Asymmetric transformation (deracemisation) of an atropisomeric bisheterocyclic amine

Article abstract of DOI:10.1039/a905018c

Heating (±)-6,8,14,17-tetrahydro-7H-[1,2,5]triazepino[3,2b:7,1-b']diquinazoline-14,17 -dione 1 with (+)-CSA gives (-)-1 with > 90% ee via a crystallisation-induced asymmetric transformation, the first non-racemic example of a C₂ symmetric bishe

Full text of DOI:10.1039/a905018c

Asymmetric transformation (deracemisation) of an atropisomeric
bisheterocyclic amine†
Michael P. Coogan,*a David E. Hibbsb‡ and Emma Smarta
a Department of Chemistry, University of Durham, Science Laboratories, South Road, Durham, UK DH1 3LE.
E-mail: m.p.coogan@durham.ac.uk
b Department of Chemistry, University of Wales, Cardiff, Park Place, Cardiff, UK CF1 3TB
Received (in Liverpool, UK) 21st June 1999, Accepted 11th August 1999
Heating ( )-6,8,14,17-tetrahydro-7H-[1,2,5]triazepino[3,2-
b+7,1-bA]diquinazoline-14,17-dione 1 with (+)-CSA gives
(2)-1 with > 90% ee via a crystallisation-induced asym-
metric transformation, the first non-racemic example of a C2
symmetric bisheterocycle which is atropisomeric by virtue of
retarded rotation around an N–N bond.
Treatment of a solution of 2 in THF with aqueous NH3 gave
6,8,14,17-tetrahydro-7H-[1,2,5]triazepino[3,2-b+7,1-bA]diqui-
nazoline-14,17-dione 1 [m/z 331 (100%)]§ in 74% yield; none
of the anticipated diamine 5 was detected. In amine 1 the
diastereotopic protons in each methylene unit [d 4.00, 3.93 (J
13.2 Hz)] confirmed that rotation around the N–N bond was
slow on the NMR time scale; again in VT NMR experiments
coalescence was not observed at 135 °C, indicating a minimum
The efficacy of substituted 1,1A-binaphthyls as ligands for
metals in catalytic asymmetric reactions has prompted interest
in other aspects of atropisomerism: recently hindered aryl
amides2a and novel binaphthyl ligands and analogues2b have
been the focus of interest. One system which displays hindered
rotation due largely to an electronic rather than steric barrier is
the N–N bond of peracylhydrazines.3 Atkinson et al.4 have
shown that the N–N bond in 3-diacylaminoquinazolin-4-ones is
a chiral axis when the acyl groups are different and, by analogy,
N–N linked biquinazolinones would be expected to display
atropisomerism (axial chirality). However, although a range of
3,3A-biquinazoline-4,4A-diones have been prepared,5 the issue of
chirality was not addressed. A stereostable chiral axis in such
compounds could lead to the design of novel ligands for metals
in catalytic asymmetric reactions. Having verified the presence
of chirality it was hoped that, rather than resolution of
enantiomers, a deracemisation reaction (asymmetric trans-
formation, AT)1 could be developed. Vedejs and co-workers
have obtained diastereomerically pure materials by AT at the
stereogenic but stereolabile heteroatoms of chiral organobor-
on6a and organophosphorus6b compounds. Notably, Vedejs and
Donde6b achieved crystallisation-induced AT at a stereogenic
phosphorus utilising an alkoxycarbonylphosphine to both
introduce a chiral auxiliary, and lower the barrier to inversion,
allowing racemisation at around the melting point (Scheme 1).
In the work described here a high barrier to rotation (racemisa-
tion) coexists with high crystallinity, and the electronic nature
of the barrier is exploited, allowing it to be moderated via
protonation.
1
barrier to rotation of 85 kJ mol21. No separation of H NMR
signals was observed in the presence of acetylmandelic acid7 or
tartaric acid. However, in the presence of 2 equiv. of (+)-CSA,
two sets of AB pattens were observed [d 4.89, 4.75, 4.70 and
4.68 (J 14.4 Hz for all signals)] equating to the two
diastereoisomeric salts. To investigate possible AT, a sample of
1 and 2 equiv. of (+)-CSA was suspended in toluene and heated
1
at reflux for 16 h, then evaporated and analysed by H NMR.
Astonishingly, only one set of diastereotopic protons was
observed (d 4.89, 4.70), along with decomposition products.
Base washing gave 1 identical with authentic material which on
treatment with CSA displayed only one set of methylene
signals, implying that the sample was essentially a single
enantiomer. To ascertain whether this phenomenon was due to
selective decomposition or AT, optimisation experiments were
undertaken in various solvents at a range of temperatures and
concentrations. Optimum conditions were heating 1 at reflux
with benzene containing (+)-CSA (2 equiv.) for 40 h, which
gave an apparent diastereoisomer ratio of > 10+1 (judged by
NMR) with negligible decomposition. The recovered 1 after
base washing accounted for 82% of the starting material, which
is incompatible with a selective decomposition of one enantio-
mer, and had specific rotation [a]D 2643 (c 0.1, CH2Cl2)
(Scheme 3), raised by a single recrystallisation to 2688 (c
0.011, CH2Cl2), at which point (+)-CSA as an NMR shift
2,2A-Bis(bromomethyl)-3,3A-biquinazoline-4,4A-dione 2 was
prepared in two steps from 35 in 58% yield via the unstable bis-
bromoacetate 4 (Scheme 2). The bromomethylene protons of 2
are observed as an AB system [d 4.43, 4.32 (J 12.0 Hz)]
indicating chirality (slow rotation around the N–N bond on the
NMR time scale). VT NMR showed line broadening above
60 °C, however even at 135 °C free rotation was not observed
around the N–N bond (non-coalescence) which indicates a
minimum barrier to rotation in the order of 85 kJ mol21
.
Scheme 1 Conditions: i, crystalline state, 50 °C, 22 h.
Scheme 2 Reagents and conditions: i, H2NNH2·H2O (0.5 equiv.), EtOH,
reflux, 16 h, 84%; ii, BrCH2COBr, Et3N, DMF, 0 °C; iii, TsOH (10 mol%),
PhMe, reflux (Dean and Stark), 58% for 2 steps; iv, 35% aq. NH3 (100
equiv.), THF, 74%.
† See ref. 1.
‡ Corresponding author for crystallography.
Chem. Commun., 1999, 1991–1992
This journal is © The Royal Society of Chemistry 1999
1991

