Palladium(0)-catalyzed cascade one-pot synthesis of isoxazolidines

Article abstract of DOI:10.1016/j.tetlet.2005.11.148

A highly diastereoselective cascade reaction protocol has been developed for the synthesis of isoxazolidine derivatives utilizing aryl halides, O-homoallyl hydroxylamine and palladium(0) in a one-pot reaction.

Full text of DOI:10.1016/j.tetlet.2005.11.148

Tetrahedron
Letters
Tetrahedron Letters 47 (2006) 927–930
Palladium(0)-catalyzed cascade one-pot synthesis of isoxazolidines
Krishna Gopal Dongol* and Boon Ying Tay
Institute of Chemical and Engineering Sciences, 1 Pesek Road, Jurong Island, Singapore 627833, Singapore
Received 29 September 2005; revised 25 November 2005; accepted 30 November 2005
Available online 20 December 2005
Abstract—A highly diastereoselective cascade reaction protocol has been developed for the synthesis of isoxazolidine derivatives
utilizing aryl halides, O-homoallyl hydroxylamine and palladium(0) in a one-pot reaction.
Ó 2005 Elsevier Ltd. All rights reserved.
The use of palladium-catalyzed cascade reactions in the
synthesis of heterocyclic compounds of biological inter-
est continues to receive widespread attention.1 Recently,
heterocyclic rings such as tetrahydrofurans2 and pyrrol-
idines3 were synthesized with good to excellent stereo-
selectivity by the cascade method, employing a suitable
Pd0 catalyst and the requisite aryl halides and alkenes.
There are no reports, however, describing the use of a
palladium-mediated cascade sequence for substrates
containing an alkene with tethered nitrogen, in which
the construction of both C–N and C–C bonds could
conceivably be achieved in a one-pot event.4 We decided
to investigate this area using functionalized O-homoallyl
hydroxylamines such as 4 where the N–O bond would
serve as a tether in facilitating an intramolecular pro-
cess. This chemistry could potentially offer a convenient
entry to 3,5-disubstituted isoxazolidines, which serve as
valuable synthetic building blocks for the construction
of bioactive natural products5,6 and as precursors for
b-amino ketones.7
To examine this hypothesis, a series of N-Boc protected
hydroxylamines 4 were prepared from the correspond-
ing alcohols 18 via the Mitsunobu protocol.9 O-Homo-
allylic alcohol 1 was treated with N-hydroxyphthalimide,
triphenylphosphine (PPh3) and diethyl azodicarboxylate
(DEAD) in THF furnishing hydroxylamine ether 2. The
phthaloyl moiety (Phth) was then cleaved by treatment
with hydrazine hydrate at 0 °C to room temperature,
leading to O-homoallyl hydroxylamine 3.10 Protection
NH2
NPhth
O
O
N
-hydroxylphthalimide
OH
1
hydrazine hydrate
R
R
R
DEAD, PPh3,
0 °C - rt, DCM
3
2
THF, 0 °C - rt
di-tert-butyl dicarbonate
NaOH/DCM
ArI,
Boc
R
Yield
Boc
N
Pd2(dba)3
NH
Ph
5a Ph,
5b
5c 2-furyl
80%
O
O
NaOtBu
p-MeOC6H4, 84%
R
R
76%
dppe, toluene
reflux
5
4
Scheme 1.
Keywords: Cascade reaction; Palladium(0); Isoxazolidine; O-Homoallyl hydroxylamine.
*
Corresponding author. Tel.: +65 6796 3917; fax: +65 6316 6184; e-mail: Krishna_dongol@ices.a-star.edu.sg
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.11.148

