D. R. Kelly et al. / Tetrahedron Letters 43 (2002) 9331–9333
9333
1
converted to the hydroxyketone 16. H NMR analysis
of the salt reaction in deuterochloroform (in which the
salt is almost insoluble) showed a trace of hydroxyke-
tone 16, which may have been formed during evapora-
tion of the reaction mixture. This was confirmed when
the two reactions and a blank were repeated with
monitoring by IR. The endoperoxide isomerisation
methanol or t-butanol relative to methylene chloride.
These results strongly suggest that alkoxides do not play
a significant role in the mechanistic pathway to hydroxy-
ketones. Kelly, D. R.; Bansal, H.; Morgan, J. J. G.
Tetrahedron, in preparation.
7. N,N-Dimethylisoxazolidinium salts undergo a compara-
ble elimination: Casuscelli, F.; Chiacchio, U.; Rescifina,
A.; Romeo, R.; Romeo, G.; Tommasini, S.; Uccella, N.
Tetrahedron 1995, 51, 2979–2990.
−
1
showed a smooth increase in absorption at 1717 cm
due to the carbonyl group of the hydroxyketone 16, but
there was no change in the salt reaction. Perturbation
of the carbonyl absorption by the presence of the salt
was excluded by a control experiment, in which an
aliquot of the salt reaction was spiked with 10 mol% of
the hydroxyketone 16. This mixture showed the
expected carbonyl absorption. Finally, treatment of the
salt 15 with stronger bases such as potassium t-butox-
ide or DBU in THF caused conversion to the hydroxy-
ketone within 15 min and 2 days, respectively. These
results quite clearly exclude the O-alkyl-hydroxylam-
monium salt 4 as an intermediate in the isomerisation
of the endoperoxide 2. Efforts are continuing to
develop catalysts for the enantioselective Kornblum–
DeLaMare rearrangement of prochiral endoperoxides.
8
. Adam, W.; Stossel, R.; Treiber, A. J. Org. Chem. 1995,
0, 2879–2884.
6
9
. In the absence of a trap, Grob fragmentation to give two
ketones occurs: Adam, W.; Heil, M. J. Am. Chem. Soc.
1992, 114, 5591–5598.
1
1
0. Endoperoxide 2 was prepared by cycloaddition of singlet
oxygen to 1,3-cyclohexadiene and endoperoxide 14 by
diimide reduction: Bloodworth, A. J.; Eggelte, H. J. J.
Chem. Soc., Perkin Trans. 1 1984, 2069–2072.
1. Prepared by reaction of enantiomerically pure ‘cis-tolu-
ene glycol’ with 2,2-dimethoxypropane, zinc chloride and
cycloaddition of singlet oxygen: Hudlicky, T.; Luna, H.;
Barieri, G.; Kwart, L. D. J. Am. Chem. Soc. 1988, 110,
4735–4741.
1
2. O,N,N-Trimethylhydroxylamine is methylated by methyl
iodide in the absence of solvent: Jones, L. W.; Major, R.
T. J. Am. Chem. Soc. 1928, 50, 2742–2747.
Note: All new compounds described in this paper were
1
13
fully characterised by 360 MHz H NMR, 90 MHz,
C
NMR, IR, mass spectrometry and combustion analysis
or HRMS.
13. Sengul, M. E.; Ceylan, Z.; Balci, M. Tetrahedron 1997,
53, 10401–10408.
1
4. Under identical conditions the pK of pyridinium hydro-
a
chloride was 4.5 which is in good agreement with previ-
Acknowledgements
ous measurements. pK 4.45, MeOH:H O, 1:1, Stul’, Y.
a
2
B.; Emel’yanov, V. I. J. Org. Chem. USSR 1984, 20,
159–1160.
5. Bissot, T. C.; Parry, R. W.; Campbell, D. H. J. Am.
1
We thank the former Zeneca Ltd for support of a
CASE award to J.J.G.M. and the EPSRC for an alloca-
tion of spectrometer time at the Mass Spectrometry
Centre, University of Swansea.
1
Chem. Soc. 1957, 79, 796–800. This is the only report that
we could find of the pK of an unconjugated N,N-disub-
a
stituted-aminoether lacking other basic or acidic groups;
for related compounds, see: Khan, M. N.; Arifin, Z.;
George, A.; Wahab, I. A. Int. J. Chem. Kinetics 2000, 32,
References
146–152.
1
6. Kelly, D. R.; Harrison, S.; Jones, S.; Masood, A.; Mor-
gan, J. J. G. Tetrahedron Lett. 1993, 34, 2689–2690.
7. Methylene chloride reacts slowly with triethylamine to
give chloromethyltriethylammonium chloride. A blank
lacking the starting material was run to monitor the
formation of chloromethyltriethylammonium chloride.
This was used as the solvent blank in the IR monitored
experiments. Less than 1% conversion occurred during
the course of the reaction as reported previously: Wright,
D. A.; Wulff, C. A. J. Org. Chem. 1970, 35, 4252. See
also: Almarzoqi, B.; George, A. V.; Isaacs, N. S. Tetra-
hedron 1986, 42, 601–607. We have also found that
chloromethylquinuclidinium chloride does not catalyse
endoperoxide isomerisation and does not affect the rate
of isomerisation by dimethylamine.
1
. Schwaebe, M. K.; Little, R. D. Tetrahedron Lett. 1996,
3
7, 6635–6638.
1
2
3
. Balci, M. Chem. Rev. 1981, 81, 91–108.
. Kornblum, N.; DeLaMare, H. E. J. Am. Chem. Soc.
1
951, 73, 880–881.
4. Roberts, L. J.; Salomon, R. G.; Morrow, J. D.; Brame,
C. J. FASEB J. 1999, 13, 1157–1168.
5
6
. Salomon, R. G. Acc. Chem. Res. 1985, 18, 294–301.
. Both the deprotonation and peroxide attack mechanisms
require prior protonation of the distal peroxide oxygen
atom or at least transient formation of an alkoxide.
5
However, Salomon has shown that the presence of acetic
acid favours formation of the hydroxyketone 7b plus the
enone 8b and we have shown that the rate of isomerisa-
tion of the endoperoxide 2 is greatly accelerated in