Single-electron/pericyclic cascade for the synthesis of dienes

Article abstract of DOI:10.1002/anie.201403234

The highly efficient and diastereoselective synthesis of E dienes has been accomplished through radical cyclization of bromoallyl hydrazones. This methodology has been further extended to generate these products through a one-pot condensation/radical cyclization/cycloreversion cascade from simple aldehyde starting materials in high yields (>75 %) and high diastereoselectivities (>95:5). Mechanistic investigations suggest that the cascade reaction proceeds through a cyclic diazene intermediate prior to the cycloreversion.

Full text of DOI:10.1002/anie.201403234

.
Angewandte
Communications
DOI: 10.1002/anie.201403234
Synthetic Methods
Hot Paper
Single-Electron/Pericyclic Cascade for the Synthesis of Dienes**
Natalie E. Campbell and Glenn M. Sammis*
Abstract: The highly efficient and diastereoselective synthesis
of E dienes has been accomplished through radical cyclization
of bromoallyl hydrazones. This methodology has been further
extended to generate these products through a one-pot
condensation/radical cyclization/cycloreversion cascade from
simple aldehyde starting materials in high yields (> 75%) and
high diastereoselectivities (> 95:5). Mechanistic investigations
suggest that the cascade reaction proceeds through a cyclic
diazene intermediate prior to the cycloreversion.
hydrazones into cyclic diazenes,[10] which readily undergo
cycloreversions to access dienes in high diastereoselectivi-
ties.[11] Currently, there is only one method for the synthesis of
dienes from hydrazones, and it is limited to aryl hydrazones.[9]
A new and direct method for the conversion of hydrazines
into dienes would not only represent the first example of
accessing cyclic diazenes from hydrazones, but it also may
provide a solution for the unsolved challenge of directly
forming dienes from alkyl hydrazones.
We hypothesized that we may be able to rapidly access
cyclic diazenes from the corresponding sulfonylated hydra-
zone 1 (Scheme 1). Generation of either an anion or radical
S
ynthetic chemists have long exploited the ability to convert
hydrazone derivatives into a wide variety of highly reactive
diazo intermediates.[1] In the 1910s both Wolff[2] and Kishner[3]
used the base-mediated conversion of a hydrazone into an
alkyl diazene to effect an overall reduction (Figure 1).
Scheme 1. Proposed route to dienes from hydrazones.
using the vinyl halide would lead to a 6-endo cyclization to
form the cyclic hydrazide 2,[12,13] followed by elimination of
the sulfonamide to form the key cyclic diazene 3.[14] The
diazene 3 will then readily undergo a cycloreversion to form
the desired diene 4.[15] Each of these reactions will have low
activation energies, thus the entire process should proceed
during a single reaction step. The challenge with this route is
finding an appropriate method for facilitating the key 6-endo
cyclization. While 6-endo cyclizations of vinyl anions are
known,[12] an anionic strategy may lead to undesired enoliza-
tion or addition reactions. Furthermore, an anionic reaction
requires a geometrically pure vinyl halide,[16] which places
significant limitations on the availability of the key starting
material. A radical-based approach is an intriguing alterna-
tive. Hydrazones are known acceptors for vinyl radicals,[17]
and radicals avoid the complications of enolization[18] as well
as the synthesis of a diastereomerically pure precursor.[19]
We focused our studies on alkyl hydrazones which
currently cannot be readily converted into dienes
(Scheme 2). The aldehyde 5a was first condensed with
bromoallyl hydrazine 6a, which is readily prepared in
a single step from tosyl hydrazine.[20] The corresponding
hydrazone was then subjected to a refluxing solution of
tributyltin hydride and azobis(isobutyronitrile) (AIBN) to
afford the desired diene (4a) in 80% conversion along with
approximately 20% of debrominated hydrazone 7a. As 7a is
Figure 1. Representative diazo intermediates accessed from hydra-
zones. TBS=tert-butyldimethylsilyl.
Numerous other synthetic methodologies have also utilized
diazene intermediates, including vinyl diazenes in the Shapiro
reaction,[4] allyl diazenes in Hutchins[5] and Kabalka[6] reac-
tions, and alkyl silylated diazenes from Myers et al.[7] Hydra-
zones have also been used to access diazoalkanes in Bamford–
Stevens reactions[8] and allylated diazoalkanes in recent
examples by Thomson and co-workers.[9] One transformation
that is notably absent from the literature is the conversion of
[*] N. E. Campbell, Prof. G. M. Sammis
University of British Columbia, Department of Chemistry
2036 Main Mall, Vancouver V6T 1Z1 (Canada)
E-mail: gsammis@chem.ubc.ca
[**] This work was supported by the University of British Columbia
(UBC) and the Natural Sciences and Engineering Research Council
of Canada (NSERC).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201403234.
