Perruthenate ion. Another metal oxo species able to promote the oxidative cyclisation of 1,5-dienes to 2,5-disubstituted cis-tetrahydrofurans

Article abstract of DOI:10.1016/j.tetlet.2003.10.173

Perruthenate ion, from tetrapropylammonium perruthenate (TPAP), in the presence of tetrabutylammonium periodate (TBAPI) as reoxidant catalyses the stereospecific and stereoselective oxidative cyclisation of 1,5-dienes to cis-2,5-disubstituted tetrahydrofu

Full text of DOI:10.1016/j.tetlet.2003.10.173

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 303–308
Perruthenate ion. Another metal oxo species able to promote
the oxidative cyclisation of 1,5-dienes to 2,5-disubstituted
cis-tetrahydrofurans
*
Vincenzo Piccialli and Teresa Caserta
Dipartimento di Chimica Organica e Biochimica, Universit ꢀa degli Studi di Napoli, ‘Federico II’, Via Cynthia 4, 80126 Napoli, Italy
Received 18 August 2003; revised 24 October 2003; accepted 29 October 2003
Abstract—Perruthenate ion, from tetrapropylammonium perruthenate (TPAP), in the presence of tetrabutylammonium periodate
TBAPI) as reoxidant catalyses the stereospecific and stereoselective oxidative cyclisation of 1,5-dienes to cis-2,5-disubstituted
(
tetrahydrofurans in good to moderate yields. NMO also works as co-oxidant in the process but is less effective and at least 0.7 equiv
TPAP are required. Acidic conditions promote the formation of THF diols. 1,5-Dienes with two terminal double bonds give poor
yields of THFꢀs or cleavage products.
ꢀ
2003 Elsevier Ltd. All rights reserved.
ꢀ
4
1d;e
The metal oxo reagents OsO , RuO and MnO are the
tion of 1,5-dienes to cis-THF diols,
covered a few years ago in our laboratories.
a process dis-
4
4
1c
only known species able to accomplish the stereospecific
and stereoselective conversion of 1,5-dienes to cis-2,5-
bis(hydroxymethyl)tetrahydrofurans (cis-THF diols).
This is a very effective process by which four stereogenic
centres are created in a single stepfrom an achiral
reagent with complete internal stereocontrol. Stereo-
chemically complex tetrahydrofuran fragments, useful
in the synthesis of biologically active natural substances,
1
3
As a continuation of our studies in this area, we report
here the discovery of a novel oxidative process strictly
related to the above-mentioned ones. In particular,
ꢀ
perruthenate ion (RuO
4
), from commercially available
4
tetrapropylammonium perruthenate (TPAP), in the
presence of tetrabutylammonium periodate (TBAPI) or
4-methylmorpholine N-oxide (NMO) as reoxidants, can
catalyse the stereoselective conversion of 1,5-dienes into
2
such as Annonaceous acetogenins, can be easily accessed
through the above processes.
2,5-cis-disubstituted THFꢀs.
Interest in this field appears still to be keen. Recent
ꢀ
advances in the above chemistry include the MnO4
-
In a preliminary experiment treatment of geranyl acetate
(1, GA, Scheme 1) with excess TPAP (10 equiv) in
CH Cl cleanly afforded THF ketol 2, but the process
2 2
stopped at a ca. 10–20% conversion (Table 1). This
result established the ability of the perruthenate ion to
promoted formation of enantiomerically enriched
mono-THF diols from dienones in the presence of a
5a
1n
chiral phase-transfer catalyst, and an effective
improvement of the OsO -catalysed oxidative cyclisa-
4
OAc
TPAP (excess)
0-20% conversion
OAc
O
1
O
H
HO
Geranyl acetate (1)
2
Scheme 1.
ꢀ
4
Keywords: Oxidative cyclisation; 1,5-Dienes; cis-Tetrahydrofurans; RuO ; TPAP; TBAPI; NMO.
