Radical-chain deoxygenation of tertiary alcohols, protected as their methoxymethyl (MOM) ethers, using thiols as polarity-reversal catalysts

Article abstract of DOI:10.1039/b000533i

The deoxygenation of tertiary alcohols can be accomplished by heating their MOM ethers in the presence of a peroxide initiator and a thiol catalyst: the proposed radical-chain mechanism is supported by EPR spectroscopic studies.

Full text of DOI:10.1039/b000533i

Radical-chain deoxygenation of tertiary alcohols, protected as their
methoxymethyl (MOM) ethers, using thiols as polarity-reversal catalysts
Hai-Shan Dang, Paola Franchi† and Brian P. Roberts*
Christopher Ingold Laboratories, Department of Chemistry, University College London, 20 Gordon Street, London,
UK WC1H 0AJ. E-mail: b.p.roberts@ucl.ac.uk
Received (in Liverpool, UK) 19th January 2000, Accepted 14th February 2000
The deoxygenation of tertiary alcohols can be accomplished
by heating their MOM ethers in the presence of a peroxide
initiator and a thiol catalyst: the proposed radical-chain
mechanism is supported by EPR spectroscopic studies.
Several free-radical based procedures are available for the
deoxygenation of secondary alcohols (RsOH). In particular,
methods that involve radical-chain reactions of thiocarbonyl
derivatives RsOC(NS)X with tributyltin hydride have found
wide application in organic synthesis.1 However, because of the
thermal instability of many corresponding thiocarbonyl deriva-
tives of tertiary alcohols, the number of methods available for
the deoxygenation of the latter is more limited.1 Conversion to
an ether is commonly used to protect the alcohol functionality
during chemical manipulations2 and it would be convenient if
appropriate ether derivatives of tertiary alcohols (generalised as
RtOCHR2) could be induced to undergo radical-chain deprotec-
tion and deoxygenation in a single step to yield the deoxy
compound RtH, as shown in Scheme 1. Although the b-scission
Scheme 2
corresponding deoxy compounds 2b–d in good yields
(Scheme 2); only traces of RtH were formed in the absence of
the thiol catalyst and most of the MOM ether could be recovered
Scheme 1
step B is expected to be rapid above room temperature,3 the
hydrogen transfer step A will be close to thermoneutral and also
suffer from adverse polar effects in the transition state (a
nucleophilic radical Rt· is required to abstract hydrogen to give
unchanged.‡
In order to explore its applicability to more complex systems,
we made use of the procedure in the conversion of diacetone d-
glucose 3 to 3-deoxy-1,2:5,6-di-O-isopropylidene-3-C-methyl-
a-d-allofuranose 4 [eqn. (4)]. Swern oxidation followed by
another nucleophilic radical RtOCR2), making it a relatively
˙
slow process at moderate temperatures.4 Consonant with this
analysis, no 2-methyladamantane 2a was formed after heating
the methoxymethyl (MOM) ether of 2-methyl-2-adamantanol
1a in refluxing octane for 2 h in the presence of the 2,2-di-tert-
butylperoxybutane (DBPB, 2 3 3 mol%) as a thermal source of
initiating alkoxyl radicals. We reasoned that step A should be
subject to polarity-reversal catalysis4 by a protic catalyst such as
a thiol when, in the case of a MOM ether, it would be replaced
by the cycle of polarity-matched reactions (1) and (2). Coupled
with the b-scission reaction (3), this sequence would then
provide a pathway for the thiol-catalysed deprotection and
deoxygenation of the MOM ethers of tertiary alcohols to give
RtH and methyl formate.
treatment of the derived ketone with methylmagnesium iodide
afforded the tertiary alcohol 5,7 which was converted to the
corresponding MOM ether 6. The latter was heated under reflux
in octane in the presence of TBST (3 3 3 mol%), DBPB (3 3
3
mol%) and 2,4,6-trimethylpyridine§ (collidine,
1
3
When the previously-attempted reductive deprotection of 1a
was repeated in the presence of tri-tert-butoxysilanethiol5
(TBST; 2 3 3 mol%), the reaction now proceeded smoothly and
2-methyladamantane was isolated in 87% yield. Similar
treatment of the tertiary alkyl MOM ethers 1b–d gave the
† On study leave from the University of Bologna, Italy.
DOI: 10.1039/b000533i
Chem. Commun., 2000, 499–500
This journal is © The Royal Society of Chemistry 2000
499