6228
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 6228 –6231

Angewandte
Chemie
Table 1: Substrate scope of diene formation from bromoallyl hydrazones.
Entry[a] R1
R2
R3
Product Yield [%][b] d.r. (E/Z)[c]
Scheme 2. Hydrazone formation and cascade diene synthesis starting
from the siloxy butyl derivative 5a. AIBN=azobis(isobutyronitrile),
THF=tetrahydrofuran, Ts=para-toluenesulfonyl.
1
2
3
H
CH3
CH3 CH3 4c
H
H
4a
4b
79
82
84
90:10
>95:5
90:10
4
5
6
H
CH3
CH3 CH3 4 f
H
H
4d
4e
94
84
79
>95:5
>95:5
>95:5
presumably formed from direct hydrogen transfer to the vinyl
radical, we next investigated slow addition of tributyltin
hydride to minimize this undesired pathway. Gratifyingly,
under these slow addition conditions, only 4a was observed by
1H NMR spectroscopy and it could be isolated in 79% yield
as a 90:10 mixture of E to Z isomers.[21,22]
For this methodology to be practical for the preparation of
substituted dienes, the synthesis of the requisite substituted
hydrazine must be highly efficient. We developed a rapid
synthetic route beginning with the known carboxylic acids 8b
and 8c (Scheme 3).[23] This readily available starting material
7
8
9
H
CH3
CH3 CH3 4i
H
H
4g
4h
82
79
86
>95:5
>95:5
>95:5
10
11
12
H
CH3
CH3 CH3 4l
H
H
4j
4k
80
82
67
>95:5
>95:5
90:10
13
14
15
H
CH3
H
H
4m
4n
76
72
74
>95:5
90:10
>95:5
CH3 CH3 4o
[a] Reactions were carried out on a >0.5 mmol scale. [b] Yield of the
mixture of diastereomers isolated after flash chromatography. [c] The
diastereomeric ratio was determined by 1H NMR spectroscopy of the
crude reaction mixture. Boc=tert-butoxycarbonyl, Trt=triphenylmethyl.
Scheme 3. Synthesis of tri- and tetrasubstituted bromoallyl hydrazines
6b and 6c. Ph=phenyl, TBAI=tetrabutylammonium iodide, THF=te-
trahydrofuran.
tuted diene 4l was slightly diminished, the mono- and
disubstituted dienes 4j and 4k were isolated with high
diastereoselectivity. Finally, we investigated the synthesis of
piperidyl dienes (entries 13–15) given the importance of
nitrogen-containing heterocycles to the pharmaceutical
industry.[26] Gratifyingly, the desired dienes could be synthe-
sized in good yields[27] and high diastereoselectivities.
To be competitive with modern methods for the synthesis
of dienes,[28] we next explored whether simple aldehydes
could be directly converted into the corresponding diene in
a single reaction pot. We began our studies with hydrazones
derived from citronellal (Scheme 4). Subjecting the aldehyde
5b to the one-pot, sequential procedure with bromoallyl tosyl
hydrazines 6a, 6b, and 6c afforded the dienes 4d, 4e, and 4 f,
respectively, in comparable yields and diastereoselectivities to
can be rapidly converted into the desired hydrazine in three
steps, which involve reduction of the carboxylic acid, Appel
conversion into the allyl bromide,[24] and subsequent substi-
tution using sodium tosyl hydrazinide to afford 6b and 6c in
74 and 43% yield, respectively.[25]
Investigations into the substrate scope commenced with
the examination of the syntheses of more highly-substituted
dienes (Table 1, entries 1–3). Subjecting both the resulting tri-
(entry 2) and tetrasubstituted bromoallyl hydrazones
(entry 3) to the optimized reaction conditions afforded the
di- (4b) and trisubstituted (4c) diene products in 82% and
84% yields, respectively, and are comparable to the yield of
the monosubstituted diene 4a (entry 1). We next investigated
increasing the steric bulk b to the hydrazone in citronellal
hydrazones (entries 4–6). With the exception of an increase in
yield for the monosubstituted diene 4d, the yields were
comparable for both the di- and trisubstituted alkenes (4e and
4 f). However, the diastereoselectivities observed were gen-
erally greater, with the E/Z ratios of all three greater than
95:5. Gratifyingly, no reaction was observed with the alkene
on citronellal, which would be susceptible under oxidation
reaction conditions.[9] Increasing the steric bulk a to the
hydrazone (entries 7–9) led to no observed loss of yield or
diastereoselectivity. Further increasing the steric bulk in the
a-position provided the corresponding diene in good yield
(entries 10–12). While the diastereoselectivity of trisubsti-
Scheme 4. One-pot hydrazone formation followed by one-electron/
pericyclic cascade for the synthesis of dienes 4d, 4e, and 4 f.
