Synthesis and properties of a chiral bis-tetrahydroisoquinoline proton sponge

Article abstract of DOI:10.1039/b110499n

A new chiral proton sponge has been prepared, and the reasons for its unusually high basicity elucidated by a quantum chemical study.

Full text of DOI:10.1039/b110499n

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis and properties of a chiral bis-tetrahydroisoquinoline
proton sponge†
Mark C. Elliott,* Eve Williams and Siân T. Howard
2
Department of Chemistry, Cardiff University, PO Box 912, Cardiff, UK CF10 3TB.
E-mail: elliottmc@cardiff.ac.uk; Fax: ϩ44 (0)29 20874030; Tel: ϩ44 (0)29 20874686
Received (in Cambridge, UK) 15th November 2001, Accepted 29th November 2001
First published as an Advance Article on the web 17th December 2001
A new chiral proton sponge has been prepared, and the
reasons for its unusually high basicity elucidated by a
quantum chemical study.
mixture of diastereoisomers of 2, from which the dl isomer was
easily separated and characterized as the double hydrobromide
salt 7. Liberation of the free base from 7 and treatment with
excess methyl iodide led to the formation of the double hydro-
iodide salt 8, from which the free base 3 was readily liberated
and purified by recrystallisation from ethanol–diethyl ether,
and characterised by single crystal X-ray diffraction (Fig. 1).6 ‡
Since Alder’s landmark synthesis of 1,8-bis(dimethylamino)-
naphthalene 1,1 the synthesis of novel organic superbases con-
tinues to generate considerable interest2,3 with recent work
focusing on chiral targets.4 Despite its potential in this area,
and more generally in asymmetric catalysis, the bis-tetrahydro-
isoquinoline ring system has seldom been reported. Herein we
report preliminary results in the preparation of derivatives of
compound 2, including the important observation that 3 is a
C2-symmetric chiral proton sponge, and compare the basicity
of compound 3 with that of compound 1 and of TMEDA (4).
The syntheses of both diastereoisomers of 2 have been previ-
ously reported.5 However, while this route provided the desired
compound, it features a hydrogenation step which we found to
be capricious, and so an alternative synthesis was developed as
shown in Scheme 1. Reaction of phenylethylamine 5 with ethyl
Fig. 1 Structure of 3 from single crystal X-ray diffraction data.
Treatment of 3 with additional methyl iodide cleanly provided
9, with no evidence for further methylation.
The basicity of 3 was determined by competitive NMR
studies with 1 in deuterated acetonitrile. A 1 : 1 mixture of 3
and the perchlorate salt of 1 showed clearly that the salt of 1
was approximately 50% deprotonated under these conditions,
leading to the conclusion that 3 and 1 have essentially identical
pKa values (18.2)7 in acetonitrile. While the free base and salt of
1 exchange protons relatively slowly at room temperature, the
corresponding exchange in 3 is rapid on the NMR timescale.
Two different products were isolated from the reaction of 3
with 1,2-dibromoethane. At room temperature little reaction
occurred, but the desired bridged double salt underwent
demethylation to give 10 as a single stereoisomer (unknown
stereochemistry) at nitrogen. At higher temperatures the only
product isolated was the single salt of 3 (Scheme 2). In both
cases the balance of material was recovered 3.
Scheme 1 Reagents and conditions: (i) EtOCHO; (ii) polyphosphoric
acid; (iii) Zn, Me3SiCl, BrCH2CH2Br then HBr, H2O; (iv) NaOH, H2O,
CH2Cl2; (v) CH3I; (vi) CH3I, THF.
formate, followed by cyclisation with polyphosphoric acid
gave dihydroisoquinoline 6 cleanly. Dimerisation of 6 using zinc
with 1,2-dibromoethane and chlorotrimethylsilane gave a 1 : 1
† Dedicated to Professor Harry Heaney on the occasion of his 70th
birthday.
Scheme 2 Reagents and conditions: (i) BrCH2CH2Br, 60 ЊC; (ii)
BrCH2CH2Br, 25 ЊC.
DOI: 10.1039/b110499n
J. Chem. Soc., Perkin Trans. 2, 2002, 201–203
This journal is © The Royal Society of Chemistry 2002
201

