A new chiral oxazoline derived from camphor and its conversion to (R)-2-methylalkanoic acids

Article abstract of DOI:10.1016/S0957-4166(00)00197-X

(1S,5R,7R)-(-)-10,10-Dimethyl-3-ethyl-4-oxa-2-azatricyclo[5.2.1.0^1,5]dec-2-ene 2 was prepared in 95% yield from (1S)-1-amino-2-exo-hydroxyapocamphane 1. The chiral oxazoline could be alkylated (LDA/THF/-78°C/RX, RX=ethyl, n-propyl, n-butyl iodides or benzyl bromide) to 3 in 95% yield and >95% diastereoselectivity, and the products hydrolysed to (R)-2-methylalkanoic acids 4 (43-47% yield, 93-98% e.e.). Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(00)00197-X

Tetrahedron: Asymmetry 11 (2000) 2249±2253
A new chiral oxazoline derived from camphor and its
conversion to (R)-2-methylalkanoic acids
Sosale Chandrasekhar* and Amina Kausar
Department of Organic Chemistry, Indian Institute of Science, Bangalore 560 012, India
Received 6 April 2000; accepted 16 May 2000
Abstract
(1S,5R,7R)-(^)-10,10-Dimethyl-3-ethyl-4-oxa-2-azatricyclo[5.2.1.01,5]dec-2-ene 2 was prepared in 95%
yield from (1S)-1-amino-2-exo-hydroxyapocamphane 1. The chiral oxazoline could be alkylated (LDA/
THF/^78ꢀC/RX, RX=ethyl, n-propyl, n-butyl iodides or benzyl bromide) to 3 in 95% yield and >95%
diastereoselectivity, and the products hydrolysed to (R)-2-methylalkanoic acids 4 (43±47% yield, 93±98%
e.e.). # 2000 Elsevier Science Ltd. All rights reserved.
Among the methodologies utilising a stoichiometric equivalent of a chiral auxiliary, the alkyl-
ation of chiral oxazolines has retained its importance in the armoury of asymmetric synthesis.
Introduced by Meyers1 3 in the mid 1970s, the methodology enables the preparation of chiral 2-
alkylalkanoic acids in excellent yields and enantiopurity. Although the elegant and pioneering
studies of Meyers and co-workers have de®ned an area of asymmetric synthesis which is as
exciting as it is utilitarian, scope for elaboration remains, if only by virtue of the importance of
the process: for clearly, a diverse pool of auxiliariesÐin both enantiomeric forms wherever
possibleÐonly enhances the value of an asymmetric synthetic methodology. We set out to design
a new chiral oxazoline based on a proven chiral frameworkÐthat of camphorÐand in the event
developed a methodology that o€ered considerable advantages over existing ones.
The Meyers methodology1 3 relies on the diastereoselective alkylation of a chiral 2-ethyl-
oxazoline anion, the oxazoline moiety thus serving as a chiral carboxyl protecting group and the
anion as a masked enolate. Initial work by the Meyers group on (4S,5S)-2-ethyl-4-methoxy-
methyl-5-phenyl-1,3-oxazoline indicated that a combination of three factors was essential for
good stereoselectivity (Scheme 1): the kinetically controlled formation of the (Z)-azaenolate, the
presence of the C5-phenyl group and the chelation of the lithium counterion by the C4-methoxy-
methyl group. The formation of the (Z)-azaenolateÐnon-equilibrating with the (E)-formÐallows
for the play of facial selectivity, which is then partly de®ned by the steric hindrance o€ered by the
* Corresponding author. E-mail: sosale@orgchem.iisc.ernet.in
0957-4166/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00197-X