*
Corresponding author. Tel.: +39-81-674111; fax: +39-81-674393; e-mail: vinpicci@unina.it
0
040-4039/$ - see front matter ꢀ 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2003.10.173

304
V. Piccialli, T. Caserta / Tetrahedron Letters 45 (2004) 303–308
Table 1
All the above evidence can be summarised as follows.
a
Entry
TPAP (equiv)
NMO (equiv)
2/GA
Both the TPAP and NMO amounts concur to push the
reaction ahead. THF formation is strongly affected by
the degree of alkylation of the diene: less alkylated
dienes require higher amounts of reagents. In addition,
1,5-dienes possessing terminal double bonds are unre-
active under the above conditions.
1
2
3
4
5
6
7
8
9
10
0.05
0.1
0.5
0.7
1
––
20
1/4
1.5/1
1/1.5
2.3/1
1/0
3–5
10
10–12
2
3
4
5
3
4
4
2.3/1
3.6/1
1
Due to the high amounts of reagents required, the above
process appeared rather unattractive. Therefore, we
turned our attention to the use of tetrabutylammonium
periodate (TBAPI) as co-oxidant, a reagent able to
deliver the periodate ion in an organic medium. Grati-
fyingly, 5% TPAP and 5 equiv of TBAPI gave, in 4.5 h,
complete consumption of geranyl acetate and the for-
1
12.0/1
1/0
1
1
1
1
0
1
2
2
5.7/1
7.3/1
1/0
2
3
a
1
Calculated by H NMR on the crude reaction mixture.
mation of the THF ketone 2 and the corresponding diol
6
1
3, as a ca. 1:1 mixture. By leaving the mixture over-
night the complete over-oxidation of 13 to 2 took place
coincidentally giving the same 82% yield of THF 2
(Table 3, entry 1), as was obtained with the TPAP/NMO
oxidising system described before.
promote THF formation, therefore, our next objective
was to force the process to completion.
With this aim, the use of the system TPAP (cat)/NMO/
4
ꢁ
A molecular sieves, known to accomplish alcohol
Oxidation of the other 1,5-dienes under these conditions
led to good yields of the expected THF products. In
particular, dienes belonging to the geraniol series gave
yields comparable to, or slightly higher than, those
obtained by the process with NMO (Table 3, entries 2
and 3), while for the disubstituted diene 8 the yield was
improved to 68% (entry 5). In some cases 10% TPAP
proved more effective (entries 4–6). Geranic acid methyl
ester was cyclised in a 41% yield (entry 4), but excess
TBAPI (10 equiv) was required. The rather low yield for
this process was due to the poor mass recovery after
work-up. This behaviour is unclear and is peculiar to
this conjugated diene. It is possible that the remainder of
material for the mass balance could be made upof
highly polar, water soluble, or volatile substances, given
4
6
oxidation, was investigated. Disappointingly, the use
of 5–10% TPAP in the presence of 3–5 equiv of NMO,
still gave only a modest 35–40% conversion of GA into
THF 2 (Table 1, entry 3), once again obtained as the
sole reaction product. However, complete GA con-
sumption was reached in the presence of 0.7 equiv TPAP
and excess NMO (10–12 equiv), affording 2 in a good
8
2% isolated (HPLC) yield (Table 1, entry 5; Table 2,
entry 1). These conditions worked well with neryl ace-
tate (3) and 2,6-dimethyl-2,6-octadienol diacetate (4) for
which high yields (76% for both) of the expected THF
products were obtained (Table 2, entries 3 and 5,
respectively), but proved ineffective for the related ger-
anic acid methyl ester (6), which gave no THF product
while a 50% conversion was obtained for ethyl (E,E)-
nonadeca-4,8-dienoate (8). Oxidation of both these
dienes could be forced to completion only in the pres-
ence of larger amounts of both TPAP and NMO
1
the high purity of crude 7, as estimated from its H
NMR.