the mixed acetal. However, for the tertiary alkyl MOM ethers
examined in this work, the major fate of any radicals of the type
RtOCH2OCH2 (or of radicals which might possibly be formed
˙
by abstraction of hydrogen from the Rt group)** must be
‘repair’ by hydrogen-atom transfer from the thiol to regenerate
the starting MOM ether.
Financial support for this work was provided by the EPSRC.
We thank Professors K. J. Hale and W. B. Motherwell for
helpful discussions and we are grateful to Peroxid-Chemie
GmbH for a gift of 2,2-di-tert-butylperoxybutane.
Notes and references
‡ Representative procedure: a solution containing the MOM ether 1a
(210 mg, 1.0 mmol), TBST (9 mg, 3 mol%) and DBPB (18 ml of a 50% w/w
solution in mineral oil, 3 mol%) in dry octane (1.2 cm3) was stirred and
heated under gentle reflux (bath temperature 130 °C, pre-heated) under an
atmosphere of argon. After 40 min, more initiator and thiol (3 mol% of
each) were added and heating was continued for a further 2 h. The solvent
was removed by evaporation under reduced pressure and the residue was
subjected to flash chromatography on silica gel (hexane eluent) to give
2-methyladamantane 2a (131 mg, 87%), mp 146–147 °C (lit.6 mp
146–148 °C).
§ The yield was improved in the presence of collidine, probably because this
acts as a scavenger of acid resulting from reactions between the initiator and
the thiol.8 Collidine was also present during the reductive deprotection of 1d
(Scheme 2).
¶ The value of (kb/2kt) at a given temperature is equal to ([But·]/
[8]){[But·] + [8] + [9]}.11 The Arrhenius relation for the self-reaction of tert-
butyl radicals in fluorobenzene was taken12 to be log10(2kt/M21 s21) =
11.6 2 10.2/q and the rate constants for the diffusion-controlled reactions of
But· with 8 and with 9 are assumed to be equal to 2kt.
Fig. 1 EPR spectrum recorded during UV irradiation of a fluorobenzene
solution containing ButOCH2OMe and di-tert-butyl peroxide at 261 °C.
The two central ‘lines’ (showing second-order fine structure) from the tert-
butyl radical are indicated by asterisks and the doublet of quartets from 8 by
filled circles; the remaining lines arise from 9. At 261 °C, the splitting
constants for 8 (g = 2.0031) are 11.3 G (1 Ha) and 0.91 G (3 Hg); for 9 (g
= 2.0032) they are 18.0 G (2 Haaverage) and 0.87 G (2 Ht). The central
multiplet in the spectrum of 9 is broadened because rotation about the Ca–O
bond is occurring at an intermediate rate on the EPR timescale.
10 mol%), to afford, after chromatography, a 91:9 mixture of 49
and its C-3 epimer in a total isolated yield of 90%. The
predominance of 4 in the epimeric product mixture is evidently
a result of the preferential attack by TBST at the exo-face of the
intermediate radical 7.
∑ The relative molar rate constants (k8/k9) for hydrogen abstraction by ButO·
from ButOCH2OMe to give 8 and 9 were determined by measuring relative
radical concentrations in the temperature range 270 to 220 °C.13 The value
of k8/k9 is given by {[8] + [But·]}/[9] and was shown to conform to the
Arrhenius relation log10 (k8/k9) = 20.59 + 4.4/q; the extrapolated value of
k8/k9 at 126 °C is 0.97.
In order to gain further insight into the mechanism of this
redox process, the reaction of ButOCH2OMe with tert-butoxyl