View Article Online
Table 1 Main structural features (HF/6-31G** and crystallographic)
Free base
Monoprotonated cation
rNؒؒN/Å
NCCN/Њ
NCCN/Њ
rNؒؒN/Å
rNؒؒH/Å
N–HؒؒN/Њ
1
2.791
2.679
3.781
3.672
2.955
3.760
–
–
2.684
2.585
–
1.706
1.772
–
155.1
133.1
–
3 syn
anti
11.8
179.6
143.2
46.1
175.5
15.1
–
X-ray
4 syn
46.2
174.9
2.749
3.776
2.099
3.945
119.7
73.0
anti
Table 2 B3LYP/6-31ϩG**//HF/6-31G** proton affinities (PA)a and
strain energies (SE) for the bases; plus hydrogen bond ϩ strain energies
(HB ϩ SE)ϩ and hydrogen bond energies (HB) and intramolecular pro-
ton transfer barriers (PT) for the monoprotonated cations, all in kJ
molϪ1 a
shows that compound 4 on the other hand is essentially strain-
less (the small negative SE is within the expected error for a
thermoneutral isodesmic reaction). Additionally the energies of
the syn and anti forms are within 1 kJ molϪ1
.
Considering now the syn protonated species, 1Hϩ and 3Hϩ
exhibit similar N–H ؒ ؒ ؒ N hydrogen bonds. Compound 3Hϩ
also has the very low barrier to intramolecular proton transfer
typical of diamine proton sponges, and appears to have no
stable anti protonated form. Serendipitously 4Hϩ does have a
stable anti conformer, so the ≈ 38 kJ molϪ1 syn/anti energy dif-
ference provides a good estimate of hydrogen bond strength.
The gas phase PA of 3 is slightly higher than that of 1.
An additive scheme8e describing its PA is
PA
SE
PT
(HB ؉ SE)؉
HB
1
3
1030.7 [1030 2]b
1037.0
998.6
26.9
0
2
15
–
Ϫ66.8
Ϫ17.8
–
–
Ϫ78.2
42.3
Ϫ2.4
–
≈ Ϫ38
≈ Ϫ37.6
–
4
11
975.9
a The PA’s, SE’s and (HB ϩ SE)ϩ values include scaled HF/3-21G
thermal energy corections.9 b Experimental gas phase PA measured by
Lau et al.10 Values for 1 taken from reference 8(e).
PA(3) = PA(11) ϩ [SE(3Hϩ) Ϫ SE(3)] Ϫ HB(3Hϩ)
Thus, the PA of 3 can be measured relative to a suitable,
structurally-related monoamine 11 and is the sum of strain
release on protonation and intramolecular hydrogen bond
stabilization of 3Hϩ. These latter two contributions we will
refer to as the ‘excess PA’. Taking HB(3Hϩ) ≈ HB(4Hϩ) this
gives 1037 = 976 ϩ 23 ϩ 38 kJ molϪ1, so the 61 kJ molϪ1 excess
PA comes mostly from hydrogen bonding. This is in line with
other sponges so far analysed in this way: for example, in
the case of 3, when the reference monoamine is chosen as
1-(dimethylamino)naphthalene8e we have 1031 = 937 ϩ 16 ϩ 78
(H-bond provides more than 80% of the 94 kJ molϪ1 excess
PA).
Quantum chemistry provides a powerful tool in proton
sponge design.8 Using the methodology described recently8e we
have explored the syn and anti conformers of 3 and 4 and their
mono-protonated cations, and have quantified the relative
contributions of hydrogen bonding and strain release to their
basicities. Structures were first optimised and frequency-tested
at the HF/3-21G level, followed by HF/6-31G** optimisation
and B3-LYP/6-31 ϩ G** single points at these geometries.
Strain energies (SE) for 1 and 3 have been calculated from iso-
desmic reactions (Schemes 3 and 4). Proton transfer energies
The main difference between 1Hϩ and 3Hϩ is clearly in the
absolute magnitude of the hydrogen bond, that in 3Hϩ being
only half the strength of 1Hϩ due to less favourable geometric
constraints. However, the fact that the (gas-phase) basicity of
3 is marginally higher than that of 1 appears to be due to the
higher intrinsic basicity of the aliphatic amine.
We believe that derivatives of compound 2 will have a range
of useful applications in asymmetric synthesis. These studies
are underway, and will be reported in due course.
Scheme 3 Isodesmic reaction used to obtain proton sponge strain.
Acknowledgements
The authors thank the EPSRC (use of the UK Computational
Chemistry Facility), and Cardiff University for a studentship
(EW). We are grateful to Dr M. P. Coogan and Mr R. Haigh for
the X-ray crystallographic study of compound 3. We also thank
Professor R. W. Alder and Dr G. C. Lloyd-Jones (both of the
University of Bristol) for helpful discussions.
Scheme 4 Isodesmic reaction used to obtain cation H-bond energy.
(PT) were also estimated from calculations on C2-symmetrised
cations.
Non-bonded distances rNؒؒN and N–C–C–N torsion angles
in both the free base and the cation are reported in Table 1,
along with hydrogen bond geometries (cation only). Gas phase
absolute PAs and other relevant electronic properties are
reported in Table 2.
References
‡ CCDC reference number 167915. See http://www.rsc.org/suppdata/
p2/b1/b110499n/ for crystallographic files in .cif or other electronic
format.
The calculations reveal that both 3 and 4 have stable syn (C2
symmetric) and anti (C1 or C2) conformers. In agreement with
the crystal structure of 3, the anti conformer is predicted to be
the most stable form, being some 17 kJ molϪ1 more stable than
syn in the gas phase. This is also consistent with the H NMR
data which suggest substantial conformational change (pre-
1 R. W. Alder, P. S. Bowman, W. R. S. Steele and D. R. Winterman,
J. Chem. Soc., Chem. Commun., 1968, 723.
2 T. Isobe, K. Fukuda and T. Ishikawa, J. Org. Chem., 2000, 65,
7770.
3 H. A. Staab, A. Kirsch, T. Barth, C. Krieger and F. A. Neugebauer,
Eur. J. Org. Chem., 2000, 8, 1617.
4 P. Hodgson, G. C. Lloyd-Jones, M. Murray, T. M. Peakman and
R. L. Woodward, Chem. Eur. J., 2000, 6, 4451.
5 A. T. Nielson, J. Org. Chem., 1970, 35, 2498.
1
sumably anti
syn) on protonation. Isodesmic reactions reveal
that 3 is quite strained (some 15 kJ molϪ1 more strained than 1)
presumably due to steric interactions between the benzene
rings. Calculation of the corresponding isodesmic reaction
6 Selected crystallographic data for 3 are as follows: C20H24N2 (Mr
292.41), monoclinic, P21/c, a = 11.039(2), b = 7.791(2), c = 19.001 Å.
202
J. Chem. Soc., Perkin Trans. 2, 2002, 201–203