928
K. G. Dongol, B. Y. Tay / Tetrahedron Letters 47 (2006) 927–930
Boc
NH
Boc
NH
R'I,
Pd2(dba)3,
P-ligand
Boc
N
O
O
O
R'
+
R'
R
R
R
Cs2CO3
toluene, 20 h
reflux
6
7
4
c
d
R = 2-furyl
R = Me
R' = Ph
R' = Ph
a
b
R = Ph
R' = Ph
R =
p-MeOC6H4 R' = Ph
Scheme 2. Substituent effects.
Table 1. Substituent effects
Entry
R
R0
GC yielda (%)
Cis/transb ratio of 6
6
7
a
b
c
d
e
f
Ph
Ph
Ph
Ph
Ph
52
64
74
11
8
48
36
26
89
92
39
6:1
10:1
3:1
—
—
2:1
p-MeOC6H4
2-Furyl
CH3
p-MeOC6H4
p-MeOC6H4
p-MeOC6H4
p-O2NC6H4
61
a The ratio (%) was based on gas chromatographic analysis.
b The cis/trans ratio of 6 was based on 1H NMR analysis. In the case of entries 4 and 5, we are unable to isolate pure cyclized compounds 6.
of 3 as its N-Boc carbamate then completed the synthe-
sis of 4, setting the stage for the investigation of palla-
dium-mediated isoxazolidine formation (Scheme 1).
Unfortunately, no improvement in product distribution
was observed when the catalyst loading was varied from
1 to 20 mol %.
In the initial studies, an attempted Pd2(dba)3-catalyzed
reaction between N-Boc hydroxylamine ether 4 and iodo-
benzene in the presence of NaOtBu failed to provide
the desired isoxazolidine 6. Instead, amide 5 resulting
from direct amination of iodobenzene with 4 was
observed as the sole product. A possible explanation is
that the strong base, NaOtBu, abstracts the amide
N-hydrogen and thus facilitates N-arylation.11 To
circumvent the formation of this undesired product,
Cs2CO3 was employed as an alternative,12 and under
this modified procedure, isoxazolidine 6b, for example,
was obtained in 60% isolated yield, together with the
Heck coupling adduct 7b, in 21% yield as shown in
Scheme 2.
However, the choice of the phosphine ligand does play a
role in the observed product ratio. In the presence of
P(Cy)3 or P(o-Tol)3, an approximately 3:1 ratio in
favour of isoxazolidine 6b was observed. In contrast, a
reversed product distribution favouring the Heck adduct
7b was found when P(tBu)3 was employed. These results
are summarized in Table 2 in reference to the reaction in
Scheme 2. The best result for the isoxazolidine 6b over
Heck type product 7b was obtained when 1 mol% P(o-
Tol)3 was used.
A mechanistic rationale for the observed products is
summarized in Figure 1. The first part of the catalytic
cycle is believed to proceed through the standard Heck
pathway, where an initial oxidative insertion of Pd0 into
the aryl halide R0X is followed by syn carbopalladation
across the terminal olefin to afford intermediate 9.
To determine the scope of this reaction, a series of
reactions were carried out with a variety of R-substi-
tuted N-Boc-hydroxylamine ethers (Ph, p-MeOC6H4,
Me, 2-furyl) and aryl iodides (R0 = p-MeOC6H4, Ph,
p-O2NC6H4).
The relative rates of the subsequent b-hydride elimina-
tion versus nitrogen coordination in intermediate 9
The product distribution is summarized in Table 1 in
reference to Scheme 2.
Table 2. Ligand effects
Entry
P-ligands
GC yielda (%)
It was evident that the substituents R and R0 influenced
the ratio of products obtained to some degree. In gen-
eral, when R was electron donating in nature, the cyclic
product 6 was favoured, while if R0 was an electron
donating group the Heck type product 7 was preferred.
6b
7b
1
2
3
4
dppe
52
72
79
30
48
28
21
70
P(Cy)3
P(o-Tol)3
P(tBu)3
R = p-MeOC6H4 and R0 = Ph.
a The product ratio and conversion was based on gas chromatographic
(GC) analysis based on 100% conversion.
In order to suppress the formation of the undesired Heck
product 7, the effect of catalyst loading was examined.

