Structure, conformation, stereodynamics, and absolute configuration of the atropisomers of fluorenylidene derivatives

Article abstract of DOI:10.1021/jo1023427

The barrier for the interconversion of the conformational atropisomers of an aryl fluorenylidene derivative was determined by variable-temperature NMR technique. In the case of a more hindered compound the two atropisomers were isolated and the structure

Full text of DOI:10.1021/jo1023427

pubs.acs.org/joc
SCHEME 1
Structure, Conformation, Stereodynamics, and
Absolute Configuration of the Atropisomers
of Fluorenylidene Derivatives
Lodovico Lunazzi, Michele Mancinelli, and
Andrea Mazzanti*
Department of Organic Chemistry “A. Mangini”, University
of Bologna, Viale Risorgimento 4, Bologna 40136, Italy
of the isopropylphenyl ring is not coplanar with that of the fluor-
enylidene moiety. At low temperature the rotation about the
bond joining these two planes is slow in the NMR time scale
and in these conditions this bond corresponds to a stereo-
genic (chiral) axis. Consequently, the molecule displays an
asymmetric conformation (C1 point group) that entails the
existence of two conformational enantiomers (atropisomers):7
such an asymmetry makes the isopropyl methyl groups dia-
stereotopic, thus anisochronous.8 The measured ΔGq value
corresponds, therefore, to the rotation barrier required for
the interconversion of the two atropisomers. Indeed DFT
calculations9 show that the two mentioned planes are twisted
by 54.7° (see Figure S-1 of the SI) and also indicate that in the
transition state the two planes become essentially coplanar,
with the isopropyl group pointing toward the ethylenic
hydrogen. The energy computed for such a rotational transi-
tion state is 8.5 kcal mol-1 higher than the ground state, in
good agreement with the experimental barrier.
mazzand@ms.fci.unibo.it
Received November 25, 2010
The barrier for the interconversion of the conformational
atropisomers of an aryl fluorenylidene derivative was deter-
mined by variable-temperature NMR technique. In the case
of a more hindered compound the two atropisomers were
isolated and the structure determined by X-ray diffraction.
The absolute configuration was assigned by theoretical
interpretation of the Electronic Circular Dichroism spec-
trum (ECD).
When a bulkier substituent, like bromine, replaces the
1
ethylenic hydrogen (compound 2, X = Br) the H NMR
spectrum displays two diastereotopic methyl groups even at
ambient temperature. These signals remain anisochronous
also at higher temperatures (þ120 °C), implying that the two
atropisomers, generated by the chirality axis, could be con-
figurationally stable. This is confirmed by the analysis on an
enantioselective HPLC column, where compound 2 displays
two separated lines for the two atropisomers (Figure S-2 in
the SI). DFT calculations indicate that the transition state,
having the isopropylphenyl coplanar with bromine (Figure S-3
of the SI), has an energy 29 kcal mol-1 higher than the
ground state, when the isopropyl is pointing toward the
bromine atom. This transition state shows, in addition, that
the double bond is twisted by about 39°, in order to lower the
steric clash between the ortho-hydrogen of the phenyl and
the H-1 hydrogen of the fluorene ring.10 Such a theoretical
Hindered derivatives of fluorenyl and of analogous deri-
vatives have been synthesized and investigated as models for
studying the so-called molecular motors.1-6 With the pur-
pose of contributing to the comprehension of some of the
related properties we synthesized the fluorenyl derivatives
1-3 (Scheme 1) and investigated their stereochemistry by
means of dynamic NMR and ECD spectroscopy as well as by
X-ray diffraction and DFT calculations.
The variable-temperature 13C spectrum of compound 1
(X=H) shows that the methyl signal broadens on cooling
and eventually splits into a pair of equally intense lines
at -106 °C (Figure 1). The line shape simulation provides a
set of rate constants, corresponding to a free energy of activation
(ΔGq) of 8.9 kcal mol-1. This feature indicates that the plane
(7) Kuhn, R. Moleculare Asymmetrie. In Stereochemie; Freudenberg, K.,
Ed.; Franz Deuticke: Leipzig, 1933; pp 803-824. Oki, K. Recent Advances in
Atropisomerism. In Topics in Stereochemistry; Allinger, N. L., Eliel, E. E,
Wilen, S. H, Eds.; Wiley Interscience: New York, 1983; Vol. 14, pp 1-76. For
a review on atropisomerism see: Bringmann, G.; Price Mortimer, A. J.;
Keller, P. A.; Gresser, M. J.; Garner, J.; Breuning, M. Angew. Chem., Int. Ed.
