).
By simulating the experimental EPR spectra at different
temperatures by using well-established procedures7 based on
the density matrix theory,8 we could determine the activaction
parameters for the N–Ca bond rotation as DS‡ = 0.60 0.25
e.u. DH‡ = 5.97 0.08 kcal mol21 and DS‡ = 0.41 0.20 e.u.
DH# = 6.11 0.07 kcal mol21 (see ESI†). These values give a
free energy difference of 0.2 kcal mol21 and a population ratio
of 1.41 for the two rotamers at 20 °C.
Given the similarity of the b-proton splittings of the two
rotamers, the main factor determining the magnetic non-
equivalence of these splitting is the free energy difference of the
two conformations with the intrinsic term being responsible for
9
only 18% of the observed value of DaHb
.
In conclusion, it has been demonstrated, for the first time, that
the observed magnetic non-equivalence of diastereomeric
protons in the EPR spectra of chiral radicals mainly arise from
the population difference of the various conformations. It
should be pointed out that the optical pure nitroxide 3b
represents a suitable probe to investigate the effect of chiral
perturbations on the conformational equilibrium of a radical
containing a chiral centre. This can be important in determining
the factors giving rise to preferential stabilisation of a given
conformation and, consequently, to stereoselective reactions of
acyclic radicals.10
Financial support by the University of Bologna, MURST, and
CNR (Rome) is gratefully acknowledged. We thank Dr E.
Mezzina for helpful discussions.
Notes and references
1 B. C. Gilbert and M. Trenwith, J. Chem. Soc., Perkin Trans. 2, 1973,
1834.
2 B. C. Gilbert, J. P. Larkin and R. O. C. Norman, J. Chem. Soc., Perkin
Trans. 2, 1971, 1272.
3 See, for example:(a) G. Lagercrantz and M. Setaka, Acta Chem. Scand.
B, 1974, 28, 619; (b) C. C. Felix and R. C. Sealy, J. Am. Chem. Soc.,
1981, 103, 2831; F. Moriya, K. Makino, N. Suzuki, S. Rokushika and H.
Hatano, J. Am. Chem. Soc., 1982, 104, 830; (c) S. Brumby, J. Phys.
Chem., 1983, 87, 1917; (d) Y. Zhang and G. Xu, Magn. Reson. Chem.,
1989, 27, 846; (e) P. Smith, W. H. Donovan, C. E. Mader, L. M.
Dominguez and W. T. Koscielniak, Magn. Reson. Chem., 1995, 33,
395.
4 Actually, only the two rotamers with the chiral substituent avoiding the
tert-butyl group have been detected in radical 3b (vide infra).
5 The same EPR spectra were observed by using the (R) or the (S)
enantiomer of 3b, as expected.
6 C. Heller and H. M. McConnell, J. Chem. Phys., 1960, 32, 1535.
7 (a) J. H. Freed and G. K. Fraenkel, J. Chem. Phys., 1963, 39, 326; (b) A.
Hudson and G. R. Luckhurst, Chem. Rev., 1969, 69, 191.
8 P. Franchi, M. Lucarini, G. F. Pedulli and D. Sciotto, Angew. Chem., Int.
Ed., 2000, 39, 263.
9 If the energies of the two rotamers were identical, the values of aH1 and
aH2 would be 1.086 and 1.127 mT (DaHb = 0.041 mT) in the averaged
spectrum.
Fig. 1 EPR spectra of (R)-3b recorded in toluene at 95 °C (a), 20 °C (b), 276
°C (c) and computer simulation of the latter one (d).
10 D. P. Curran, N. A. Porter and B. Giese Stereochemistry of Radical
Reactions, VCH, Weinheim, 1996.
CHEM. COMMUN., 2002, 560–561
561