The behaviour of 2,5-dimethyl-1,5-hexadiene under the
new conditions paralleled that of geranic acid methyl
(
2 equiv/25 equiv, entries 7 and 9, respectively); the fact
5b
that THF 7 from geranic acid methyl ester was obtained
in 34% yield indicated that conjugation of one of the
double bonds of the diene affected its reactivity (com-
pare entries 1 and 7). This result is in sharp contrast with
ester in that a THF product, lactone 12, was obtained
(note that the oxidation with TPAP/NMO gave no THF
product), albeit in a poor 22% yield (entry 6). This
product is likely to have been produced by over-oxida-
tion of one of the primary alcohol functions of the ini-
tially formed THF diol to an aldehyde, followed by
internal hemiacetalisation and further oxidation to give
the lactone. The remaining material (ca. 70%) was made
upof 2,5-hexanedione derived from cleavage of both
double bonds. Though the yield of 12 is low, the pos-
sibility of obtaining, in a single step, a compound more
complex than the parent THF diol, without additional
the oxidation of this substrate with RuO
yields comparable to, or higher than, that for the
4
for which
1b
unconjugated GA are obtained. 2,5-Dimethyl-1,5-
hexadiene (10) and 1,5-hexadiene (11) gave no THF
product (entries 11 and 13) even with excess (up to
3
0 equiv) NMO.
Oxidations of geranyl acetate conducted with stoichio-
metric, or higher, amounts of TPAP were also tested
7
protection–deprotection steps, appears nontrivial. 1,5-
Hexadiene still gave no THF, possibly due to the for-
(
Table 1). With 1 equiv of TPAP at least 5 equiv of
NMO were required for the process to go to completion
entry 9) and a further increase in NMO (upto 12 equiv)
mation of volatile (C ) aldehyde product(s) produced by
4
scission, as observed for 2,5-dimethyl-1,5-hexadiene,
and/or because of the volatility of the diene itself.
(
did not affect the yield of the THF, which was invariably
in the 70–78% range. On the other hand, less NMO
(
4 equiv) required excess TPAP (3 equiv) (entry 12) but
the process was faster (30 min) once again giving 2 in
2% yield.
Thus, the process with TPAP/TBAPI is more efficient
due to the acceptably low amounts of both reagents
(TPAP 5–10%, TBAPI 5 equiv) used, higher yields for
8

V. Piccialli, T. Caserta / Tetrahedron Letters 45 (2004) 303–308
305
Table 2
Entry
1,5-Diene
TPAP
equiv)
NMO
(equiv)
AcOH
(equiv)
Product
Yield
(%)
(
OAc
OAc
1
2
0.7
2
12
3
––
O
O
82
HO
H
H
2
OAc
OAc
500
13 (55)
2 (22)
1
O
+ 2
OH
HO
1
3
3
4
0.7
2
12
2
––
2
76
OAc
500
14 (44)
2 (33)
O
+2
3
HO
OH
H
1
4
AcO
OAc
5
6
0.7
2
12
2
––
76
O
5
O
HO
H
AcO
OAc
AcO
OAc
4
500
15 (62)
5 (18)
+
5
O
1
OH
OHH
5
CO Me
2
7
8
2
4
25
5
––
34
O
7
O
HO
HO
H
CO Me
2
6
CO Me
2
500
16 (20)
7 (8)
O
+ 7
OH
H
1
6
H21C10
(CH )CO Et
2 2
9
2
4
25
15
––
59
O
9
O
O
H
H
H
H C
(CH ) CO Et
2 2 2
2
1
10
H21C10
(CH )CO Et
2
2
8
10
500
17 (17)
9 (29)
O
1
O
OH
+ 9
H
7
1
1
2
2
2
20
2
––
No THF
No THF
––
––
1
500
1
0
1
1
3
4
2
2
20
2
––
No THF
––
500
<5
1
1
O
OH
HO
H
H
1
8
some dienes, and the ability of this pair to force diene 10
to cyclise.
oxidation of the THF diol products initially formed in
the process with TPAP/TBAPI was slow, an attempt
was made to stopthe oxidation at the diol stage. We
were unable to obtain this product alone even quenching
the process at an early stage (1–2 min); however, a
Our next objective was to seek out conditions to obtain
THF diols. Based on the observation that the over-

306
V. Piccialli, T. Caserta / Tetrahedron Letters 45 (2004) 303–308
Table 3
Entry
1,5-Diene
TPAP/TBAPI (equiv)
Product
Yield (%)
1
2
3
4
5
1
3
4
6
8
0.05/5
0.05/5
0.05/5
0.1/10
0.1/5
2
2
5
7
9
82
76
79
41
68
6
10
0.1/5
12
22
O
O
O
RuO -
4
NMO or TBAPI
(Ru oxidation)
OAc
OAc
OAc
O O
Ru
O O
Ru
GA
2
0
1
9
(
O)n
-
O
O
O
Cyclisation
OAc
OH
(
a)
O
OAc
H
H
HO
HO
1
3 Ox
O
Ru species +
O
b
a
b
O
OAc
(
b) Oxidative cleavage
21
a
Ru
b
O
2
O
H
(O)n
Figure 1.