2005, 44, 5384–5427.
(1) Feringa, B. L. J. Org. Chem. 2007, 72, 6635–6652.
(8) Mislow, K.; Raban, M. Top. Stereochem. 1967, 1, 1. Jennings, W. B.
Chem. Rev. 1975, 75, 307. Eliel, E. L. J. Chem. Educ. 1980, 57, 52. Casarini,
D.; Lunazzi, L.; Macciantelli, D. J. Chem. Soc., Perkin Trans. 2 1992, 1363–
1370.
(2) ter Wiel, M. K. J.; Feringa, B. L. Tetrahedron 2009, 65, 4332–4339.
(3) Geertsema, E. M.; van der Molen, S. J.; Martens, M.; Feringa, B. L.
Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 16919–16924.
(4) Kulago, H.; Mes, E. M.; Klok, M.; Meetsma, A.; Brouwer, A. M.;
Feringa, B. L. J. Org. Chem. 2010, 75, 666–679.
(5) Pijper, T. C; Pijper, D.; Pollard, M. M.; Dumur, F.; Davey, S. G.;
Meetsma, A.; Feringa, B. L. J. Org. Chem. 2010, 75, 825–838.
(6) Chen, W.-C.; Lee, Y.-W.; Chen, C.-Y. Org. Lett. 2010, 12, 1472–1475.
(9) B3LYP/6-31G(d) level. Gaussian 03, Revision E.01, Frisch, M. J. et al.;
Gaussian, Inc., Wallingford, CT, 2004. See the SI for the complete reference.
(10) In the alternative transition state where the isopropyl points away
from the bromine the energy is even higher (38 kcal mol-1) and the twisting
angle of the double bond is about 50°.
DOI: 10.1021/jo1023427
r
Published on Web 01/20/2011
J. Org. Chem. 2011, 76, 1487–1490 1487
2011 American Chemical Society

Lunazzi et al.
JOCNote
FIGURE 2. Top: DFT computed structure of the diasteroisomers
3a and 3b (in each case only one of the two possible enantiomers is
displayed). Bottom: Experimental X-ray structure of 3b.
To confirm this indication, we approached the determina-
tion of the absolute configuration of compound 3b by
making use of chirooptical techniques. In particular we
simulated the Electronic Circular Dichroism spectra by
TD-DFT calculations, because this technique has become
highly reliable and it has been successfully employed to assign
the absolute configuration of complex organic molecules.13
Starting from the geometry determined in the solid state, a
conformational search was carried out by minimizing all the
possible conformations obtainable by the 3-fold rotation of
the MeCHOH group and of the OH group. All the con-
formations were optimized by using DFT at the B3LYP/
6-31G(d) level, and the harmonic vibrational frequencies
of each conformation were calculated at the same level to
confirm their stability (no imaginary frequencies were
observed), and to evaluate the free energy of each conforma-
tion by ZPE correction. After DFT minimization, in addi-
tion to the conformation of 3b reported in Figure 2 (which
corresponds to the global minimum), two other conforma-
tions were found to fall in a 2 kcal mol-1 range (see Figure
S-5 in the SI). These three conformations were selected for
the simulation of the ECD spectrum. Using the optimized
geometries, the electronic excitation energies and rotational
strengths were calculated by using TD-DFT at the
BH&HLYP/6-311þþG(d,p) level14 and assuming S,M ab-
solute configuration. The results are shown in Figure S-6 of
the SI (GS-1 to GS-3).15 The patterns of the three spectra are
FIGURE 1. Temperature dependence of the 13C methyl signal
(150.8 MHz in CD2Cl2) of compound 1 (left) and line shape
simulations (right) with the rate constants reported.