We thank Dr Alan Kenwright, Ian McKeag and Catherine
Heffernan for VT NMR studies and Dr P. G. Steel for helpful
discussions and the generous donation of materials.
Notes and references
§ All new compounds were fully characterised. Selected data for 1: mp
218.2–220.6 °C (decomp.) (EtOH); dH(300 MHz, CDCl3) 8.38 (2H, dd, J
8.1, 1.2, 2 3 H6), 7.91 (2H, ddd, J 8.1, 7.2, 1.5, 2 3 H7), 7.81 (2H, dd, J
7.5, 1.5, 2 3 H9), 7.63 (2H, ddd, J 8.1, 7.5, 1.2, 2 3 H8), 4.08, 4.02 (4H,
2 3 d, J 13.5, 2,2A-CH2), 2.01 (br, NH). For 2: mp 228.6 °C (EtOAc);
dH(300 MHz, CDCl3) 8.31 (2H, ddd, J 8.1, 1.5, 0.6, 2 3 H6), 7.93 (4H, m,
2 3 H7 + H8), 7.60 (2H, ddd, J 6.9, 1.5, 1.2, 2 3 H9), 4.43 (2H, d, J 12.0,
2 3 H2A), 4.32 (H, d, J 12.0, 2 3 H2B).
Scheme 3 Reagents and conditions: i, (+)-CSA (2 equiv.), PhH (3.8 ml
mol21), reflux, 40 h; ii, HBr, EtOH–H2O or CoSO4, EtOH–H2O.
¯
¶ Crystal data for 6: C18H15N5O3, M = 349.35, triclinic, Space group P1,
m = 0.101 mm21, R1 = 0.2872, wR2 = 0.1412, a = 4.684(4), b =
13.470(5), c
88.302(10)°, V = 816.6(8) Å3, T = 293(2) K, Z = 2, reflections collected/
unique 2307/2201 [R(int) 0.0696]. CCDC 182/1375. See
= 13.786(6) Å, a = 73.144(10), b = 78.90(2), g =
reagent indicated that the material was enantiopure ( > 98%
ee).
=
The use of a paucity of a ‘poor’ solvent is essential; a solution
of 1 with 2 equiv. of CSA (1,2-dichloroethane) at 83 °C (vs. 80
°C for benzene) for 40 h gave only the racemic material and, on
prolonged heating (72 h), decomposition. Two samples of 1 (0.1
mmol) plus 2 equiv. of CSA heated at reflux for 16 h in 10 ml
(entirely dissolved) and 1 ml (two phase) of benzene gave
diastereoisomer ratios 1+1 and 3+1, respectively. The use of
higher temperatures was detrimental: 2 h at 135 °C led to
complete decomposition. These data are indicative of a
crystallisation induced AT [see ref. 1(c) for a full explanation of
the principles involved] rather than the formation of a
thermodynamically favourable salt. The racemisation process
would appear to be acid catalysed: a solution of (2)-1 and CSA
showed appreciable racemisation (diastereoisomer ratio 4+1)
after 1 week at room temperature, whereas a solution of the free
base showed minimal racemisation after 14 days, and the rate of
racemisation of 1 even at 110 °C was slow (t1/2 > 24 h at 110
°C). The crystalline material retained its original ee during the
same period. It is unlikely that protonation of the secondary
