Synthetic studies on the seven- and eight-membered rings by the intramolecular Nozaki-Hiyama reaction of the allylic phosphates

Article abstract of DOI:10.1016/j.tetlet.2004.09.142

Synthetic studies on the seven- and eight-membered rings by the intramolecular Nozaki-Hiyama reaction of the allylic phosphates are described. The yield greatly depends on the structure of substrate; however, some complex substrates afforded desired products in high to excellent yield.

Full text of DOI:10.1016/j.tetlet.2004.09.142

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 8653–8657
Synthetic studies on the seven- and eight-membered rings by the
intramolecular Nozaki–Hiyama reaction of the allylic phosphates
Mitsuhiro Iwamoto, Masayuki Miyano, Masayuki Utsugi, Hatsuo Kawada and
Masahisa Nakada*
Department of Chemistry, School of Science and Engineering, Waseda University, 3-4-1 Ohkubo, Shinjuku-ku, Tokyo 169-8555, Japan
Received 21 August 2004; revised 17 September 2004; accepted 22 September 2004
Available online 8 October 2004
Abstract—Synthetic studies on the seven- and eight-membered rings by the intramolecular Nozaki–Hiyama reaction of the allylic
phosphates are described. The yield greatly depends on the structure of substrate; however, some complex substrates afforded
desired products in high to excellent yield.
Ó 2004 Elsevier Ltd. All rights reserved.
Some biologically important natural products, particu-
larly terpenoids, contain seven- or eight-membered car-
bocyclic ring in their structures. For example, these
carbon skeletons can be found in TaxolTM, guanacaste-
pene, erinacine E, and variecolin (Fig. 1).1
strain, and the transannular interaction; hence, their
synthesis has been a challenging problem.
Nozaki–Hiyama reaction is a Cr(II) mediated C–C bond
forming reaction of various halides with aldehydes.2
This reaction has been studied extensively because of
their potential utility, and has also been applied to
numerous total syntheses of complex natural products
due to their high chemoselectivity, high yield, and excel-
lent compatibility with various functional groups. How-
ever, to prepare both functional groups, an allylic halide
and an aldehyde, in a same molecule is sometimes trou-
blesome because of their high reactivity. Hence, applica-
tion of the intramolecular Nozaki–Hiyama allylation
(Fig. 2) to natural product synthesis has been limited.3
These ring systems, especially eight-membered rings, are
difficult to construct because of entropy reasons, ring
Nozaki and co-workers and Knochel and co-workers
reported the intermolecular Nozaki–Hiyama reaction
of allylic phosphates with aldehydes. To our knowledge,
however, the intramolecular Nozaki–Hiyama reaction
of allylic phosphates has never been reported so far.4
This intramolecular reaction would be useful for the
Figure 1.
Keywords: Nozaki–Hiyama reaction; Intramolecular reaction; Allylic
phosphate; Carbocyclic ring, Seven- and eight-membered ring.
*
Corresponding author. Tel./fax: +81 3 5286 3240; e-mail: mnakada@
waseda.jp
Figure 2.
0040-4039/$ - see front matter Ó 2004Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.09.142