barrier accounts for the observed configurational stability of
the atropisomers of 2. However, we were unable to isolate the
two atropisomers because of the very small chromatographic
separation, despite the various attempts on a number of
enantioselective columns. To achieve such a separation and
to maintain a very high enantiomerization barrier, we pre-
pared compound 3, where the MeCHOH group is intro-
duced in the place of bromine by reacting the lithiate of 2 with
acetaldehyde. Compound 3 bears, in addition to the chirality
axis, also a stereogenic carbon atom, so that two configur-
ationally stable diastereoisomers are expected to occur. The
synthetic procedure yields in fact two diastereoisomers, as
revealed by the HPLC separation showing the signals of a
major 3a (80%) and of a minor 3b (20%) compound as
shown in the middle trace of Figure S-4 of the Supporting
Information (first and second eluted, respectively). Each of
these diastereoisomers is expected to exist as a pair of
enantiomers due to the chirality axis and indeed when an
enantioselective HPLC column is employed, four peaks are
observed. The first and the second peak correspond to the
two enantiomers of 3b, the third and fourth peak correspond
to the two enantiomers of 3a (see Figure S-4 in the SI). X-ray
diffraction of a single crystal of the second eluted enantiomer
of the minor diastereoisomer 3b provided the structure re-
ported in Figure 2 (bottom, right). Although the absence of a
sufficiently heavy atom prevented an unambiguous assign-
ment of the absolute configuration by the anomalous dis-
persion method,11 the S,M configuration reported in Figure 2
gives a fit of the experimental data (Flack parameter12) that
is slightly better than that obtained by assuming the R,P
configuration.
(13) Mazzeo, G.; Giorgio, E.; Zanasi, R.; Berova, N.; Rosini, C. J. Org.
Chem. 2010, 75, 4600–4603. Pescitelli, G.; Di Pietro, S.; Cardellicchio, C.;
Annunziata, M.; Capozzi, M.; Di Bari, L. J. Org. Chem. 2010, 75, 1143–1154.
ꢀ
Jacquemin, D.; Perpete, E. A.; Ciofini, I.; Adamo, C.; Valero, R.; Zhao, Y.;
Truhlar, D. G. J. Chem. Theory Comput. 2010, 6, 2071–2085. Gioia, C.; Fini,
F.; Mazzanti, A.; Bernardi, L.; Ricci, A. J. Am. Chem. Soc. 2009, 131, 9614–
9615. Stephens, P. J.; Pan, J. J.; Devlin, F. J.; Cheeseman, J. R. J. Nat. Prod.
2008, 71, 285–288. For reviews see: Bringmann, G.; Bruhn, T.; Maksimenka,
K.; Hemberger, Y. Eur. J. Org. Chem. 2009, 2717–2727. Bringmann, G.;
Gulder, T. A. M.; Reichert, M.; Gulder, T. Chirality 2008, 20, 628–642.
Berova, N.; Di Bari, L.; Pescitelli, G. Chem. Soc. Rev. 2007, 36, 914.
(14) Gaussian 09, Revision A.01; Frisch, M. J. et al.; Gaussian, Inc.,
Wallingford, CT, 2009. See the SI for the complete reference. In Gaussian
HF
LSDA
09 the BH&HLYP functional has the form: 0.5EX
0.5ΔEX
þ 0.5EX
þ
þ ECLYP. Becke, A. D. J. Chem. Phys. 1993, 98, 1372–77.
Becke88
(11) Peerdeman, A. F.; van Bommel, A. J.; Bijvoet, J. M. Nature 1951,
168, 271–272. For a review on the use of X-ray Crystallography for the
determination of the absolute configuration see: Flack, H. D.; Bernardinelli,
G. Chirality 2008, 20, 681–690.
(15) The rotational strengths were calculated in the length and velocity
representations. The resulting values are very similar, and the errors due
to basis set incompleteness are therefore very small, or negligible. See:
Stephens, P. J.; McCann, D. M.; Devlin, F. J.; Cheeseman, J. R.; Frisch, M. J.
J. Am. Chem. Soc. 2004, 126, 7514–7521.
(12) Flack, H. D. Acta Crystallogr. 1983, A39, 876–881.
1488 J. Org. Chem. Vol. 76, No. 5, 2011

Lunazzi et al.
