Efficient chromatographic resolution of a configurationally fragile atropisomeric diphenol via its N-(α)-Boc-tryptophan diesters

Article abstract of DOI:10.1016/j.tetasy.2005.03.011

Racemic 4,5-bis-(2-hydroxy-phenyl)-benzo[f]isoindole-1,3-dione 2, an atropisomeric diphenol prone to thermal isomerisation, has been efficiently resolved by conversion into its N-(α)-Boc-tryptophan diesters. Easy chromatographic separation of the diastereomeric pair of diesters, followed by ester cleavage under mild conditions gave each enantiomer in good yield and high enantiomeric purity. X-ray diffraction on a single crystal of one of the diastereomeric diesters allowed attribution of the absolute configuration of the stereogenic axes.

Full text of DOI:10.1016/j.tetasy.2005.03.011

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 1681–1684
Efficient chromatographic resolution of a configurationally fragile
atropisomeric diphenol via its N-(a)-Boc-tryptophan diesters
a
Malek Nechab, Bhavesh M. Panchal, Christian Philouze,
b
a
a
Cathy Einhorn and Jacques Einhorn
a,
*
a
Laboratoire d’ e´ tudes dynamiques et structurales de la s e´ lectivit e´ /UMR 5616, Universit e´ Joseph Fourier, BP 53, 38041,
Grenoble Cedex 9, France
b
Sun Pharma Advanced Research Center, Tandalja, Vadodara 390020, Gujarat, India
Received 2 February 2005; accepted 10 March 2005
Available online 14 April 2005
Abstract—Racemic 4,5-bis-(2-hydroxy-phenyl)-benzo[f]isoindole-1,3-dione 2, an atropisomeric diphenol prone to thermal isomeri-
sation, has been efficiently resolved by conversion into its N-(a)-Boc-tryptophan diesters. Easy chromatographic separation of the
diastereomeric pair of diesters, followed by ester cleavage under mild conditions gave each enantiomer in good yield and high enan-
tiomeric purity. X-ray diffraction on a single crystal of one of the diastereomeric diesters allowed attribution of the absolute con-
figuration of the stereogenic axes.
Ó 2005 Elsevier Ltd. All rights reserved.
1.Introduction
R
O
OH
O
N-Hydroxyphthalimide (NHPI) and several of its struc-
tural analogues are valuable catalysts for the aerobic
oxidation of organic compounds under mild condi-
O
N OH
N OH
N H
1
O
O
R
O
OH
tions. Room temperature oxidations have even been
performed, using NHPI or NHPI-analogues in associa-
NHPI
1
b,c
1
2
tion with acetaldehyde or CuCl.
We have previously
reported the first synthesis of axially chiral N-hydroxy-
imides and their use as catalysts for some representative
asymmetric oxidations. They displayed modest but nev-
chiral HPLC column, although effective on an analytical
scale, is impractical on a preparative scale, owing to its
low solubility in most of the usual solvents. On the other
hand, diphenol 2 is configurationally fragile: though a
solution of pure (±)-2 in toluene shows only negligible
isomerisation after 24 h at room temperature, its heating
at reflux for a few hours resulted in an atropisomerisa-
2
ertheless encouraging enantioselectivities. The discov-
ery of more efficient chiral analogues of NHPI is a
goal of current interest, due to various potential applica-
tions in asymmetric catalysis. Promising candidates are
C -symmetrical compounds of type 1, R being a variable
2
substituent. Several catalysts of type 1 have already been
3
tion of the stereogenic axes and equilibration to a ca.
4
prepared in their racemic form starting from racemic 2,
which has been obtained previously with a high diaste-
reomeric purity via thermal isomerisation in the solid
1:1 mixture of cis:(±)-2. This thermal lability renders
the resolution of (±)-2 particularly challenging. Indeed,
many classical methods developed for the resolution of
BINOL or other atropisomeric diphenols via covalent
linking to a chiral auxiliary, separation of the diastereo-
meric pair and final cleavage of the auxiliary were inef-
fective when applied to (±)-2. Herein we report an
efficient resolution of (±)-2 via its N-(a)-Boc-L-trypto-
phan diesters, and the determination of the absolute ste-
reochemistry of the stereogenic axes by X-ray diffraction
on a single crystal of one of the diastereomeric diesters.