3
5–45% yield of the THF diol was produced, as esti-
yield was encouraging and further proved the ability of
AcOH to facilitate the diol formation. 2,5-Dimethyl-1,5-
hexadiene (entry 12) still gave no THF, while 1,5-hexa-
diene gave only a <5% amount of the THF diol 18 (entry
14), isolated as its tosyl derivative, while no mono- or di-
aldehyde corresponding to 18 was detected.
1
mated by H NMR. This evidence suggests that both the
THF diol and ketol products were simultaneously
formed from a common intermediate through parallel
routes (see Fig. 1, later in the discussion).
The oxidative cyclisation process could be driven to-
wards the formation of THF diols under acidic condi-
The use of trifluoroacetic acid (TFA 20 equiv, TPAP
2 equiv, NMO 3 equiv) in place of AcOH was tested
in the oxidation of GA, but slightly lower (ca. 10%)
yields were obtained. On the other hand, the oxidation of
the same substrate with TPAP/TBAPI under acidic
conditions proved to be less clean compared with the
analogous process in the presence of NMO. It must be
pointed out, however, that fine tuning of the reaction
conditions for these processes has not yet been carried
out.
1d;e
ꢀ1l;m;n
tions. Precedents in OsO
suggest the use of acidic conditions for these transfor-
4
and MnO chemistry
4
mations. AcOH (500 equiv), TPAP 2 equiv and NMO 2–
6
3
equiv were selected as a good compromise in order
that the THF diol would predominate over the corre-
sponding THF ketone. In this manner oxidation of
geraniol-type substrates gave the THF diols with
acceptable 44–62% yields (Table 2, entries 2, 4 and 6) in
2
short times (15–20 min). It is worth noting that D iso-
meric 1 and 3 gave C-2 epimeric THF diols 13 and 14,
indicating that the process was stereospecific, as
observed for the cyclisations promoted by the above
A doubt could arise about the possibility that the true
oxidising agent could be RuO produced in some way
4
during the process. Though at the present level of
knowledge this cannot be firmly ruled out, we believe that
it is not plausible due to the too many differences with the
1
related oxo-species. Interestingly, no, or extremely slow,
over-oxidation of the diol was observed under these
conditions. In addition, the overall yields of THF
products did not significantly change in comparison with
the process without AcOH. For diene 8, once again more
forcing conditions were required (TPAP 4 equiv, NMO
4
analogous RuO -mediated process. In particular:
(a) A higher level of stereoselectivity was observed for
the perruthenate-mediated cyclisation since no
trans-THF isomer was formed, as is observed (in
1
5 equiv) to push the process to completion (entry 10).
However, these conditions prevented the formation of
the expected THF diol corresponding to diketone 9, but
allowed the formation of ketol 17. Geranic acid methyl
ester was still cyclised in an unsatisfactory 28% (overall)
yield. The fact that THF diol 16 was obtained in 20%
1b
very low yields) in the process with RuO .
4
(b) No scission products were observed for geraniol-
type substrates and for the unbranched diene. These
products are generally formed in minor amounts as

V. Piccialli, T. Caserta / Tetrahedron Letters 45 (2004) 303–308
307
side products in the process with RuO . In addition,
the yields of the THF products obtained from these
dienes are higher with the new process.
4
the ꢁCentro di Metodologie Chimico-Fisiche dellꢀUni-
versit ꢀa di Napoli ꢁFederico II’’ for NMR facilities.