JOCNote
100 s). This suggests that CH3CHO (which is added only a
few seconds after the formation of the Li intermediate) reacts
with the Li intermediate when the latter is in a locked
conformation (see Figure S-9 of the SI). This feature might
account for the occurrence of the mentioned diastereoselec-
tivity.
As the measurement of the free energy barrier of the
diastereoisomers 3a and 3b is inaccessible to the dynamic
NMR technique, a kinetic approach was used. When a
DMSO/D2O (10:1 v/v) solution of the isolated diastereo-
isomer 3a (which is computed to be the least stable) is kept at
þ135 °C, the interconversion into 3b begins to take place
until the thermodynamic equilibrium is reached, whereby the
3b/3a ratio becomes 62:38, as opposed to the kinetic ratio 3b/
3a = 20:80. This proves that the diastereoisomer 3b is the
thermodynamically more stable species, as predicted by calc-
ulations. The kinetic path of this process was followed by
monitoring the variation of the NMR signal ratio as a
function of the time (see Figure S-10 of the SI) and the free
energy of activation for the 3b into 3a interconversion was
found equal to 34.6 kcal mol-1 (at þ135 °C).
FIGURE 3. Experimental ECD spectrum (black trace) of the sec-
ond eluted atropisomer of the minor diastereoisomer 3b and TD-
DFT computed traces (red and blue) obtained assuming the abso-
lute SM configuration.
different, in particular between the spectrum calculated for
the lowest energy ground state and the other two. For this
reason the shape of the final simulated spectrum could be
strongly influenced by errors in the evaluation of the relative
populations of the conformations. However, the best calcu-
lated geometry corresponds to the experimental structure
observed in the solid state. The same calculations were repeated
by using a different model (CAM-B3LYP16) in order to
check the solidity of the theoretical approach (Figure S-6
of the SI, bottom).17 The final simulated ECD spectra shown
in Figure 3 were obtained taking into account the 78:18:4
ratio derived from Boltzmann distribution based on the ZPE
corrected energy values of the three conformers of 3b. The
theoretical simulations obtained by the TD-DFT approach
employing two different functionals13,17 show that the ex-
perimental trace is reproduced quite satisfactorily by assum-
ing the S,M configuration: consequently the absolute con-
figuration of the other (first eluted) atropisomer of 3b must
be R,P.
The computed DFT structure of 3b (Figure 2, top right)
is very close to that experimentally determined by X-ray,
so that the computed structure for 3a (Figure 2, left) can
be likewise considered a satisfactory representation of the
major diastereoisomer.18 In addition, the ECD spectrum of
the second eluted enantiomer of the major diastereoisomer
3a was theoretically reproduced following the same proce-
dure reported above,9 by assuming the S,P configuration
(Figures S-7 and S-8 of the SI). It also should be stressed that
the computed energy for the diastereoisomer 3b is lower by
1.8 kcal mol-1 than that of 3a, whereas the proportion of the
latter (80%) was found much higher than that of 3b (20%) in
the reaction mixture. This apparent discrepancy could be due
to the fact that the reaction to produce 3 is under kinetic
rather than thermodynamic control. Indeed the Li inter-
mediate required to produce 3 by low-temperature reaction
with CH3CHO has a computed rotation barrier around the
Ar-CLi bond equal to 12.6 kcal mol-1. This implies that, at
the temperature of the formation of the lithiate (-85 °C), the
corresponding lifetime of the rotamer is quite long (about
The energy for the corresponding rotation transition state
(Figure S-11 of the SI) was computed to be 32.8 kcal mol-1
higher than that of the ground state of 3b.19
The calculated barrier is lower than the measured Gibbs
free energy barrier (34.6 kcal). This could be due to a negative
entropy of activation for the rotational process due to re-
stricted freedom of the proximate isopropyl and MeCHOH
groups in the rotational transition state TS-1.20
Experimental Section
9-(2-Isopropylbenzylidene)-9H-fluorene (1)21. To an oven-
dried round-bottomed flask were added 1.00 g (6.02 mmol)
of fluorene and 1.86 g of t-BuOK (16.54 mmol), followed by
50 mL of THF. The reaction was heated to reflux with vigorous
stirring, then 2-isopropylbenzaldehyde was added (1.07 g,
7.22 mmol), and the reaction was allowed to reflux overnight.