4
state.
The availability of the pure enantiomers of 2 is therefore
crucial for our project. Direct resolution of (±)-2 on a
*
Corresponding author. Tel.: +33 4 76 51 48 39; fax: +33 4 76 51 48
6; e-mail: jacques.einhorn@ujf-grenoble.fr
3
0
957-4166/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2005.03.011

1682
M. Nechab et al. / Tetrahedron: Asymmetry 16 (2005) 1681–1684
2
.Results and discussion
2.2 equiv of DCC and 10 mol % of DMAP in anhydrous
dichloromethane was complete after 2 h at room tem-
perature (TLC monitoring). Column chromatography
of the crude reaction mixture was performed on 200 g
of silica gel, using mixtures of dichloromethane/ether
(9:1 to 1:1) as eluents. The first eluted diester was ob-
tained diastereomerically pure in 94% yield, as a white
microcrystalline solid. Numerous attempts to obtain a
monocrystal suitable for X-ray diffraction failed. The
second eluted diester was isolated diastereomerically
pure in 91% yield as a yellowish foam, which crystallised
after 24 h stirring at room temperature in a 4:1 mixture
of hexane/ethyl acetate. A monocrystal suitable for X-
ray diffraction was obtained by slow diffusion of hexane
into an ethyl acetate solution of this compound. Figure
We have previously reported a convenient and efficient
0
chromatographic resolution of BINOL and 6,6 -di-
5
bromo-BINOL. A preliminary screening revealed that
esterification of (±)-BINOL with various amino acid
derivatives gave diastereomeric esters, whose chromato-
graphic separation factors were highly dependent on the
amino acid pattern. Unusually high separation factors
were observed using N-(a)-Boc-L-tryptophan. Both
selective mono- and diesterifications have been per-
formed while the diastereomeric pairs of mono- or di-
esters were efficiently separated by simple column
chromatography techniques. We have attempted similar
mono- and diesterifications of (±)-2 with several amino
acid derivatives, using the standard DCC/DMAP ester-
ification method. In contrast with the case of BINOL
7
1 shows an ORTEP plot obtained from the X-ray data.
0
and that of 6,6 -dibromo-BINOL, clean conversions of
(
±)-2 into monoesters were not achieved and intractable
mixtures of starting 2, mono- and diesters were always
obtained. Fortunately, clean diesterifications were more
easily performed, without any disturbance caused by the
free imide functionality. Again, the largest TLC separa-
6
tion factors were observed in the tryptophan series:
TLC of the crude mixture of diesters 3a and 3b (Scheme
1
) gave two spots with R = 0.56 and R = 0.46
f1 f2
(
ethyl acetate as the eluent. Elution with a 4:1 mixture
a = R /R = 1.2) with a 1:1 mixture of hexane and
f1 f2
of dichloromethane and ethyl acetate gave R = 0.41
f1
and R = 0.21 (a = 2), while a 4:1 mixture of dichloro-
f2
methane and ether gave R = 0.30 and R = 0.13
(
f1
f2
a = 2.3).
We then undertook the synthesis and separation of 3a
and 3b on a preparative scale. Diesterification of 5 mmol
of (±)-2, using 2.2 equiv of N-(a)-Boc-L-tryptophan,
Figure 1. ORTEP plot of the second eluted diastereomer 3b.