2250
S. Chandrasekhar, A. Kausar / Tetrahedron: Asymmetry 11 (2000) 2249±2253
C5-phenyl group on the `top' face. Furthermore, the above chelation of the counterion was
deemed equally critical, as it allowed the alkyl halide to be guided to the `bottom' face via co-
ordination of the halide atom to the lithium ion: in the absence of this interaction, presumably,
steric hindrance by the C4-substituent would dominate with a resulting loss of facial selectivity.
Scheme 1.
Such are the broad contours of the Meyers methodology. A consideration of the mechanism
leads to a key conclusion: the fact that both the C5-phenyl and C4-methoxymethyl groups are
required indicates that the C5-phenyl group is not by itself ecient in blocking the `top' face of
the azaenolate (Scheme 1). Clearly, a larger group than phenyl at C5 may obviate the necessity for
the above chelation-guided process at the `bottom' face of the azaenolate. In principle, a geo-
metrically de®ned azaenolate in conjunction with ecient blocking of one face of this enolateÐ
whether at C4 or C5Ðshould provide high selectivity. The present studies address these
possibilities (indeed, more recent work has indicated that a t-butyl group at C5 can meet the
above criterion: naphthalene appended with a chiral oxazoline derived from t-leucinol added
alkyl lithium in excellent diastereoselectivity.4 However, we are not aware of the extension of
these studies to include simple alkylation processes, so their generality has apparently not been
de®ned).
The camphor framework has inspired the design of several chiral auxiliaries analogous to the
oxazoline, most notably the oxazolidinone5 but also the imidazolidinone6 and the oxazinone,7 not
to mention8 the sultam. The analogues were mostly derived from either 3-amino-2-hydroxy-
apocamphane or 1-amino-2-hydroxyapocamphane. In the present studies, an oxazoline was
prepared starting from the known5b,9,10 (1S,2R,4R)-1-amino-2-hydroxyapocamphane 1 (Scheme
2), obtained via NaBH4 reduction of (1S)-(+)-1-aminoapocamphane, derived from the commer-
cially available (1S)-(+)-camphorsulphonic acid via oxidation to (1S)-(+)-ketopinic acid11 and a
Curtius rearrangement9,10 of its azide. Treatment of amino alcohol 1 with ethyl iminopropionate
hydrochloride1 3 and a large excess of sodium bicarbonate in re¯uxing chloroform for 48 h,
a€orded (1S,5R,7R)-10,10-dimethyl-3-ethyl-4-oxa-2-azatricyclo[5.2.1.01,5]dec-2-ene 2 in 95%
yield (Scheme 2). The oxazoline 2 was characterised spectroscopically: the IR showed absorption
at 1630 cm^1, ascribed to CˆN; the 1H NMR showed a doublet of a doublet at ꢀ 4.12 ascribed to
OCH; the 13C NMR showed resonances at ꢀ 169.8 (C3), 88.5 (C5) and 81.0 (C1); and a satis-
factory high resolution mass was obtained.
Scheme 2. (i) EtOC(Et)ˆNH/NaHCO3; (ii) LDA/THF/^78ꢀC/RX; (iii) Ac2O/Á±KOH/MeOH/Á±H3O+

S. Chandrasekhar, A. Kausar / Tetrahedron: Asymmetry 11 (2000) 2249±2253
2251
The alkylations of the chiral oxazoline 2 were investigated via lithium diisopropylamide
mediated deprotonation in tetrahydrofuran solution at ^78ꢀC.1 3 The alkyl halides employed
were ethyl, n-propyl and n-butyl iodides, and benzyl bromide (Scheme 2). The alkylations were
generally complete in 2 h, and yielded only one diastereomeric product 3 as far as could be
discerned by NMR, in 95% yield (Table 1). However, when the reaction was performed at ^20ꢀC
two diastereomeric products were discernible, as indicated by the resonance of the (O-C-H)
proton at C5: this appeared as a doublet of a doublet in the ꢀ 4.04±4.16 region with a mean
separation of 0.05 ppm between the diastereomers; the diastereomeric excess was consistently
33% at ^20ꢀC. All the alkylated products 3 were characterised by IR, 1H and 13C NMR, and high
resolution mass spectra.
Table 1
Yields and diastereomeric excesses (d.e.) obtained for the alkylation of the
oxazoline 2 with RX at ^78ꢀC, and yields and enantiomeric excesses
obtained for the hydrolyses of the alkylated oxazolines 3 to the carboxylic
acids 4
The hydrolyses of the alkylated oxazolines 3 were then investigated, previous reports1 3 having
indicated that these could be accompanied by considerable racemisation. Indeed, the hydrolysis
of the benzyl derivative 3d in 6N HCl furnished only racemic 2-methyl-3-phenylpropanoic acid;
with 6N or 18N H2SO4 an enantiomeric excess of only 40% was realised. It seemed likely1 3 that
the racemisation occurred during the slow cleavage of the partially hydrolysed intermediates,
particularly the hydroxyamide derived by cleavage of the O±C3 bond in 3d. This was avoided by
cleaving the oxazoline ring with re¯uxing acetic anhydride over 24 h to obtain the putative N-
acetyl-O-acyl derivative of the amino alcohol 1 (`acyl' as derived from 4): this was not isolated but
hydrolysed further in re¯uxing methanolic KOH to the ®nal products, the 2-methylalkanoic acids
4 and the N-acetyl derivative of 1 (Scheme 2). The acids 4 were thus obtained in around 45%
1
yield (after distillation), and characterised by their physical constants, H NMR spectra and