After cooling to room temperature, the reaction mixture was
poured into a large excess of 1 M HCl and crushed ice, and it was
extracted with Et2O. The organic layer was washed with brine
and dried over Na2SO4. After removal of the Et2O the crude was
purified by silica chromatography with hexanes/Et2O 10/1 v/v.
Anaytical samples were obtained by semipreparative HPLC
chromatography (Phenomenex Luna C8, 250 ꢀ 10 mm, eluent
CH3CN/H2O 90:10 v/v, 5 mL/min). 1H NMR (600 MHz,
CD2Cl2, 25 °C, δ 5.32): δ 1.21 (6H, d, J = 7.00 Hz), 3.23 (1H,
septet, J = 7.0 Hz), 6.98-7.04 (2H, m), 7.25-7.32 (2H, m),
7.35-7.48 (5H, m), 7.73 (1H, d, J = 7.6 Hz), 7.76 (1H, d, J =
7.4 Hz), 7.84 (1H, s), 7.87 (1H, d, J = 7.4 Hz). 13C NMR (150.8
MHz, CD2Cl2, 25 °C, δ 53.8): δ 23.6 (2 CH3), 31.35 (CH), 119.95
(CH), 120.0 (CH), 120.7 (CH), 125.0 (CH), 125.7 (CH), 126.2 (CH),
127.0 (CH), 127.3 (CH), 127.4 (CH), 128.5 (CH), 128.7 (CH), 128.9
(CH), 129.9 (CH), 135.8 (q), 137.2 (q), 137.3 (q), 139.5 (q), 139.6 (q),
(19) In the alternative transition state where the isopropyl points away
from the MeCHOH group the energy is even higher (45.6 kcal mol-1).
(20) An alternative conversion path between the two diastereoisomers
might correspond to a dehydration process of the secondary alcohol to
alkene, where the rotation of the phenyl moiety could take place with a lower
barrier, followed by rehydration. In this case, however, the presence of a large
molar excess of D2O in the NMR sample should lead to a partial (or total)
deuteration of the CH and Me groups, a feature that it is not experimentally
observed. This confirms that the energy barrier determined is that of the
rotation of the aryl group.
(16) Yanai, T.; Tew, D.; Handy, N. Chem. Phys. Lett. 2004, 393, 51–57.
(17) The use of more than one model enhances the reliability of the
method, and it can serve as an evaluation of the amplitude of the possible
error. See: Check, C. E.; Gilbert, T. M. J. Org. Chem. 2005, 70, 9828–9834.
(18) Neither of the enantiomers of the major diastereisomer 3a yielded
crystals suitable for a structural determination by X-ray diffraction.
(21) Dane, E. L.; Maly, T.; Debelouchina, G. T.; Griffin, R. G.; Swager,
T. M. Org. Lett. 2009, 11, 1871–1874.
J. Org. Chem. Vol. 76, No. 5, 2011 1489

Lunazzi et al.
JOCNote
141.3 (q), 147.7 (q). HRMS(EI): m/z calcd for C23H20 296.15650,
found 296.15611.
7.31-7.34 (1H, m), 7.37-7.43 (2H, m), 7.48-7.49 (2H, m), 7.64
(1H, d, J = 7.3 Hz), 7.76-7.77 (1H, m), 8.01 (1H, d, J = 7.6 Hz).
13C NMR (150.8 MHz, CDCl3, 25 °C, TMS): δ 21.8 (CH3), 24.4
(CH3), 25.0 (CH3), 31.2 (CH), 67.7 (CH), 119.2 (CH), 119.9
(CH), 125.9 (CH), 126.4 (CH), 126.6 (CH), 126.9 (2 CH), 127.5
(CH), 127.9 (CH), 128.3 (CH), 128.9 (CH), 129.5 (CH), 135.1
(Cq), 136.5 (Cq), 137.0 (Cq), 138.8 (Cq), 140.2 (Cq), 141.3 (Cq),
147.0 (Cq), 147.2 (Cq). HRMS(EI): m/z calcd for C25H24O:
340.18272, found 340.18244.