O
O
NHBoc
NH
NH
OH
O
O
NHBoc
CO2Me
ii
N H
NH + 2
N
H
O
O
NHBoc
O
OH
OH
O
O
3a (94%±
()±)2 (99%±
i
N H
O
OH
O
O
NHBoc
CO2Me
(
± ±)2
NH
NH
OH
O
O
NHBoc
ii
N H
NHBoc
NH + 2
N
H
O
O
O
OH
O
3b (91%±
(+±)2 (98%±
Scheme 1. Resolution of (±)-2 via its N-(a)-Boc-L-tryptophan diesters. Reagents and conditions: (i) (a) DCC, DMAP, N-(a)-Boc-L-tryptophan,
CH Cl , rt; (b) column chromatography. (ii) (a) LiOHÆH O, MeOH, rt; (b) TFA.
2
2
2

M. Nechab et al. / Tetrahedron: Asymmetry 16 (2005) 1681–1684
1683
An (aS,aS) absolute stereochemistry of the stereogenic
axes can be deduced from the (S) absolute stereochemis-
try of the stereogenic centres of the amino acid moieties.
Consequently, the stereochemistry of the second eluted
diester corresponds to formula 3b (Scheme 1). The ste-
reochemistry of the first eluted diester therefore corre-
sponds to formula 3a. It is worth mentioning that
similar behaviours have been observed for mono- and
at d = 39.52. Mass spectra were recorded on a Thermo
Quest Finnigan spectrometer. Elemental analyses were
performed at the Service Central dꢀAnalyses du CNRS.
Optical rotations were measured on a Perkin–Elmer 341
polarimeter. IR spectra were recorded as KBr pellets on
a Magna-IR 550 spectrometer.
4.2. Synthesis and separation of diesters 3a and 3b
0
diesters of BINOL and of 6,6 -dibromo-BINOL with
N-(a)-Boc-L-tryptophan: in each case, the stereogenic
axis of the faster running diastereomer had an (aR)-
Racemic diphenol 2 (1.9 g, 5 mmol), DCC (2.28 g,
11 mmol), DMAP (0.062 g, 0.5 mmol), and N-(a)-Boc-
L-tryptophan (3.348 g, 11 mmol) were dissolved in anhy-
drous dichloromethane (150 mL) and the mixture stirred
at room temperature for 2 h. After filtration of dicyclo-
hexylurea, the solvent was removed at a reduced pressure.
The residue was chromatographed over silica gel (200 g)
using a 9:1 mixture of dichloromethane/ether for the elu-
tion. The proportion of ether was progressively raised to
50%. Diester 3a was eluted first (2.23 g, 94% yield), fol-
lowed by diester 3b (2.16 g, 91% yield).
8
configuration.
As diphenol 2 is prone to thermal atropisomerisation,
the ultimate success of the resolution rested on the clean
removal of the chiral auxiliary under mild conditions.
Classical saponification using aqueous NaOH or KOH
gave somewhat erratic results. Gratifyingly, almost
instantaneous cleavage occurred at room temperature,
by using stoichiometric amounts of LiOH in methanol.
Acidification with trifluoroacetic acid, followed by col-
9
umn chromatography, gave resolved 2, along with
recovered enantiomerically pure N-(a)-Boc-L-trypto-
phan as its methyl ester (Scheme 1). Thus, 3a gave
4.2.1. (aR,aR)-4,9-Bis(phenyl-2-yl)-benzo[f]isoindole-1,3-
dione [(2S)-2-tert-butoxycarbonylamino-3-(1H-indol-3-
yl)]dipropionate 3a. White crystalline 20 solid, mp
5
1
0
9
3
9% yield of pure (aR,aR)-(À)-2 (ee >99%). Similarly,
195 °C (from hexane/ethyl acetate). ½aŠ ¼ À4:1 (c
D
1
0
1
b gave 98% yield of pure (aS,aS)-(+)-2 (ee >99%).