2252
S. Chandrasekhar, A. Kausar / Tetrahedron: Asymmetry 11 (2000) 2249±2253
speci®c rotations. The above N-acetyl derivative was isolated and hydrolysed with re¯uxing 6N
HCl for 12 h to recover the chiral auxiliary, the amino alcohol 1 in 60% yield.
The speci®c rotations showed that the e.e. values for the acids 4 obtained as above were in the
range 93±98% (Table 1). Thus, although the yields of 4 are moderate as compared to the Meyers
methodology1 3 (ꢁ45% vs 62±84%), the e.e. values are substantially higher in the present method
(ꢁ95% vs. ꢁ75%). Also to be considered is that the reported methodology1 3 in some cases
employs lower temperatures for the alkylation than in the present case (Table 1).
From the known1 3 correlation of con®guration and rotational sign for 4, the presently
obtained formsÐbeing levorotatoryÐmust possess the R con®guration at C2. This suggests a
transition state for the alkylation of the oxazoline 2 as depicted in Scheme 3: the (Z)-azaenolate
su€ers electrophilic attack at the face that corresponds to the endo side of the camphane moiety.
Molecular models indicate that the alternative face is severely hindered by the gem-dimethyl
bridge (C10): its methyl group syn to the oxazoline moiety looms over one face of the azaenolate;
also, the overall rigidity of the camphane framework ensures that this steric shielding cannot
easily be overcome by simple deformations. Interestingly, the (C10) bridge corresponds to the C5-
substituent, and the C6-methylene bridge (in 2) to the C4-substituent, in Meyers oxazoline (cf.
Scheme 1). Also noteworthy is that the camphane skeleton ensures that the above two bridging
substituents in the present oxazoline 2 are mutually gauche rather than anti as in the Meyers case
(Scheme 1). Clearly, the origins of the stereoselectivity in these two cases are di€erent.
Scheme 3.
A temperature dependence study indicated that the temperature at which the alkylation
was performedÐrather than the temperature at which the deprotonation was performedÐ
determined the stereoselectivity. Thus, deprotonation of the oxazoline 2 as above at either ^78ꢀC
or ^20ꢀC, followed by alkylation with n-butyl iodide at ^78ꢀC, a€orded the (hex-2-yl)oxazoline 3c
in high diastereoselectivity (d.e. >95%, by NMR). However, deprotonation of 2 at ^78ꢀC
followed by alkylation (n-BuI) at ^20ꢀC furnished 3c in very low diastereoselectivity (d.e. ꢁ30%,
by NMR). These results most likely indicate a weighted equilibrium favouring the Z-form of the
oxazoline anion over the E-form at ^78ꢀC, followed by preferential `underside' alkylation
(Scheme 3). The alternative would involve a selectiveÐbut temperature independent and hence
unlikelyÐformation of the Z-form.
In fact, the above results accord well with the detailed results of the Meyers group,1 3 which
showed that the temperature of metallation did not a€ect the stereoselectivity in the case of LDA
(although it did in the case of n-BuLi). The present mechanism departs from the Meyers
mechanism in allowing for equilibration between the (E)- and (Z)-azaenolates, with the latter

S. Chandrasekhar, A. Kausar / Tetrahedron: Asymmetry 11 (2000) 2249±2253
2253
being more stable and so prevailing at lower temperatures (it may well be that the internal
chelation prevents isomerisation of the azaenolate in the Meyers oxazoline).
In summary, a novel chiral oxazoline has been introduced, which compares favourably
with available analogues. Its utility in the synthesis of chiral 2-alkylalkanoic acids has been
demonstrated. Both the ®rst alkylation step and the subsequent hydrolysis step are characterised
by very high stereoselectivity, although the hydrolysis occurs in relatively moderate yields.
Further work that would extend the scope of the present study is planned.
Acknowledgements
University Grants Commission (New Delhi) is thanked for ®nancial support (to A.K.).
References
1. Gant, T. G.; Meyers, A. I. Tetrahedron 1994, 50, 2297±2360.
2. Lutomski, K. A.; Meyers, A. I. Asymmetric Synthesis; Academic Press: New York, 1984; Vol. 3 (Part B); pp. 213±
274 and references cited therein.
3. Meyers, A. I.; Knaus, G.; Kamata, K.; Ford, M. E. J. Am. Chem. Soc. 1976, 98, 567±570 and references cited
therein.
4. Meyers, A. I.; Shipman, M. J. Org. Chem. 1991, 56, 7098±7102 and references cited therein.
5. (a) Banks, M. R.; Blake, A. J.; Cadogan, J. I. G.; Dawson, I. M.; Gosney, I.; Grant, K. J.; Gaur, S.; Hodgson,
P. K. G.; Knight, K. S.; Smith, G. W.; Stevenson (nee Thomson), D. E. Tetrahedron 1992, 48, 7979±8006. (b) Yan,
T.-H.; Chu, V.-V.; Lin, T.-C.; Wu, C.-H.; Liu, L.-H. Tetrahedron Lett. 1991, 32, 4959±4962.
6. Palomo, C.; Oiarbide, M.; Gonzalez, A.; Garcia, J. M.; Berree, F. Tetrahedron Lett. 1996, 37, 4565±4568.
7. Ahn, K. H.; Lim, A.; Lee, S. Tetrahedron: Asymmetry 1993, 4, 2435±2436.
8. Oppolzer, W.; Moretti, R.; Thomi, S. Tetrahedron Lett. 1989, 30, 5603±5606.
9. Ishidate, M.; Kawada, A. Pharm. Bull. (Tokyo) 1956, 4, 483±486; Chem. Abstr. 1957, 51, 14625b.
10. Beak, P.; Harris, B. R. J. Am. Chem. Soc. 1974, 96, 6363±6372.
11. Haslanger, M. F.; Heikes, J. Synthesis 1981, 801±802 and references cited therein.