(
(
c) The reactivity of geranic acid methyl ester appears to
be quite different, at least under the tested condi-
tions, from that of other geraniol-type substrates,
probably due to conjugation. On the contrary, this
substance gives the cis-THF product with yields
comparable to, or higher than, that for GA and
References and Notes
1
4
. RuO : (a) Charlsen, P. H. J.; Katsuki, T.; Martin, V. S.;
Sharpless, K. B. J. Org. Chem. 1981, 46, 3936–3938; (b)
Piccialli, V.; Cavallo, N. Tetrahedron Lett. 2001, 42,
1b
4
NA in the process with RuO .
d) 1,5-Dienes possessing terminal double bonds were
unreactive in the presence of TPAP but give the
expected cis-THF diols with RuO4.
4
695–4699; (c) OsO
4
: de Champdor ꢂe , M.; Lasalvia, M.;
Piccialli, V. Tetrahedron Lett. 1998, 39, 9781–9784; (d)
Donohoe, T. J.; Winter, J. J. G.; Helliwell, M.; Stemp, G.
Tetrahedron Lett. 2001, 42, 971–974; (e) Donohoe, T. J.;
Butterworth, S. Angew. Chem., Int. Ed. 2003, 42, 948–
ꢀ
In addition, some experiments aimed at excluding the
formation of RuO in the cyclising conditions have been
carried out. In particular, trans-7-tetradecene in CH Cl
9
51; (f) MnO
4
: Klein, E.; Rojahn, W. Tetrahedron 1965,
1, 2353–2358; (g) Baldwin, J. E.; Crossley, M. J.;
4
2
2
2
Lehtonen, E.-M. M. J. Chem. Soc., Chem. Commun.
979, 918–919; (h) Walba, D. M.; Wand, M. D.; Wilkes,
failed to give the expected scission or diol products,
usually obtained with RuO , on treatment with excess
TPAP (10 equiv) alone or with the TPAP /NMO
1
4
M. C. J. Am. Chem. Soc. 1979, 101, 4396–4397; (i)
Walba, D. M.; Edwards, P. D. Tetrahedron Lett. 1980,
ðcatÞ
8
system, while the same alkene gave heptanal on RuO
oxidation in the same solvent.
4
2
1, 3531–3534; (j) Spino, C.; Weiler, L. Tetrahedron Lett.
1987, 28, 731–734; (k) Walba, D. M.; Przybyla, C. A.;
Walker, C. B. J. J. Am. Chem. Soc. 1990, 112, 5624–5625;
(l) Kocienskyi, P. J.; Brown, R. C. D.; Pommier, A.;
Procter, M.; Schmidt, B. J. Chem. Soc., Perkin Trans.
Based on the above evidence and previous studies by
9
Lee et al. it seems reasonable that at least for the pro-
1
998, 9–39; (m) Brown, R. C. D.; Hughes, R. M.; Keily,
cess conducted under neutral conditions, the first step
could involve the formation of a cyclic ruthenium(V)
diester 19 through attack of the perruthenate anion at
one of the double bonds (Fig. 1). These authors reported
J.; Kenney, A. Chem. Commun. 2000, 1735–1736; (n)
Brown, R. C. D.; Keily, J. F. Angew. Chem., Int. Ed.
2
001, 40, 4496–4498; For related oxidative monocyclisa-
tions of 5,6-dihydroxyalkenes, see; Walba, D. M.; Stoudt,
G. S. Tetrahedron Lett. 1982, 23, 727–730.
. (a) Zeng, L.; Oberlies, N. H.; Shi, G.; Gu, Z.-M.; He, K.;
McLaughlin, J. L. Nat. Prod. Rep. 1996, 275–306; (b)
Alali, F. Q.; Liu, X. X.; McLaughlin, J. L. J. Nat. Prod.
1999, 62, 505–540.
. (a) See Refs. 1b,c. and Piccialli, V. Tetrahedron Lett. 2000,
41, 3731–3733; (b) Bifulco, G.; Caserta, T.; Gomez-
Paloma, L.; Piccialli, V. Tetrahedron Lett. 2002, 43,
9
that, for the oxidation of trans-cinnamate, this species
decomposes slowly to cleavage products, and also pro-
posed that it is oxidised to a Ru(VI) species in a fast
2
3
9
step. These precedents agree well with the observation
that no scission products were detected in our process
and suggest that the transformation to a more oxidised
ruthenium diester species 20 (note that (O)n only indi-
cates a higher oxidation state for Ru), by the co-oxidant,
or, in its absence, by the perruthenate ion itself, could
follow. It seems rather obvious that this species should
be different from the Ru(VI) species involved in the
9265–9269, corrigendum Tetrahedron Lett. 2003, 44,
3429; (c) Bifulco, G.; Caserta, T.; Gomez-Paloma, L.;
Piccialli, V. Tetrahedron Lett. 2003, 44, 5499–5503.