1.01, THF). H NMR (DMSO-d ) d (ppm) 1.13 and
6
1.23 (2br s in a 0.2:1 ratio, 18H), 1.76–2.14 (m, 4H),
3.79–3.87 (m, 2H), 6.65–7.09 (m, 12H), 7.16–7.36 (m,
6H), 7.53–7.60 (m, 4H), 7.65–7.70 (m, 2H), 10.65 (br s,
3.Conclusion
1
3
2
H), 11.48 (br s, 1H). C NMR (DMSO-d ) d (ppm)
6
Racemic, configurationally fragile diphenol 2 has been
converted into its diastereomeric N-(a)-Boc-L-trypto-
phan diesters 3a and 3b, which have been easily sepa-
rated by column chromatography. X-ray diffraction on
a monocrystal of the second eluted diester allowed the
attribution of the absolute configuration of the stereo-
genic axes. Finally, mild cleavage of the chiral auxiliaries
gave each enantiomer of 2 with an ee >99%. The overall
yield of the resolution was 93% for (aR,aR)-(À)-2, and
25.07, 28.03, 54.18, 78.28, 109.32, 111.50, 117.20,
118.48, 121.02, 122.50, 123.59,125.52, 125.62, 126.45,
127.56, 127.67, 129.05, 129.05, 129.74, 131.02, 133.54,
134.52, 136.00, 148.19, 155.32, 167.18, 170.65. IR (KBr
À1
pellet) m (cm ) 3404, 3061, 2980, 2931, 1771, 1722,
À
1494, 1471, 1167, 743. MS (CI , NH +isobutane) m/z
3
(%) 953(84), 853(100), 753(31), 667(10), 567(3). Anal.
Calcd for C H N O : C, 70.50; H, 5.39; N, 7.34.
Found: C, 69.85; H, 5.38; N, 7.47.
5
6
51
5
10
8
9% for (aS,aS)-(+)-2, when carried out on a 5 mmol
scale. The transformation of enantiomerically pure 2
into various enantiomerically pure catalysts of type 1
is now under investigation.
4.2.2. (aS,aS)-4,9-Bis(phenyl-2-yl)-benzo[f]isoindole-1,3-
dione [(2S)-2-tert-butoxycarbonylamino-3-(1H-indol-3-
yl)]dipropionate, 3b. Yellowish crystals, mp 217 °C
1
1
2
0
(
from hexane/ethyl acetate). ½aŠ ¼ À10:2 (c 1.09,
D
1
THF). H NMR (DMSO-d ) d (ppm) 1.07 and 1.21
6
4.Experimental
(2br s in a 0.2:1 ratio, 18H), 1.92–2.50 (m, 4H), 4.01–
4.09 (m, 2H), 6.77–7.30 (m, 12H), 7.39–7.55 (m, 6H),
7.72–7.79 (m, 4H), 7.86–7.91 (m, 2H), 10.94 (br s, 2H),
4
.1. General
1
3
1
1.55 (br s, 1H). C NMR (DMSO-d ) d (ppm) 25.18,
6
Dichloromethane was distilled over calcium hydride be-
fore use. Methanol and other chemicals were of com-
mercial grade and used as received. Analytical TLC
was performed on Merck silica gel plates (60 F254). Col-
umn chromatography was carried out on silica gel 60
27.99, 54.43, 78.25, 109.28, 111.40, 117.34, 118.36,
120.92, 122.28, 123.59, 125.53, 125.63, 126.40, 127.65,
127.77, 129.02,129.76, 131.28, 133.60, 134.57, 135.94,
148.31, 155.29, 167.08, 170.38. IR (KBr pellet) m
À1
(cm ) 3404, 3061, 2980, 2931, 1771, 1722, 1494, 1167,
À
(
70–230 mesh). Chiral HPLC was performed on a Chir-
743. MS (CI , NH +isobutane) m/z (%) 953(33),
3
alpak AS column. Melting points were measured on a
Buchi Melting Point B-545 apparatus and are uncor-
rected. NMR spectra were recorded on a Brucker
853(100), 753(79), 667(17), 567(9). Anal. Calcd for
C H N O : C, 70.50; H, 5.39; N, 7.34. Found: C,
70.65; H, 5.46; N, 7.50.