9-(Bromo(2-isopropylphenyl)methylene)-9H-fluorene (2). To
an oven-dried round-bottomed flask was added 1.17 g of 9-(2-
isopropylbenzylidene)-9H-fluorene (3.95 mmol, in 30 mL of
glacial acetic acid). The resulting suspension was heated to
reflux and additional acetic acid was added dropwise until a
homogeneous solution was formed. The mixture was cooled
to room temperature and a slight excess of a bromine solution
(4.75 mmol, 0.47 M in acetic acid) was added dropwise over
5 min. The resulting red solution was allowed to stir overnight at
room temperature until a white precipitate formed. Then NaOH
(0.158 g 3.95 mmol) and absolute ethanol (25 mL) were added.
The reaction was refluxed for 60 min and then cooled to room
temperature. The solution was acidified with 0.5 M HCl and
extracted with Et2O. The organic layer was then washed with
brine and dried over Na2SO4. After removal of the Et2O the
material was purified by silica chromatography with hexanes/
Et2O 10:1 v/v. Anaytical samples were obtained by semipre-
parative HPLC chromatography (Phenomenex Luna C8, 250 ꢀ
3b (2° eluted): 1H NMR (600 MHz, CD3CN, 25 °C, 1.95
ppm): δ 0.81 (3H, d, J = 6.7 Hz), 1.22 (3H, d, J = 7.0 Hz), 1.52
(3H, d, J = 6.7 Hz), 2.88 (1H, d, J = 6.2 Hz), 3.03 (1H, septet,
J = 6.7 Hz), 5.79 (1H, d, J = 8.1 Hz), 5.82 (1H, quintet, J =
6.6 Hz), 6.79 (1H, t, J = 7.5 Hz), 7.11 (1H, d, J = 7.9 Hz), 7.19
(1H, t, J = 7.3 Hz), 7.30-7.33 (1H, m), 7.40-7.45 (2H, m),
7.48-7.53 (2H, m), 7.73 (1H, d, J = 7.3 Hz), 7.83-7.86 (1H, m
Hz), 7.96-7.98 (1H, m). 13C NMR (150.8 MHz, CD3CN, 25 °C,
117.5 ppm): δ 22.5 (CH3), 23.2 (CH3), 24.0 (CH3), 31.6 (CH),
66.8 (CH), 119.3 (CH), 119.9 (CH), 125.5 (CH), 126.22 (CH),
126.23 (CH), 126.61 (CH), 126.63 (CH), 127.6 (CH), 127.7
(CH), 128.1 (CH), 128.5 (CH), 128.53 (CH), 132.3 (Cq), 137.2
(Cq), 137.5 (Cq), 138.8 (Cq), 140.0 (Cq), 140.7 (Cq), 148.1 (Cq),
151.2 (Cq). HRMS(EI): m/z calcd for C25H24O 340.18272,
found 340.18239
1
10 mm, eluent CH3CN/H2O 90:10 v/v, 5 mL/min). H NMR
(600 MHz, CD3CN, 25 °C, δ 1.95): δ 1.00 (3H, d, J = 7.0 Hz),
1.28 (3H, d, J = 7.0 Hz), 3.17 (1H, septet, J = 7.0 Hz), 6.02 (1H,
d, J = 7.9 Hz), 6.86-6.89 (1H, m), 7.26-7.32 (2H, m),
7.36-7.40 (1H, m), 7.46 (1H, dt, J = 1.2, 7.8 Hz), 7.50-7.58
(3H, m), 7.76 (1H, d, J = 7.5 Hz), 7.85 (1H, d, J = 7.5 Hz), 8.89
(1H, d, J = 7.9 Hz). 13C NMR (150.8 MHz, CD3CN, 25 °C,
δ 117.5): δ 23.1 (CH3), 23.4 (CH3), 30.6 (CH), 119.8 (CH),
120.1 (CH), 124.5 (q), 124.8 (CH), 126.2 (CH), 127.1 (CH),
127.2 (CH), 127.3 (CH), 127.5 (CH), 128.2 (CH), 128.8 (CH),
129.7 (CH), 130.3 (CH), 136.4 (q), 137.6 (q), 138.1 (q), 139.8 (q),
140.9 (q), 141.3 (q), 146.1 (q). HRMS(EI): m/z calcd for
C23H19Br 374.06701, found 374.06736.