View Article Online
8 (a) A. Gobbi and G. Frenking, J. Am. Chem. Soc., 1993, 115, 2362;
β = 104.31(3)Њ, Z = 4, V = 1583.5(6) Å3, Dc = 1.227 g cmϪ3, Mo-Kα
radiation, λ = 0.71073 Å, 2.21 < θ < 25.29Њ. 11232 reflections were
collected of which 2881 were unique. 1641 reflections having I >
2.0σ(I ) were used for refinement (272 parameters). Final R indices
(I > 2.0σ(I )) R1 = 0.0461 and wR2 = 0.0929. Crystallographic data
for 3 have been deposited at the CCDC, 12 Union Road, Cam-
bridge, UK CB2 1EZ and copies can be obtained on request, free
of charge, by quoting the publication citation and the deposition
number 167915.
(b) R. Notario and J. Elguero, J. Chem. Soc., Chem. Commun.,
1995, 154; (c) M. Peräkylä, J. Org. Chem., 1996, 61, 7420;
(d ) E. Fujiwara, K. Omoto and H. Fujimoto, J. Org. Chem., 1997,
62, 7234; (e) S. T. Howard, J. Am. Chem. Soc., 2000, 122, 8238;
( f ) Z. B. Maksic and B. Kovacevic, J. Org. Chem., 2000, 65, 3303.
9 R. F. Hout, B. A. Levi and W. J. Hehre, J. Comput. Chem., 1982, 3,
234.
10 Y. K. Lau, P. P. S. Saluja, P. Kebarle and R. W. Alder, J. Am. Chem.
Soc., 1978, 100, 7328.
7 A. F. Pozharskii, Russ. Chem. Rev., 1998, 67, 1.
J. Chem. Soc., Perkin Trans. 2, 2002, 201–203
203