Radical azidonation of benzylic positions with iodoninm azide

Article abstract of DOI:10.1002/1521-3773(20010202)40:3<623::AID-ANIE623>3.0.CO;2-G

Introduction of an azido substituent at the α position of benzyl ethers can be achieved by treating them with IN₃ in refluxing acetonitrile. Some of the products obtained after 20 min - 5 h are given.

Full text of DOI:10.1002/1521-3773(20010202)40:3<623::AID-ANIE623>3.0.CO;2-G

COMMUNICATIONS
[10] Only allylmagnesium bromide successfully added to 1b (in Et2O at RT
for 7 h using 20 mol% of [Fe(acac)3]) and the corresponding allylated
product was obtained in quantitative yield.
[11] The produced vinyl anion seems to be unstable at room temperature.
Yield of 2b and its deuterium content were decreased to 90% and
82%-D, respectively when the reaction mixture was kept at ambient
temperature for 3 h before quenching with DCl/D2O, the reaction was
conducted under the same conditions as in Equation (5).
1) 5% [Fe(acac)3]
BuLi (3 equiv)
toluene
E
OBn
Bu
−20 °C, 4 h
OBn
OBn
(6)
2) E+
−20 °C - RT, 2 h
1b
3
HO
Bu
Ph
SiMe2H
[12] Ethyllithium and hexyllithium were also added to 1b to afford alkenes
in high yields (Et: 96%, Hex: 92%).
Bu
OBn
3a 73%
3b 83%
Et
Ph
HO
Bu
Et
HO
Bu
OBn
3c 80%
OBn
3d 69%
Radical Azidonation of Benzylic Positions with
Iodonium Azide**
Experimental Section
Christel Viuf and Mikael Bols*
Full experimental details are in the Supporting Information. A represen-
tative procedure is as follows: to a flask purged with argon, a solution of
[Fe(acac)3] in toluene (0.025m, 0.025 mmol, 1 mL), toluene (4 mL), and a
substrate 1b (0.5 mmol) were introduced. The flask was cooled to 408C
and a solution of butyllithium in hexane (1.5m, 1.5 mmol) was added to the
mixture. Reaction temperature was immediately raised to 208C and the
mixture was stirred at 208C for 2 h. The reaction was quenched with 1n
HCl. After the conventional workup and purification by chromatography
on silica gel, a pure product 2b was obtained in 97% yield.
We recently reported a procedure for the conversion of
benzyloxy alcohols like 1 into benzylidene derivatives like 2
with N-iodosuccinimide (NIS) and light (Scheme 1).[1] During
Received: September 15, 2000 [Z15812]
Scheme 1. The NIS-promoted transformation of benzyloxy alcohols into
[1] B. J. Wakefield in Organolithium Methods (Eds.: A. R. Katritzky, O.
Meth-Cohn, C. W. Rees), Academic Press, London, 1988.
[2] Zr: a) E. Negishi, T. Takahashi, Synthesis 1988, 1 ± 19; Ti: b) D. C.
Brown, S. A. Nichols, A. B. Gilpin, D. W. Thompson, J. Org. Chem.
1979, 44, 3457 ± 3461; Cu: c) J. F. Normant, A. Alexakis, Synthesis
1981, 841 ± 870; Zn: d) F. Bernadou, L. Miginiac, Tetrahedron Lett.
1976, 3083 ± 3086; e) M. Bellassoued, Y. Frangin, Synthesis 1978, 838 ±
840; Ni: f) B. B. Snider, R. S. E. Conn, M. Karras, Tetrahedron Lett.
1979, 1679 ± 1682; See also: g) P. Knochel in Comprehensive Organic
Synthesis, Vol. 4 (Eds.: B. M. Trost, I. Fleming, M. F. Semmelhack),
Pergamon, Oxford, 1991, pp. 865 ± 911.
[3] a) J. E. Mulvaney, Z. G. Gardlund, S. L. Gardlund, J. Am. Chem. Soc.
1963, 85, 3897; b) J. E. Mulvaney, D. J. Newton, J. Org. Chem. 1969, 34,
1936 ± 1939; c) W. Bauer, M. Feigel, G. Müller, P. von R. Schleyer, J.