1-(9H-fluoren-9-ylidene)-1-(2-isopropylphenyl)propan-2-ol
(3a and 3b). To an oven-dried round-bottomed flask was added
100 mg of 9-(bromo(2-isopropylphenyl)methylene)-9H-fluorene
(0.266 mmol,in 20 mL of dry THF). The mixture was cooled
to -85 °C and n-BuLi (0.18 mL, 0.29 mmol, 1.6 M in hexane,
1.1 equiv) was added dropwise. The solution became immediately
red, and after 1 min, 0.03 mL of acetaldehyde (neat, 0.53 mmol)
was added, yielding an almost colorless solution. The mixture
was then warmed to room temperature and the reaction was
quenched with H2O and extracted with Et2O. The organic layer
was washed with brine and dried over Na2SO4. After removal of
the Et2O the crude was purified by silica chromatography with
hexanes/Et2O 10:1 v/v. The resulting mixture of the two diaster-
eoisomers 3a and 3b was separated by semipreparative HPLC
chromatography (Phenomenex Synergy Polar-RP 4 μm, 250 ꢀ
21.2 mm, eluent CH3CN/H2O 90:10 v/v, 20 mL/min). Separa-
tion of the two enantiomers of 3a and 3b was obtained by
semipreparative HPLC on an enantioselective column (Daicel
Chiralcel AD-H 5 μm, 250 mm ꢀ 21.2 mm, eluent hexane/
iPrOH 95:5 v/v, 20 mL/min). 3a (1° eluted): 1H NMR (600 MHz,
CDCl3, 25 °C, TMS): δ 0.86 (3H, d, J = 6.7 Hz), 1.24 (3H, d, J =
7.0 Hz), 1.46 (3H, d, J = 6.7 Hz), 1.80 (1H, d, J = 5.5 Hz), 2.90
(1H, septet, J = 6.7 Hz), 5.79 (1H, quintet, J = 5.5 Hz), 5.85
(1H, d, J = 7.9 Hz), 6.79-6.83 (1H, m), 7.15-7.21 (2H, m),
ECD Spectra. Standard UV absorption spectra were recorded
at 25 °C in acetonitrile on the racemic mixtures, in the 200-
400 nm spectral region. Maximum molar absorption coeffi-
cients were recorded at 193 nm for 3b (ε = 59 300) and 193 nm
for 3a (ε = 54 400). ECD spectra were recorded at 24 °C in
acetonitrile solutions (5 ꢀ 10-5 M), using a path length of 0.2 cm.
Spectra were recorded in the range 180-400 nm; reported Δε
values are expressed as L mol-1 cm-1
.
NMR Spectroscopy. NMR spectra were obtained at 600 MHz
for 1H and at 150.8 MHz for 13C. The assignments of the 1H and
13C signals were obtained by bidimensional experiments (edited-
gHSQC and gHMBC sequences). The variable-temperature
1
spectra were recorded at 600 MHz for H and 150.8 MHz for
13C; temperature calibration and line shape simulation methods
were described elsewhere.22
Acknowledgment. The authors received financial support
from the University of Bologna (RFO) and from MIUR,
Rome (PRIN national project “Stereoselection in Organic
Synthesis, Methodologies and Applications”).
Supporting Information Available: DFT computed ground
states and transition states structures for 1, 2, 3a, and 3b;
enantioselective HPLC traces for compounds 2, 3a, and 3b;
computed components of the ECD spectrum of 3a and 3b, TD-
DFT simulation of the ECD spectrum of 3a, kinetics of the
thermal equilibration of 3a, crystallographic data of 3b, and 1H,
13C NMR spectra and computational data of 1-3. This material
is available free of charge via the Internet at http://pubs.acs.org.
(22) For a recent review see: Casarini, D.; Lunazzi, L.; Mazzanti, A. Eur.
J. Org. Chem. 2010, 2035–2056.
1490 J. Org. Chem. Vol. 76, No. 5, 2011