amine will assist rotation around the N–N bond, thus it is
proposed that protonation of the quinazolinone 1-(imine)
nitrogens (hence the requirement for CSA, a strong acid) and the
resultant amidinium ion resonance to N-3 reduces the barrier to
rotation around the N–N bond. Attempts to grow crystals
containing heavy atoms suitable for the determination of
absolute configuration with HBr or aqueous CoSO4 led not to
the respective salt/complex but to the rearranged hydrolysis
product 6 (Scheme 3) as determined by single crystal diffrac-
tion.¶ This facile hydrolysis is atypical of biquinazolinones.
http://www.rsc.org/suppdata/cc/1999/1991/ for crystallographic data in .cif
format.
1 (a) Asymmetric Transformation refers to ‘an asymmetric transformation
of the second knd’, R. Kuhn, Chem. Ber., 1932, 65, 49; (b) For a recent
review of controlled racemisation and asymmetric transformation, see
E. J. Ebbers, G. J. A. Ariaans, J. P. M. Houbiers, A. Bruggink and B.
Zwannenburg, Tetrahedron, 1997, 53, 9417; (c) For a recent paper on
asymmetric transformation, the introduction of which sets out the
principles and correct terminology, see N. A. Hassan, E. Bayer and J. C.
Jochims, J. Chem. Soc., Perkin Trans. 1, 1998, 3747; (d) For rare cases
of the application of asymmetric transformation to chiral allenes, see M.
Node, K. Nishide, J. Fujiwara and S. Ichihashi, Chem. Commun., 1998,
2363; Y. Naruse, H. Watanabe, Y. Ishiyama and T. Yoshida, J. Org.
Chem., 1997, 62, 3862.
2 (a) See J. Clayden, Angew. Chem., Int. Ed. Engl., 1997, 36, 949 and
references therein; (b) See e.g. S. Vyskocil, M. Smrcina and P. Kocovsky,
Tetrahedron Lett., 1998, 39, 9289; G. Chelucci, A. Bacchi, D. Fabbri, A.
Saba and F. Ulgheri, Tetrahedron Lett., 1998, 40, 553.
3 S. M. Verma and R. Prasad, J. Org. Chem., 1973, 38, 1004.
4 R. S. Atkinson, E. Barker, P. J. Edwards and G. A. Thompson, J. Chem.
Soc., Perkin Trans. 1, 1996, 1047; R. S. Atkinson, E. Barker, C. J. Price
and D. R. Russell, J. Chem. Soc., Chem. Commun., 1994, 1159.
5 P. S. N. Reddy and A. K. Bhavani, Indian J. Chem., Sect. B, 1992, 31,
740.
6 (a) E. Vedejs, S. C. Fields, R. Hayashi, S. R. Hitchcock, D. R. Powell and
M. R. Schrimpf, J. Am. Chem. Soc., 1999, 121, 2460; (b) E. Vedejs and
Y. Donde, J. Am. Chem. Soc., 1997, 119, 9293.
7 S. C. Benson, P. Cai, M. Colon, M. A. Haiza, M. Tokles and J. K. Snyder,
J. Org. Chem., 1988, 53, 5335.
Communication 9/05018C
1992
Chem. Commun., 1999, 1991–1992