8654
M. Iwamoto et al. / Tetrahedron Letters 45 (2004) 8653–8657
construction of cyclic compounds because the allylic
phosphate is stable and easy to handle.
We have been investigating a new total synthesis of
TaxolTM, however, development of a synthetic method
for the central eight-membered ring of TaxolTM (B-ring)
has been a problem. Since Nozaki–Hiyama allylation
has been an effective C–C bond forming reaction, we
have started to investigate the intramolecular Nozaki–
Hiyama allylation as a method for constructing the B-
ring of TaxolTM. During this study, we have found the
intramolecular Nozaki–Hiyama reaction of allylic phos-
phates is a good method for constructing seven- and
eight-membered carbocyclic rings. Herein we report
the results obtained so far.
Scheme 2. Reagents and conditions: (a) LDA, THF, 0°C, 15min, then
PivOCH2I, À78 to 0°C, 1h; (b) DBU, CH2Cl2, 12h, 84% (two steps);
(c) CeCl3Æ7H2O, NaBH4, MeOH, 5min, 87%; (d) Py, SOCl2, Et2O, 0°C
to rt, 1h, 87%; (e) MPMONa, TBAI, THF/DMF, 10h, 94%; (f) DDQ,
CH2Cl2/t-BuOH/potassium phosphate buffer pKa = 8 (KPB 8), 1h,
90%; (g) cat. PTSA, acetone/H2O, 50°C, 2d, 96%; (h) TBSCl,
imidazole, CH2Cl2, 9h, 98%; (i) TrisNHNH2, THF, 2d, 87%.
We have studied the construction of the eight-membered
ring in TaxolTM model compound 1, whose retrosynthetic
analysis is shown in Scheme 1. Since the intramolecular
Nozaki–Hiyama allylation of allylic phosphate 2 was
expected to provide 1, and 2 was envisioned to be prep-
ared from trisyl hydrazone 3 and aldehyde 4 via Shapiro
reaction (Scheme 1). Aldehyde 4 would be prepared
from the known compound.5
Preparation of the trisyl hydrazone 3 is shown in
Scheme 2. The known compound 56 was alkylated by
LDA and iodomethyl pivalate, followed by the b-elimi-
nation induced with DBU to afford a,b-unsaturated
ketone 6 (84%, two steps).7 Luche reduction of 6
(87%),8 subsequent transformation to the corresponding
chloride (87%), and the following reaction with sodium
p-methoxyphenylmethoxide afforded 7 (94%). MPM
group of 7 was removed by DDQ (90%), and the dioxo-
lane was hydrolyzed with p-TsOH in aqueous acetone to
afford 8 (96%). Alcohol 8 was protected as a TBS ether,
and the following condensation with trisyl hydrazine
gave trisyl hydrazone 3 (87%).
Preparation of aldehyde 13 is shown in Scheme 3. Trisyl
hydrazone 3 was converted to the corresponding alke-
nyllithium by n-butyllithium (2equiv), and which was
reacted with aldehyde 4 to produce 9 as a mixture of dia-
stereomers (91%, 6:1). The diastereoselective epoxida-
tion of the isolated major diastereomer with TBHP
Scheme 3. Reagents and conditions: (a) n-BuLi, THF, À78 to 0°C,
then 4, À78 to 0°C, 1h, 91% (6:1); (b) cat. VO(acac)2, TBHP, toluene,
0°C, 0.5h, 94%; (c) 1N LiAlH4 (in Et2O), Al(O–i-Pr)3, Et2O, 33h (10)
50% (11) 44%; (d) (Cl3CO)2CO, Py, CH2Cl2, À78°C to rt, 10h, quant.;
(e) THF/2N HCl, rt, 40min, 80%; (f) (EtO)2P(O)Cl, Py, CH2Cl2, 3h,
74%; (g) DDQ, CH2Cl2/t-BuOH/KPB 7, 2h, 91%; (h) Dess–Martin
periodinane, CH2Cl2, 1h, 95%; (i) LDA, (methoxymethyl)triphenyl
phosphonium chloride, À78°C to rt, 2h, 50%; (j) PTSA, CH2Cl2, 0.5h,
quant.
and VO(acac)2,5 followed by LiAlH4 reduction in the
presence of Al(O–i-Pr)3 to give diol 10 (50%)9 along with
triol 11 (44%). Diol 10 was converted to the cyclic car-
bonate with triphosgene (quant.), followed by the treat-
ment with diluted hydrochloric acid to produce 12
(80%), and triol 11 was independently converted to 12
by the treatment with triphosgene (quant.). Reaction
of 12 with chlorodiethylphosphate (74%), removal of
MPM with DDQ (91%), and the following Dess–Martin
oxidation generated aldehyde 13 (95%). Wittig reaction
of 13 (50%) and the following hydrolysis of the alkenyl
Scheme 1.

M. Iwamoto et al. / Tetrahedron Letters 45 (2004) 8653–8657
8655
ether under acidic condition afforded allylic phosphate 2
(quant.).
With allylic phosphate 2 in hand, the intramolecular
Nozaki–Hiyama reaction was carried out, but no reac-
tion occurred in the absence of LiI4c (Table 1, entry
2). Although the reaction in the presence of LiI at room
temperature only produced protonated product 16
(entry 3), a cyclized product 17 formed in 15% yield at
60°C (entry 4).10 In other solvents, DMF, DMSO, and
DMPU, a complicated mixture of undesired products
formed, and no product was obtained. We also prepared
allylic chloride 1511 from 14 as shown in Scheme 4, and
subjected to this reaction, however, surprisingly no
product was obtained (entry 1), and the starting material
was recovered. The reaction of 2 in DME at 70°C pro-
ceeded smoothly, and the starting material disappeared
after 1h to afford the desired product 1712 (11%) along
with 16 (39%)13 (entry 5).
Scheme 5.
and a diastereomeric mixture of seven-membered prod-
ucts 1812 was obtained in 88% yield (6.3:1) (Scheme 5).
Reaction of the corresponding acetonide 19 also af-
forded a diastereomeric mixture of seven-membered
products 20 in 85% yield, but interestingly the ratio of
diastereomers changed (3.3:1).
To examine the generality of this intramolecular
Nozaki–Hiyama allylation, we applied this reaction to
other substrates. First, 13 was subjected to this reaction,
Table 1. Intramolecular Nozaki–Hiyama reaction of 15 and 2
We succeeded thus in the synthesis of 6-8-6 tricyclic
compound 17, 6-7-6 tricyclic compounds 18 and 20. This
is a first synthesis of a cyclic compound by the intramo-
lecular Nozaki–Hiyama allylation using the allylic
phosphate.14
Entry Substrate LiI
(equiv)
Solvent Temp
(°C)
Yielda
16 (%) 17 (%)
1b
2b
3c
4
15
2
1.0
—
THF
THF
THF
THF
DME
60
60
Rt
60
70
—
—
30
60
39
—
—
—
15
11
2
1.0
1.0
1.0
2
5d
2
a Isolated yield.
b Starting material was recovered.
c Starting material (45%) was recovered.
d Reaction time was 1h. Unidentified products also formed.
Scheme 4. Reagents and conditions: (a) 1N H2SO4, THF, 1h, 98%; (b)
Dess–Martin periodinane, CH2Cl2, 1h, quant.; (c) cat. PTSA,
CH(OMe)3, CH2Cl2, MeOH, 10h, 89%; (d) DDQ, CH2Cl2, t-BuOH,
KPB 7, 88%; (e) LiCl, 2,6-lutidine, MsCl, DMF, 3h, 91%; (f) 2N HCl,
THF, 1h, 85%.
Scheme 6.