5
6
51
5
10
1
AC300 spectrometer. H NMR were recorded at
3
00 MHz in DMSO-d with chemical shifts being re-
6
4.2.3.
(aR,aR)-4,9-Bis(2-hydroxyphenyl)-benzo[f]iso-
ported as d (ppm), calibrated to the residual solvent sig-
indole-1,3-dione, (À)-2. To a solution of diester 3a
1
3
nal at d = 2.50. C NMR spectra were recorded at
5 MHz in DMSO-d , calibrated to the solvent signal
(2.22 g, 2.3 mmol) in methanol (40 mL) was added a
1 M solution of LiOHÆH O in methanol (4.6 mL,
7
6
2

1684
M. Nechab et al. / Tetrahedron: Asymmetry 16 (2005) 1681–1684
4
.6 mmol). The yellow solution was stirred for 5 min at
4. Einhorn, C.; Durif, A.; Averbuch, M.-T.; Einhorn, J.
Angew. Chem., Int. Ed. 2001, 40, 1926–1929.
room temperature. After cooling to 0 °C, trifluoroacetic
acid (0.36 mL, 4.6 mmol) was added, giving a nearly col-
ourless solution. The solvent was next removed at re-
duced pressure. The residue was chromatographed
over silica gel (100 g) using mixtures of dichlorometh-
5
. Panchal, B. M.; Einhorn, C.; Einhorn, J. Tetrahedron Lett.
2
002, 43, 9245–9248.
6. Diesterification has been performed on an analytical scale
10 lmol of (±)-2], with the following commercially
9
[
available N-protected-L-amino acids: N-Boc-proline, N-
Boc-valine, N-Boc-phenylglycine, N-Boc-phenylalanine,
N-(a)-Boc-tryptophan, N-(a)-Cbz-tryptophan, and N-(a)-
Fmoc-tryptophan. Clearly separated TLC spots were
observed only in the phenylalanine, or much better, in
the tryptophan series. N-(a)-Cbz and -Fmoc derivatives
offer no advantages over -Boc derivatives.
ane/ether (95:5 to 4:1) as eluent. N-(a)-Boc-L-trypto-
phan methyl ester was eluted first (98% recovery),
5
8
followed by (À)-2 (0.874 g, 99% yield): slightly yellow
20
amorphous solid. ½aŠ ¼ À13:25 (c 0.93, THF);
D
ee >99% (Chiral HPLC on a Chiralpack AS column,
À1
elution hexane:i-PrOH, 6:4, 1 mL min , t = 5.03 min).
R
1
7
. X-ray data were collected with a Bruker Nonius kappa
CCD diffractometer using graphite-monochromated
Mo-Ka radiation. The structure was solved by direct
methods (SIR 92: Altomare A., Cascarano G., Giaco-
vazzo C., Guagliardi A. J. Appl. Cryst. 1993, 26, 343–350)
and refined by full-matrix least-squares techniques against
F. Ortep [Johnson CK. (1976). ORTEPII. Report
ORNL-5138. Oak Ridge National Laboratory, Tennessee,
USA] was used for structure representations. Crystal
H NMR (DMSO-d ) d (ppm) 6.93–7.03 (m, 4H), 7.16–
6
7
.20 (m, 2H), 7.30–7.37 (m, 2H), 7.60–7.69 (m, 4H), 9.35
1
3
(
br s, 2H), 11.18 (br s, 1H). C (DMSO-d ) d (ppm)
6
1
1
15.46, 118.71, 121.79, 125.39, 127.81, 128.36, 129.37,
31.02, 134.88, 135.61, 154.96, 167.69. IR (KBr pellet)
À1
m (cm ) 3400, 3073, 2975, 2893, 1752, 1712, 1617,
1
1
m/z (%) 382(54), 399(100). Anal. (determined on the
ethyl acetate solvate) Calcd for C H N O : C,
10
601, 1591, 1506, 1451, 1379, 1345, 1288, 1221, 1121,
093, 1021, 837, 761 cm . MS (CI , NH +isobutane)
À1
+
3
data: C56
P322
51 5
H N O10, M = 954.03, trigonal, space group
1
0
˚
1
, a = 14.476(1), c = 21.310(2) A, c = 120.0°, V =
5
2
38
2
˚
3
À1
,
3
867.5(4) A , T = 293(2) K, l (Mo-Ka) = 0.085 mm
7
3
3.40; H, 4.50; N, 3.29. Found: C, 72.49; H, 4.56; N,
.46.