K. G. Dongol, B. Y. Tay / Tetrahedron Letters 47 (2006) 927–930
929
Acknowledgements
Boc
N
O
R'
The authors thank the Agency for Science, Technology
and Research (A*STAR) of Singapore under project
code ICES 03-142001, for financial support. The authors
also wish to thank Prof. Holger Butenschoen, University
of Hannover, Germany, for his helpful suggestions and
Dr. Felicity K. Moore, for checking the English in this
manuscript.
R'X
Pd0
R
6
HX
Pd-N insertion/
reductive
oxidative
addition
elimination
H
N
Boc
Pd
R'PdIIX
8
O
R'
R
References and notes
9
β-elimination
Boc
NH
1. Balme, G.; Bossharth, E.; Monteiro, N. Eur. J. Org.
Chem. 2003, 4101–4111.
O
2. Wolfe, J. P.; Rossi, M. A. J. Am. Chem. Soc 2004, 126,
1620–1621.
3. Ney, J. E.; Wolfe, J. P. Angew. Chem., Int. Ed. 2004, 43,
3605–3608.
Boc
NH
R'
R
7
O
R
4
4. (a) Hosokawa, T.; Murahashi, S.-I. In Handbook of
Organopalladium Chemistry for Organic Synthesis; Negi-
shi, E.-I., Ed.; Wiley and Sons, 2002; pp 2129–2169, 2211–
2225; (b) Tsuji, J. Palladium Reagents and Catalysts:
Innovations in Organic Synthesis; Wiley and Sons: Chich-
ester, 1995, pp 510–527; (c) Aftab, T.; Grigg, R.; Ladlow,
M.; Sridharan, V.; Thornton-Pett, M. Chem. Commun.
2002, 1754–1755; (d) Grigg, R.; Markandu, J. Tetrahedron
Lett. 1991, 32, 279–282; (e) Frederickson, M.; Grigg, R.;
Markandu, J.; Thornton-Pett, M.; Redpath, J. Tetra-
hedron 1997, 53, 15051–15060.
5. (a) Ravi Kumar, R. K.; Mallesha, H.; Ranagappa, B. K.
S. Eur. J. Med. Chem. 2003, 1–7; (b) Haken, P. T.; Webb,
S. B. GB Pat. 2024218, 1980; Chem. Abstr. 1980, 93,
132471; Saber, S. H.; Zhang, L. Szapacs, E. M.; Quinn,
J. A. EU Pat. 1035122 A1, 2000; Chem. Abstr. 2000, 133,
222723.
R' = p-MeOC6H4, p-O2NC6H4, Ph
Figure 1. Proposed catalytic cycle leading to 6 and 7.
determine the observed product distribution. Moreover,
the carbonyl oxygen of the N-Boc group may also offer
additional stabilization through its participation via
intramolecular chelation, thereby stabilizing intermedi-
ate 9 and suppressing b-hydride elimination. The pres-
ence of electron-donating phosphine ligands on the Pd
will accelerate C–N bond formation by accelerating
the reductive elimination,13 hence, the use of P(Cy)3 or
P(o-Tol)3 favours product 6 over b-hydride elimination
product 7.
6. Kasahara, K.; Iida, H.; Kibayashi, C. J. Org. Chem. 1989,
54, 2225–2233.
A 10:1 the cis/trans ratio14 of the 3,5-substituents in
isoxazolidine 6b was established by integration of the
1H NMR spectrum. The stereochemistry of the major
isomer of 6b was determined by NOE and NOESY15
studies, and is believed to arise from transition state A
as depicted in Figure 2.
7. Murahashi, S.-I.; Kodera, Y.; Hosomi, T. Tetrahedron
Lett. 1988, 29, 5949–5952.
8. Petrier, C.; Luche, J.-L. J. Org. Chem. 1985, 50, 910–912.
9. Bates, R. W.; Sa-Ei, K. Org. Lett. 2002, 4, 4225–4227.
10. Janza, B.; Studer, A. Synthesis 2002, 2117–2123.
11. Louie, J.; Hartwig, J. Tetrahedron Lett. 1995, 36, 3609–
3612.
12. Representative experimental procedure: A 50 mL Schlenk
flask was charged with 3.1 mg Pd2(dba)3 (1 mol %,
0.003 mmol), 145 mg Cs2CO3 (1.3 equiv, 0.44 mmol),
1.4 mg (1 mol %, 0.0034 mmol) 1,2-bis(diphenylphosph-
ino)ethane (dppe) using a glove box. The Schlenk flask
was then evacuated and back filled with argon three times.
Toluene (5 mL) and 100 mg of N-Boc hydroxylamine 4b
(0.34 mmol) and 90 mg (1.3 equiv, 0.44 mmol) iodoben-
zene were added to the flask under argon, then the
reaction mixture was refluxed for 20 h. The mixture was
filtered through a pad of silica gel and washed with ethyl
acetate. Evaporation of the solvent provided the crude
product which was further purified by flash column
A rationale for the cis selectivity can be gained by exam-
ining the reactive conformations (Fig. 2). Chelation of
PdII with the nitrogen lone pair is favoured in conforma-
tion A as compared to conformation B, leading to the
formation of the cis-diastereomer as the major product.
In summary, a novel, cascade sequence leading to the
formation of isoxazolidines 6 has been developed. A sys-
tematic study of substituent effects and the application
of this route to natural product synthesis is currently
underway.
Ar'
Boc
N
PdII
Boc
Ar'
O
N
Ar
PdII
O
Ar
N
H
Ar'
O
N
H
Ar'
O
Boc
Ar
Boc
Ar
B
A
Figure 2. Conformations leading to the cis selectivity.