8656
M. Iwamoto et al. / Tetrahedron Letters 45 (2004) 8653–8657
Next, we examined the intramolecular Nozaki–Hiyama
allylation of rather simple substrates 21, 23, 25, and 27
(Scheme 6).15 As shown in Scheme 6, 21 and 23 gave
the seven-membered products 22 (80%, a single iso-
mer)12 and 24 (34%),16 respectively; furthermore, 25
and 27 gave the eight-membered products 26
(48%)12,17 and 28 (31%),18 respectively.
natural products. Further studies on the intramolecular
Nozaki–Hiyama reaction of some other allylic phos-
phates are now in progress.
Acknowledgements
This work was financially supported in part by Waseda
University Grant for Special Research Projects. We are
also indebted to 21COE ꢀPractical Nano-Chemistryꢁ.
Reactions of 21, 23, 25, and 27 suggest that the sub-
strates affording the product with a vinyl group on its
formed ring, that is, 21 and 25, give the better yield.
Considering the Nozaki–Hiyama allylation proceeds
via a six-membered transition state, this tendency is
probably because the chromium reagent derived from
23 would cyclize via the more strained 6-7 fused transi-
tion state in contrast with that of 21; hence, the yield of
24 was lower than that of 22. The difference in yield in
the reactions of 25 and 27 could be explained by the
same assumption.
References and notes
1. For recent reviews of synthesis of medium-sized rings, see:
(a) Yet, L. Chem. Rev. 2000, 100, 2963–3007; (b) Mehta,
G.; Singh, V. Chem. Rev. 1999, 99, 881–930.
2. (a) Okude, Y.; Hirano, S.; Hiyama, T.; Nozaki, H. J. Am.
Chem. Soc. 1977, 99, 3179–3181; (b) Hiyama, T.; Okude,
Y.; Kimura, K.; Nozaki, H. Bull. Chem. Soc. Jpn. 1982,
55, 561–568; For recent reviews of organochromium
The intramolecular Nozaki–Hiyama allylation using the
allylic phosphate affording the eight-membered ring in
excellent yield is shown in Scheme 7. Phosphate 30
was prepared from 29,19 and was subjected to this reac-
tion. The substrate 30 is the same type of allylic phos-
phate as 27, however, the reaction of 30 completed at
room temperature within 1h to afford 3112 as a sole
product (91%). This result could suggest that the con-
formational requirement is critical for good yielding in
this reaction.
reagents, see: (b) Furstner, A. Chem. Rev. 1999, 99, 991–
1045.
¨
3. The Cr(II)-mediated intramolecular addition reaction
involving allylic bromide to form medium-sized or macro-
cyclic ring, see: (a) Still, W. C.; Mobilio, D. J. Org. Chem.
1983, 48, 4785–4786; (b) Shibuya, H.; Ohashi, K.; Kawa-
shima, K.; Hori, K.; Murakami, N.; Kitagawa, I. Chem.
Lett. 1986, 85–86; (c) Paquette, L. A.; Rayner, C. M.;
Doherty, A. M. J. Am. Chem. Soc. 1990, 112, 4078–4079;
(d) Wender, P. A.; McKinney, J. A.; Mukai, C. J. Am.
Chem. Soc. 1990, 112, 5369–5370; (e) Wender, P. A.;
Grissom, J. W.; Hoffmann, U.; Mah, R. Tetrahedron Lett.
1990, 31, 6605–6608; (f) Rayner, C. M.; Astles, P. C.;
Paquette, L. A. J. Am. Chem. Soc. 1992, 114, 3926–3936;
(g) Astles, P. C.; Paquette, L. A. Synlett 1992, 444–446; (h)
Paquette, L. A.; Astles, P. C. J. Org. Chem. 1993, 58, 165–
169; The Cr(II)-mediated intramolecular reaction involv-
ing the allylic mesylate to form eight-membered ring, see:
(i) Kato, N.; Tanaka, S.; Takeshita, H. Chem. Lett. 1986,
1989–1992; (j) Kato, N.; Tanaka, S.; Takeshita, H. Bull.
Chem. Soc. Jpn. 1988, 3231–3237.
4. For the use of allylic phosphates, see: (a) Takai, K.;
Nozaki, H. Abstracts of the 4th ICOS at Tokyo, B-II-
2302, 1982; (b) Jubert, C.; Nowotny, S.; Kornemann, D.;
Antes, I.; Tucker, C. E.; Knochel, P. J. Org. Chem. 1992,
57, 6384–6386; (c) Nowotny, S.; Tucker, C. E.; Jubert, C.;
Knochel, P. J. Org. Chem. 1995, 60, 2762–2772.
5. Nicolaou, K. C.; Liu, J.-J.; Yang, Z.; Ueno, H.; Sorensen,
E. J.; Claiborne, C. F.; Guy, R. K.; Hwang, C.-K.;
Nakada, M.; Nantermet, P. G. J. Am. Chem. Soc. 1995,
117, 634–644.
In summary, we have developed the intramolecular
Nozaki–Hiyama reaction of the allylic phosphates to
afford the compounds containing the seven- or eight-
membered carbocyclic ring. The yield greatly depends
on the structure of substrate, however, rather complex
substrates, 13, 19, and 30, afforded products containing
seven- or eight-membered carbon skeleton in high to
excellent yield. The allylic phosphate was stable under
some reaction conditions; therefore, the intramolecular
Nozaki–Hiyama reaction of the allylic phosphates
would be a useful method in the synthesis of complex
6. Shibuya, S.; Isobe, M. Tetrahedron 1998, 54, 6677–6698.
7. This method requires two steps, but could be applied to a
variety of ketones, so we are now investigating its scope
and limitation. For preparation of PivOCH2I, see: Bodor,
N.; Solan, K. B.; Kaminiski, J. J.; Shin, C.; Pogany, S. J.
Org. Chem. 1983, 48, 5280–5284.
8. (a) Luche, J. L. J. Am. Chem. Soc. 1978, 100, 2226–2227;
(b) Gemal, A. L.; Luche, J. L. J. Am. Chem. Soc. 1981,
103, 5454–5459.
9. No reduction proceeded in the absence of Al (O–i-Pr)3.
10. Prolonged reaction time (3days) did not improve the yield
of 17 (14%, with formation of 16 (45%)).
Scheme 7. Reagents and conditions: (a) TBAF, THF, 50°C, 1.5d,
99%; (b) (EtO)2P(O)Cl, Py, CH2Cl2, 1d, 62%; (c) excess PPTS, MeOH,
0°C, 0.5h, 97%; (d) Dess–Martin periodinane, CH2Cl2, 1h, 97%.
11. Preparation of the chloride 15 from 12 was attempted, but
the yield was very low; hence, 14 was prepared through a
different route.