radicals was investigated by EPR spectroscopy. UV irradiation
of a fluorobenzene solution containing ButOCH2OMe (1.4 M)
and di-tert-butyl peroxide (ca. 20% v/v) at 261 °C, while the
sample was in the microwave cavity of the spectrometer,10
afforded the EPR spectrum shown in Fig. 1. The three radicals
** Epimerisation8 at C-5 did not compete with the reductive deprotection
reaction of 6 to give 4.
ButOCHOMe 8, ButOCH2OCH2 9 and But· are present at this
temperature, while above ca. 235 °C only the spectra of the last
two were detectable. The rate constant (kb) for b-scission of 8 to
give tert-butyl radicals was determined relative to the rate
constant (2kt) for self-reaction of the latter, using the established
‘steady-state’ EPR method,11 and the Arrhenius relation
obtained from measurements in the temperature range 270 to
240 °C is given in eqn. (5), where q = 2.303RT kJ mol21.¶
˙
˙
1 W. B. Motherwell and D. Crich, Free Radical Chain Reactions in
Organic Synthesis, Academic Press, London, 1992, pp. 47–50, 54-58.
2 T. W. Green and P. G. M. Wuts, Protective Groups in Organic
Synthesis, Wiley-Interscience, New York, 3rd edn, 1999,
3 M. J. Perkins and B. P. Roberts, J. Chem. Soc., Perkin Trans. 2, 1975,
77.
4 B. P. Roberts, Chem. Soc. Rev., 1999, 28, 25.
5 A. Herman, B. Becker and W. Wojnowski, Z. Anorg. Allg. Chem., 1979,
450, 178.
log10(kb/s21) = (12.0 ± 0.5) 2 (33.8 ± 1.5)/q
(5)
6 S. F. Nelsen, G. R. Weisman, E. L. Clennan and V. E. Peacock, J. Am.
Chem. Soc., 1976, 98, 6893.
7 Z. Ali, S. Qureshi and G. Shaw, J. Chem. Soc., Perkin Trans. 1, 1990,
2627.
8 H.-S. Dang and B. P. Roberts, Tetrahedron Lett., 1999, 40, 4271.
9 O. R. Martin, R. C. Nabinger, Y. Ali, D. M. Vyas and W. A. Szarek,
Carbohydr. Res., 1983, 121, 302.
10 J. A. Baban and B. P. Roberts, J. Chem. Soc., Perkin Trans. 2, 1981,
161; V. Diart and B. P. Roberts, J. Chem. Soc., Perkin Trans. 2, 1992,
1761.
11 D. Griller and B. P. Roberts, J. Chem. Soc., Perkin Trans. 2, 1972, 747;
D. Griller and K. U. Ingold, Acc. Chem. Res., 1980, 13, 317.
12 H.-H. Schuh and H. Fischer, Helv. Chim. Acta., 1978, 61, 2130; J. C.
Walton, J. Chem. Soc., Perkin Trans. 2, 1987, 231.
13 A. G. Davies, D. Griller and B. P. Roberts, J. Chem. Soc. B, 1971, 1823;
D. Griller and K. U. Ingold, Acc. Chem. Res., 1980, 13, 193; H. Fischer
and H. Paul, Acc. Chem. Res., 1987, 20, 200.
The value of kb extrapolated to 126 °C (the boiling point of
octane) is 3.8 3 107 s21. Even at +20 °C, there was no EPR
evidence for the b-scission of 9 to give ButOCH2 although,
˙
because of line overlap, the latter radical could be difficult to
detect in the presence of a higher concentration of the former.
However, in similar experiments with MeOCH2OMe, only the
˙
˙
spectra of MeOCHOMe and MeOCH2OCH2 were observed up
˙
to +72 °C; in particular MeOCH2 was not detected, supporting
the conclusion that b-scission of 9 to give ButOCH2 and
˙
formaldehyde must be very much slower than the cleavage of 8
to give But· and methyl formate. Although the selectivity of
ButO· in hydrogen-atom abstraction from ButOCH2OMe will
differ quantitatively from that of (ButO)3SiS·,∑ both radicals are
electrophilic and it is probable that the thiyl radical would also
abstract hydrogen to some extent from the O-methyl group of
Communication b000533i
500
Chem. Commun., 2000, 499–500