. Ley, S. V.; Norman, J.; Griffith, W. P.; Marsden, S. P.
Synthesis 1994, 639–666.
4
RuO
possibly through a [3+2] cycloaddition
4
-mediated cyclisation. Closure of the THF ring
1c;10
and release of
5. (a) The structure of all the reaction products was proven
by comparison with authentic samples synthesised from
the same dienes through the related OsO -mediated
the THF as a diol (path a) or ketol (path b) would then
occur. On the other hand, cyclisation of diester 19 itself
cannot be a priori ruled out. Finally, one can speculate
that acidic conditions would facilitate THF release as
the diol by favouring path b over path a in 21.
4
1
c;e
cyclisation followed, where needed, by TPAP/NMO
oxidation; (b) Lactone 12 was synthesised by TPAP/NMO
oxidation of the parent THF diol, in turn obtained from
,5-dimethyl-1,5-hexadiene with KMnO
. Typical procedure for the TPAP/NMO and TPAP/NMO/
AcOH processes. To the diene in CH Cl were sequentially
added 4 A molecular sieves (500 mg/mmol), NMO dis-
solved in CH Cl (ca. 15 mg/mL) and TPAP, at rt, under
1
f
2
4
.
6
In conclusion, we have discovered a new process for
obtaining 2,5-disubstituted cis-tetrahydrofurans from
2
2
ꢁ
1
,5-dienes. Conditions have been devised to obtain
2
2
THF diols of more alkylated substrates. The process
appears to be amenable for further improvement. Efforts
towards this goal, as well as to induce chirality, are
ongoing.
stirring. When the reaction was complete (TLC monitor-
ing) the mixture was concentrated and filtered through a
pad of silica gel eluting with CHCl
The organic phase was concentrated in vacuo taken up in
CHCl and washed with 0.1 M HCl, saturated NaHCO
solution and water, then dried over Na SO and concen-
3
/MeOH (9:1 fi 8/2).
3
3
2
4
trated under reduced pressure. Pure THF products were
obtained by HPLC (hexane/EtOAc mixtures). The process
with AcOH, was performed as above; the acid was added
to the reaction mixture before TPAP. The solubility of
Acknowledgements
2 2
NMO in CH Cl /AcOH was ca. 30 mg/mL. The use of
We are grateful to MURST, Italy (PRIN 2001), for
financial support in favour of this investigation and to
molecular sieves appears not to affect appreciably both
processes but their presence facilitates the work-up.

308
V. Piccialli, T. Caserta / Tetrahedron Letters 45 (2004) 303–308
Typical procedure for the TPAP/TBAPI process. To the
diene in CH Cl were sequentially added TBAPI (ca.
0 mg/mL) and TPAP, at rt, with stirring. When the
reaction was complete (TLC monitoring) an excess of a
saturated Na Æ5H O solution was added and the
mixture stirred for 30 min whereupon the whole was
poured into a separatory funnel and the phases were
separated. The aqueous phase was further extracted with
Pure THF products were obtained by HPLC (hexane/
EtOAc mixtures).
2
2
1
7. See, for example, the synthesis of related
reported in Ref. 1e.
D-chitaric acid
2
S
2
O
3
2
8. After a prolonged treatment a < 5% of (E)-tetradec-6-
en-5-one was obtained.
9. (a) Lee, D. G.; Chang, V. S.; Helliwell, S. J. Org. Chem.
1976, 41, 3644–3646; (b) Lee, D. G.; Helliwell, S. Can. J.
Chem. 1984, 62, 1085–1092.
CHCl
EtOAc and washed twice with a saturated Na
solution, dried over Na SO and concentrated in vacuo.
3
. The organic phase was evaporated, taken up in
2
S
2
O
3
Æ5H
2
O
10. Travis, B.; Borhan, B. Tetrahedron Lett. 2001, 42, 7741–
7745, and references cited therein.
2
4