M. Iwamoto et al. / Tetrahedron Letters 45 (2004) 8653–8657
8657
12. Relative configuration was elucidated by NOE
experiment.
16. By-products were shown below. The intermolecular reac-
tion was neglected even under diluted conditions
(0.002M).
13. Since we were interested in the formation of this proto-
nated product 16, the reaction was carried out under the
same conditions as in entry 4and quenched with D 2O. The
result of this experiment showed no incorporation of
deuterium into the product; hence, the formed allylic
chromium was surmised to react with water existing
somewhere in the reaction mixture to afford 16.
14. Now we are studying the ring-enlargement reaction of the
seven-membered rings in 18 and 20 to the eight-membered
rings to construct the 6-8-6 ring system required for the
synthesis of TaxolTM.
15. Allylic phosphate 21 was synthesized from malonic acid
dimethyl ester in six steps as follows: (i) I(CH2)4OTBS,
NaH, THF, rt, 55%; (ii) 1-(4-bromo-2-butenyloxymethyl)-
4-methoxy-benzene, NaH, THF, rt, 82%; (iii) DDQ,
CH2Cl2/t-BuOH/KPB 7, rt; (iv) (EtO)2P(O)Cl, Py,
CH2Cl2, rt, 66% (two steps); (v) 2N HCl/THF; (vi) Dess–
Martin periodinane, CH2Cl2, 82% (two steps). Allylic
phosphates 23, 25, and 27 were synthesized similarly.
17. Configuration of 26 shown in Scheme 6 is that of the
major diastereomer.
18. By-products possessing similar structure to 32 and 33 were
obtained in 12% each.
19. Compound 29 was prepared according to the method in
Ref. 5.