Angew. Chem. Int. Ed. 2014, 53, 6228 –6231
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 6229

.
Angewandte
Communications
Scheme 7. Isolation of the tin-bound cyclic diazo adduct 13.
Scheme 5. One-pot hydrazone formation followed by a one-electron/
pericyclic cascade for the synthesis of the diene 4j.
tin radical and 2. To investigate this possibility, we examined
the radical cascade using a low-molecular-weight aldehyde,
isovaleroaldehyde. This allows the facile removal of the diene
product and would leave hydrazone and hydrazine inter-
mediates. After cyclization of hydrazone 1p, in vacuo removal
of all non-nitrogen-containing products, and purification by
column chromatography, the tin-bound cyclic hydrazine 13
was isolated in 2.6% yield.[32,33]
After the radical 6-endo cyclization, there are two
mechanistic possibilities for diene formation (Mechanisms A
and B; Scheme 6). We originally postulated that the diazenyl
radical 2 would first undergo fragmentation to afford
a sulfonyl radical and 3 (Mechanism A). This diazene would
then rapidly undergo a cycloreversion to release nitrogen and
afford the desired diene 4. Alternatively, 2 may fragment to
afford the stabilized radical 11 (Mechanism B). The resulting
allylic radical could subsequently fragment to form a sulfonyl
radical,[34] nitrogen gas, and the desired diene. To differentiate
between these two mechanistic pathways, we examined the
cyclization of the hydrazone 14, which contains a Boc-
hydrazone (Scheme 8). If Mechanism A is operative, the
those of the two-step process (Table 1, entries 4–6). This one-
pot process can also be applied to the reaction with the more
sterically encumbered aldehyde 5d (Scheme 5), whereas the
one-pot procedure also provided the desired diene 4j in
comparable yield and diastereoselectivity to the two-step
process.
We next explored the mechanism of the radical cascade
step. The first step of the radical cascade involves formation
of the E- and Z-vinyl radicals 9 and 10 from 1 (Scheme 6).[29]
Scheme 8. Isolation of the cyclic hydrazine 16. Cy=cyclohexyl.
Boc group should slow fragmentation to the cyclic diazene
relative to hydrogen transfer from tributyltin hydride.[35] Thus,
the cyclic hydrazine 16 should be the predominant product
and little to no diene should be observed. However, if
Mechanism B is operative, the diazene intermediate 17 will
form because the Boc group should have little effect on the
fragmentation to the allyl radical. Once formed, 17 is unstable
and should rapidly decompose to the diene. Treatment of 14
under our optimized reaction conditions exclusively afforded
16 in 86% yield, thus suggesting Mechanism A is the
predominant pathway.
We have successfully demonstrated a new method that
can be utilized for the rapid synthesis of E dienes in good
yields and high diastereoselectivities starting from aldehydes.
Not only is this methodology synthetically useful, but it is also
the first example of directly converting tosyl hydrazones into
cyclic diazene intermediates. The one-pot condensation/
Scheme 6. Possible mechanistic pathways for the formation of the
diene 4 from hydrazone 1.
The two vinyl radicals readily interconvert[30] and only 10 has
the requisite geometry to undergo the 6-endo cyclization to
the diazenyl radical 2. This geometry inversion is evident by
the observation that E-enriched hydrazones (> 20:1) only
afford products resulting from 6-endo cyclizations.[31]
If the vinyl radical undergoes a 6-endo cyclization onto the
hydrazone, we postulated that there may exist very small
amounts of a tin-bound cyclic diazo adduct, such as 13
(Scheme 7), thus resulting from radical recombination of the
6230 www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 6228 –6231

Angewandte
Chemie
[16] For a review on the utility of carbanions in diastereoselective
cyclization/pericyclic cascade procedure further enhances the
synthetic utility and makes it competitive with current diene
syntheses from aldehydes. In addition to the synthesis of
simple dienes, we are currently exploring this new method-
ology in a late-stage step for the synthesis of a natural
product.
synthesis, see: M. Braun, Angew. Chem. 1998, 110, 444 – 465;
Angew. Chem. Int. Ed. 1998, 37, 430 – 451.