À3
D
c
= 1.229 g cm , 6614 reflections measured, 3549 were
independent of symmetry and 2706 were observed
I > 1.20r(I)], R = 0.0734, wR = 0.1083, 318 parameters.
[
1
2
4
.2.4.
(aS,aS)-4,9-Bis(2-hydroxyphenyl)-benzo[f]isoin-
Crystallographic data (excluding structure factors) have
been deposited with the Cambridge Crystallographic Data
Center as supplementary publication no. CCDC 260846.
Copies of the data can be obtained free of charge on
application to CCDC, 12 Union Road, Cambridge,
CB21EZ. UK [fax: +44(0) 1223-336033; e-mail:
deposit@ccdc.cam.ac.uk].
dole-1,3-dione, (+)-2. The same procedure as above
was applied to diester 3b (1.98 g, 2.07 mmol) and gave
diphenol (+)-2 (0.770 g, 98% yield): slightly yellow
20
amorphous solid. ½aŠ ¼ þ13:9 (c 0.94, THF); ee >99%
D
Chiral HPLC on a Chiralpack AS column, elution
(
hexane:i-PrOH, 6:4, 1 mL min , t = 7.89 min). Anal.
À1
R
1
0
8. For similar observations about the resolution of BINOL
derivatives via diastereomeric bis(camphorsulfonates), see:
Chow, H.-F.; Wan, C.-W.; Ng, M.-K. J. Org. Chem. 1996,
(determined on the ethyl acetate solvate) Calcd for
C H N O : C, 73.40; H, 4.50; N, 3.29. Found: C,
7
5
2
38
2
10
2.76; H, 4.56; N, 3.49.
6
1, 8712–8714.
9
. As diphenol 2 is prone to thermal isomerisation, special
care should be taken to avoid excessive heating during its
handling, especially during solvent removal.
Acknowledgements
1
0. Crystalline solvates formed when amorphous (À)- or (+)-2
was stirred at room temperature for 24 h in a 4:1 mixture
of hexane/ethyl acetate. These crystals contained half a
molecule of ethyl acetate per molecule of (À)- or (+)-2.
1. The thermal sensitivity of 2 does not preclude application
to chiral catalysis: various NHPI type oxidations can be
Financial supports from CNRS and Universit e´ Joseph
Fourier are gratefully acknowledged.
1
References
1
b,c,2
conducted near room temperature,
at which no
1
. For a review, see: (a) Ishii, Y.; Sakaguchi, S.; Iwahama, T.
Adv. Synth. Catal. 2001, 343, 393–427; (b) Einhorn, C.;
Einhorn, J.; Marcadal, C.; Pierre, J.-L. Chem. Commun.
racemisation of 2 occurs. Moreover, several catalysts of
type 1 bearing bulky R substituents have already been
prepared. Preliminary results indicate that they are much
more thermally stable than 2. This preliminary study has
been conducted with racemic catalysts of type 1, as the
absence of cis/trans isomerisation in the racemic series will
preclude racemisation (via the cis form) under the same
conditions in the optically active series.
1997, 447–448; (c) Nechab, M.; Einhorn, C.; Einhorn, J.
Chem. Commun. 2004, 1500–1501.
. Einhorn, C.; Einhorn, J.; Marcadal-Abbadi, C.; Pierre,
J.-L. J. Org. Chem. 1999, 64, 4542–4546.
. Unpublished results.
2
3