Am. Chem. Soc. 1988, 110, 6033 ± 6046; d) L.-I. Olsson, A. Claesson,
Tetrahedron Lett. 1974, 2161 ± 2162.
benzylidene derivatives and the effect of NaN3 addition. Bn ˆ benzyl.
the investigation of this reaction we noticed that 1 was
transformed with excess NaN3 very rapidly into a variety of
products of which many had, according to mass spectrometry,
one to three hydrogen atoms replaced with azido groups.
When the analogous reaction was carried out on benzyl ethyl
ether 3 [Eq. (1), light substituted with heat to avoid decom-
position of azides] the monoazido derivative 4 was isolated in
65% yield together with small amounts of the by-products 5
and 6.[2]
[4] In the absence of a catalyst, a complex mixture without a normal
product 2b was produced.
[5] a) M. Tamura, J. K. Kochi, J. Am. Chem. Soc. 1971, 93, 1487 ± 1489;
b) S. M. Neumann, J. K. Kochi, J. Org. Chem. 1975, 40, 599 ± 606;
c) R. S. Smith, J. K. Kochi, J. Org. Chem. 1976, 41, 502 ± 509.
[6] Conditions for all reactions reported here are not fully optimized. A
sufficient amount of butyllithium per iron catalyst (5 mol%) seems to
be required during the reaction for the successful carbolithiation.
Therefore, 3 equivalents used in reactions are a large enough amount
for the present carbolithiation, and, indeed, a reaction of 1b with
2 equivalents of butyllithium gave essentially the same result (95% of
2b, cf. Table 1, entry 6).
[7] This product was possibly obtained by the addition of butyllithium to
pent-1-en-3-yne, produced by the b-elimination of phenylpropanol
from 1a. a) A. A. Petrov, V. A. Kormer, T. V. Yakovleva, Zh. Obshch.
Khim. 1960, 30, 2238; b) A. A. Petrov, Yu. I. Porfirꢁeva, V. A. Kormer,
Zh. Obshch. Khim. 1961, 31, 1518.
Since this azidonation reaction appeared faster than the
benzylidene derivative formation promoted by NIS alone and
since the red color of the reaction mixture indicated that IN3
was present, it was suggested that iodine azide was the reagent
[*] Dr. M. Bols, C. Viuf
University of Aarhus
Department of Chemistry
Langelandsgade 140, 8000 Aarhus C (Denmark)
Fax (‡45)86196199
[8] M. Nakamura, A. Hirai, E. Nakamura, J. Am. Chem. Soc. 2000, 122,
978 ± 979.
[9] BuMgBr was prepared in diethyl ether, and most of the solvent was
replaced with toluene before use.
E-mail: mb@kemi.aau.dk
[**] We thank the Danish National Research Council for financial support.
Angew. Chem. Int. Ed. 2001, 40, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4003-0623 $ 17.50+.50/0
623

COMMUNICATIONS
Table 1. Analytical data of the azido compounds 4, 8, 10, 12, and 14.[a]
causing the reaction. Indeed if IN3 was used, 3 could be
transformed to 4 in 75% yield (Scheme 2).