930
K. G. Dongol, B. Y. Tay / Tetrahedron Letters 47 (2006) 927–930
chromatography which afforded isoxazolidine 6b (73 mg,
0.2 mmol, 60%) as a colourless oil and Heck product 7b
(27 mg, 0.07 mmol, 21%) as a colourless oil.
128.7, 128.4, 127.1, 126.1, 125.4, 113.9, 86.7, 81.6, 55.3,
38.6, 28.2. EI MS: m/z (70 eV) 369 (2), 279 (2), 252 (13),
236 (89), 205 (57), 178 (63), 152 (42), 135 (100, M+1), 121
(77), 91 (67), 56 (33), 41 (57). HRMS: C22H27NO4: calcd
369.1943, found 369.1947.
Compound 6b: IR (neat): m 3314, 2977, 2934, 1732, 1683,
1615, 1515, 1440, 1367, 1305, 1250, 1173, 1035, 830 cmꢀ1
.
1H NMR (400 MHz, CDCl3): d 7.17 (m, 7H), 6.80 (d, 2H,
J 8.4 Hz) 4.74 (dd, 1H, J 6.4, 3.6 Hz), 4.44 (m, 1H), 3.73 (s,
3H, OCH3), 3.11 (dd, 1H, J 6.4, 7.2 Hz), 2.77 (dd, 1H, J
7.6, 6.0 Hz), 2.54 (m, 1H), 1.94 (m, 1H), 1.37 (s, 9H). 13C
NMR (100.4 MHz, CDCl3): d 159.7, 156.5, 137.1, 128.5,
128.1, 127.5, 127.1, 125.4, 112.9, 81.8, 80.7, 60.9, 54.3,
41.5, 41.2, 27.2. EIMS: m/z (70 eV) 369 (2), 252 (11), 236
(39), 205 (15), 178 (78), 162 (37), 135 (100), 119 (80), 91
(79), 57 (47), 41 (61). HRMS C22H27NO4: calcd 369.1943,
found 369.1943.
13. (a) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1995,
117, 4708–4709; (b) Driver, M. S.; Hartwig, J. J. F. J. Am.
Chem. Soc. 1997, 119, 8232–8245.
14. The cis/trans ratio of isoxazolidine 6b was determined as
1
10:1 from H NMR integration of the OMe resonance at
d = 3.73 ppm for major isomer and d = 3.79 ppm for
minor isomer.
15. The NOESY experimental data showed a distinct corre-
lation between the protons H-3 and H-5.
Compound 7b: IR (neat): m 3307, 2978, 2933, 1760, 1612,
Boc
1367, 1248, 1172, 1033, 830 cmꢀ1 1H NMR (400 MHz,
.
Boc
N
O
H
N
O
H
CDCl3): d 7.20–7.13 (m, 7H), 6.84 (br s, 1H, NH) 6.83 (d,
2H, J 8.2 Hz), 6.37 (d, 1H, J 15.8 Hz), 6.05 (dt, 1H, J 15.8,
7.2 Hz ), 4.74 (t, 1H, J 6.9 Hz), 3.70 (s, 3H, OCH3), 2.83
(m, 1H), 2.59 (m, 1H), 1.37 (s, 9H). 13C NMR
(100.4 MHz, CDCl3): d 159.6, 156.5, 137.4, 132.5, 131.6,
H
H
H
H
H
H
H
H
H
H
O
O
1H-noesy
noe