[17] G. K. Friestad, Top. Curr. Chem. 2012, 320, 61 – 92.
[18] a) G. A. IbꢀÇez, A. C. Olivieri, G. M. Escandar, J. Chem. Soc.
Faraday Trans. 1997, 93, 545 – 551; b) D. Pettersen, R. P. Herrera,
L. Bernardi, F. Fini, V. Sgarzani, R. Fernandez, J. M. Lassaletta,
A. Ricci, Synlett 2006, 239 – 242; c) M. Hong, H.-D. Yin, S.-W.
Chen, D.-Q. Wang, J. Organomet. Chem. 2010, 695, 653 – 662.
[19] a) S. I. Miller, W. G. Lee, J. Am. Chem. Soc. 1959, 81, 6313 –
6319; b) M. J. S. Dewar, M. Shanshal, J. Am. Chem. Soc. 1969,
91, 3654 – 3655.
[20] See the Supporting Information for further details.
[21] An erosion of diastereoselectivity was observed if the tributyltin
hydride was not distilled prior to use.
[22] Replacing AIBN for 1,1’-azobis(cyclohexanecarbonitrile)
(ABCN) yielded identical results. Other metal-hydride sources,
such as triphenyltin hydride and tris(trimethylsilyl)silane, pro-
vided very low yields.
Received: March 11, 2014
Published online: April 29, 2014
Keywords: hydrazones · pericyclic reaction · radical reactions ·
.
rearrangement · synthetic methods
[1] For recent reviews on transformations of hydrazones, see: a) S.
Kobayashi, Y. Mori, J. S. Fossey, M. M. Salter, Chem. Rev. 2011,
111, 2626 – 2704; b) R. O. Hutchins, M. K. Hutchins in Compre-
hensive Organic Synthesis, Vol. 8 (Ed.: B. M. Trost), Elsevier,
Oxford, 1991, p. 327; c) A.-Z. A. Elassar, H. H. Dib, N. A. Al-
Awadi, M. H. Elnagdi, ARKIVOC 2007, 2, 272 – 315; d) H. R.
Vollmer, Synlett 1999, 1844.
[23] R. E. Buckles, G. V. Mock, J. Org. Chem. 1950, 15, 680 – 684.
[24] R. Appel, Angew. Chem. 1975, 87, 863 – 874; Angew. Chem. Int.
Ed. Engl. 1975, 14, 801 – 811.
[25] All of the bromoallyl hydrazines are air stable and may be kept
on the bench for months without any observable decomposition.
[26] For selected reviews of piperidine analogues in pharmaceutical
sciences, see: a) S. Kꢁllstrçm, R. Leino, Bioorg. Med. Chem.
2008, 16, 601 – 635; b) I. Dragutan, V. Dragutan, A. Demonceau,
RSC Adv. 2012, 2, 719 – 736; c) A. Ahmed, K. I. Molvi, S. Nazim,
B. Irshad, T. Memon, M. Rahil, J. Chem. Pharm. Res. 2012, 4,
872 – 880.
[27] Although a small loss of yield was observed when compared to
other substrates, this could be attributed to a difficulty in
purification.
[28] A. Onishchenko, Diene Synthesis, D. Davey, New York, 1964.
[29] For an early example of halide abstraction from olefins by
trialkyltin radicals, see: C. Walling, J. H. Cooley, A. A. Ponaras,
E. J. Racah, J. Am. Chem. Soc. 1966, 88, 531 – 536.
[2] L. Wolff, Justus Liebigs Ann. Chem. 1912, 394, 86 – 108.
[3] N. Kishner, J. Russ. Phys. Chem. Soc. 1911, 43, 582 – 595.
[4] R. H. Shapiro, M. J. Heath, J. Am. Chem. Soc. 1967, 89, 5734 –
5735.
[5] R. O. Hutchins, M. Kacher, L. Rua, J. Org. Chem. 1975, 40, 923 –
926.
[6] G. W. Kabalka, D. T. C. Yang, J. D. Baker, Jr., J. Org. Chem.
1976, 41, 574 – 575.
[7] These diazenes rapidly desilylate to the alkyl diazene: a) A. G.
Myers, P. J. Kukkola, J. Am. Chem. Soc. 1990, 112, 8208 – 8210;
b) A. G. Myers, M. Movassaghi, J. Am. Chem. Soc. 1998, 120,
8891 – 8892.