4: 1H NMR: d ˆ 7.25 ± 7.41 (m, 5H), 5.38 (s, 1H), 3.85 (dqv, 1H, J ˆ 7.0,
9.5 Hz), 3.58 (dqv, 1H, J ˆ 7.0, 9.5 Hz), 1.19 ± 1.24 (t, 3H, J ˆ 7.0 Hz)
8: 1H NMR: d ˆ 7.22 ± 7.38 (m, 5H), 5.24 (s, 1H), 3.43 (s, 3H)
10: 1H NMR (diasteromeric ratio 1:1): d ˆ 7.25 ± 7.50 (m, 10H), 5.59 (s,
1H), 5.42 (s, 1H), 4.93 (m, 2H), 3.82 (m, 2H), 2.11 (s, 3H), 2.30 ± 1.90 (m,
4H), 1.82 (s, 3H), 1.68 ± 1.84 (m, 4H), 1.25 ± 1.49 (m, 8H); MS (EI): m/z:
calcd (C15H19N3O3‡Na) 312.1324, found 312.1322
N3
IN3 (2 equiv)
O
O
O
O
MeCN
∆, 20 min
75%
12: 1H NMR (diasteromeric ratio 1:1): d ˆ 7.39 ± 7.53 (m, 10H), 5.72 (s,
1H,), 5.70 (s, 1H), 5.68 (s, 1H), 5.67 (s, 1H), 4.51 (t, 1H, J ˆ 4.4 Hz), 4.49 (t,
1H, J ˆ 4.4 Hz), 4.05 (br s, 2H), 3.72 (d, 2H, J ˆ 4.4), 3.70 (d, 2H, J ˆ 4.4),
3.48 (dd, 2H, J ˆ 3.7), 3.28 (br d, 1H, J ˆ 3.7), 3.16 (br d, 1H, J ˆ 3.7); MS
(EI): m/z: calcd (C13H13N3O4‡Na) 298.0804, found 298.0804
3
7
4
8
N3
IN3 (2 equiv)
MeCN
∆, 25 min
93%
1
14: H NMR (diasteromeric ratio 1:1):[b] d ˆ 7.40 ± 7.50 (m, 5H, 5H'), 5.64
(s, 1H'), 5.55 (s, 1H), 5.47 (d, 1H, J ˆ 3.50 Hz), 5.08 (d, 1H', J ˆ 3.50 Hz),
5.01 (dd, 1H, J ˆ 3.50, J ˆ 8.1 Hz), 4.91 (dd, 1H', J ˆ 3.50 Hz, J ˆ 7.7 Hz),
4.31 ± 4.46 (m, 1H, 2H'), 4.17 ± 4.25 (m, 1H, 1H'), 4.09 (d, 1H'), 3.99 (d, 1H,
J ˆ 13.6 Hz), 3.83 (dd, 1H, J ˆ 13.6, 2.9 Hz), 2.15 (s, 3H), 2.03 (s, 3H'), 1.56
(s, 3H), 1.53 (s, 3H'), 1.37 (s, 3H'), 1.20 (s, 3H); MS (EI): m/z: calcd
(C17H21N3O6‡Na) 386.1328, found 386.1329
N3
Ph
IN3 (2 equiv)
O
O
Ph
MeCN
∆, 1 h
98%
OAc
OAc
[a] 200 MHz for all 1H NMR spectra, in CDCl3. [b] Primed protons refer to
the second diasteromer.
9
10
O
O
O
O
O
because it is more electron withdrawing than a regular ether
unit.
IN3 (3 equiv)
MeCN
O
Strong support was found for the reaction to follow a
radical mechanism: Compound 8 was not formed when 7 was
allowed to react in the presence of the radical trap N-tert-
butyl-a-phenylnitrone (Scheme 2).[8] Substrate 7 and most of
the radical trap (2 equiv) were recovered unconverted, which
shows that capture of a small amount of radicals prevented the
reaction. A certain electron-donating capacity is necessary at
the benzylic position for the reaction to occur: When an ester
is attached to this position as in benzyl benzoate no reaction is
observed. Benzyl alcohol, on the other hand, is oxidized to
benzaldehyde under the reaction conditions. Based on these
results we suggest the following reaction course: Homolysis of
IN3 results in an azide radical that abstracts a hydrogen atom
from the substrate to form a benzylic radical. This radical is
oxidized by IN3 or I2 to a benzylic cation that combines with
the azide ion to form the product.[9]
O
O
∆, 2 h
86%
N3
Ph
Ph
12
11
N3
Ph
OAc
O
O
Ph
IN3 (6 equiv)
O
O
OAc
MeCN
∆, 5 h
74%
O
O
O
O
13
14
Scheme 2. Reaction of various benzyl ethers with IN3 in MeCN.
Some related chemistry has been reported in a series of
papers by Magnus et al. They used PhIO/trimethylsilylazide
(TMSN3) at low temperature to azidonate a carbon atom a to
amino and enol ether functionalities.[3] It has also been
reported that PhI(OAc)2/NaN3 can be used to generate
carbon-centered radicals.[4] Hassner et al. have pioneered
and extensively studied the addition of IN3 and other
haloazides to alkenes.[5] They found that though the addition
of haloazides to alkenes normally follows an ionic mechanism,
in certain cases, in particular when ClN3 or BrN3 is the
reagent, the products of a radical addition to the alkene are
obtained.[5b] However, substitution of activated hydrogens by
azido groups with IN3 has not been carried out nor has the
introduction of an azido group in a benzylic position.