[8] W. R. Bamford, T. S. Stevens, J. Chem. Soc. 1952, 4735 – 4740.
[9] D. A. Mundal, K. E. Lutz, R. J. Thomson, Org. Lett. 2009, 11,
465 – 468.
[10] There are several elegant methods to access cyclic diazenes from
pyridazines. For a representative example, see: D. L. Boger,
Tetrahedron 1983, 39, 2869 – 2939.
[11] For an example of the cycloreversion of diazenes to yield
E dienes, see: A. Padwa, T. Kumagai, M. Tohidi, J. Org. Chem.
1983, 48, 1834 – 1840.
[12] For representative examples of anionic 6-endo cyclizations, see:
a) Y. Mori, K. Yaegashi, H. Furukawa, J. Am. Chem. Soc. 1997,
119, 4557 – 4558; b) S. Hanessian, T. Focken, X. Mi, R. Oza, B.
Chen, D. Ritson, R. Beaudegnies, J. Org. Chem. 2010, 75, 5601 –
5618; c) T. Sakai, A. Sugimoto, Y. Mori, Org. Lett. 2011, 13,
5850 – 5853.
[13] For representative examples of radical 6-endo cyclizations, see:
a) B. B. Snider, J. Y. Kiselgof, Tetrahedron 1998, 54, 10641 –
10648; b) R. Chuard, A. Giraud, P. Renaud, Angew. Chem.
2002, 114, 4497 – 4499; Angew. Chem. Int. Ed. 2002, 41, 4321 –
4323; c) C.-P. Chuang, A.-I. Tsai, M.-Y. Tsai, Tetrahedron 2013,
69, 3293 – 3301.
[14] For representative examples of radical and anionic eliminations
of sulfonamides, see: a) W. Paterson, G. R. Proctor, J. Chem.
Soc. 1965, 485 – 489; b) W. R. McKay, G. R. Proctor, J. Chem.
Soc. Perkin Trans. 1 1981, 2435 – 2442; c) H. Zhang, E. B. Hay,
S. J. Geib, D. P. Curran, J. Am. Chem. Soc. 2013, 135, 16610 –
16617.
[15] For representative examples to yield aromatic rings, see:
a) W. R. Dolbier, K. Matsui, H. J. Dewey, D. V. Horak, J.
Michl, J. Am. Chem. Soc. 1979, 101, 2136 – 2139; b) K. J. Stone,
M. M. Greenberg, S. C. Blackstock, J. A. Berson, J. Am. Chem.
Soc. 1989, 111, 3659 – 3671.
[30] G. D. Sargent, M. W. Browne, J. Am. Chem. Soc. 1967, 89, 2788 –
2790.
[31] All of the starting hydrazines are E enriched. The hydrazine 6a
is isolated as a 3:1 mixture of E/Z isomers while the hydrazines
6b and 6c are isolated as a > 20:1 mixture of E to Z isomers.
[32] Direct addition of the tributyltin radical to the nitrogen would be
unable to form the nitrogen–tin adduct, as this would require the
vinyl bromide to undergo a radical SN2 displacement by the
resulting intermediate carbon radical. No known examples of
this transformation could be found in the literature.
[33] It is possible that 13 is formed by the addition of the cyclic
hydrazine with tributyltin bromide. However, a control reaction,
in which tributyltin bromide was generated in situ in the
presence of Boc-hydrazine, did not afford the stannylated
product.
[34] Sulfonyl radicals are well-known to form through fragmentation
processes. For a review, see: I. Rosenstein in Radicals in Organic
Synthesis, Vol. 1, 1st ed. (Eds.: P. Renaud, M. Sibi), Wiley-VCH,
Weinheim, 2001, pp. 50 – 71.
[35] Boc groups are sufficiently resistant to radical fragmentation and
they are stable to both intramolecular and intermolecular radical
reactions. For representative examples, see: a) D. P. G. Hamon,
R. A. Massy-Westropp, P. Razzino, J. Chem. Soc. Chem.
Commun. 1991, 722 – 724; b) D. P. G. Hamon, R. A. Massy-
Westropp, P. Razzino, Tetrahedron 1995, 51, 4183 – 4194; c) P. G.
Parzuchowski, M. C. Frost, M. E. Meyerhoff, J. Am. Chem. Soc.
2002, 124, 12182 – 12191.
Angew. Chem. Int. Ed. 2014, 53, 6228 –6231
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6231