As can be seen from Scheme 2 the reaction is highly
efficient[6] when carried out in refluxing acetonitrile: The
yields of azides (Table 1) were 74 ± 98%. The transformations
9 !10, 11 !12, and 13 !14 show that the reaction tolerates
other functionalities such as ester, epoxide, or acetal units.
The reaction of 13 is more sluggish than the others requiring
more reagent and time. The reason is presumably that the
acetal functionality makes the benzylic position less reactive
Received: August 21, 2000
Revised: September 22, 2000 [Z15674]
[1] a) J. Madsen, M. Bols, Angew. Chem. 1998, 110, 3348 ± 3350; Angew.
Chem. Int. Ed. 1998, 37, 3177 ± 3178; b) J. Madsen, C. Viuf, M. Bols,
Chem. Eur. J. 2000, 6, 1140 ± 1146.
[2] The by-products 5 and 6 were presumably formed from 4 through light-
induced decomposition of the latter to a nitrene followed by H and Ph
migration, respectively.
[3] a) P. Magnus, J. Lacour, J. Am. Chem. Soc. 1992, 114, 767 ± 769; b) P.
Magnus, J. Lacour, W. Weber, J. Am. Chem. Soc. 1993, 115, 9347 ± 9348;
c) P. Magnus, J. Lacour, P. A. Evans, P. Rigollier, H. Tobler, J. Am.
Chem. Soc. 1998, 120, 12486 ± 12499; d) P. Magnus, J. Lacour, W.
Weber, Synthesis 1998, 547 ± 551.
[4] F. Fontana, F. Minisci, Y. M. Yan, L. Zhao, Tetrahedron Lett. 1993, 34,
2517 ± 2520.
[5] a) A. Hassner, Methoden Org. Chem. (Houben-Weyl) 4th ed. 1952 ± ,
Vol. E16a, 1990, pp. 1243 ± 1290; b) A. Hassner, Acc. Chem. Res. 1971,
4, 9 ± 16; c) F. W. Fowler, A. Hassner, L. A. Levy, J. Am. Chem. Soc.
1967, 89, 2077 ± 2082.
[7]
[6] General reaction conditions: A solution of IN3 was prepared by
mixing ICl (101 mg, 0.6 mmol) in MeCN (2 mL) with NaN3 (107 mg,
624
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4003-0624 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2001, 40, No. 3

COMMUNICATIONS
1.4 mmol) in MeCN (3 mL) at 108C.[5c] After 15 min the cooling bath
was removed, and the ether compound (0.3 mmol) was added. The
mixture was refluxed for between 0.4 and 5 h. Then CH2Cl2 (25 mL)
was added, and the mixture was washed with 5% Na2S2O3 solution
(20 mL). The dried (MgSO4) and evaporated organic phase was
purified by flash chromatography.
[9] We have also investigated some amine substrates in order to compare
IN3 with the Magnus reagent PhIO/TMSN3. Reaction of N,N-dimethy-
laniline with IN3 gave several products, the major one being 4-iodo-
N,N-dimethylaniline, while benzyldimethylamine led to a number of
very labile products that on decomposition gave benzaldehyde and
benzylmethylamine. This is in agreement with the formation of a-
azidoamines like those obtained in the Magnus reaction. Magnus et al.
could isolate such a-azidoamines by using nonaqeous work-up, but they
are clearly much less stable than the a-azidoethers.
[7] IN3 is an explosive substance, particularly as a solid (see ref. [5a]). We
encountered no problems, but took care to remove IN3 after the
reaction by washing with Na2S2O3 solution.
[8] K. Yelekci, X. Lu, R. B. Silverman, J. Am. Chem. Soc. 1989, 111, 1138 ±
1140.
Angew. Chem. Int. Ed. 2001, 40, No. 3
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4003-0625 $ 17.50+.50/0
625