An efficient synthesis of bridged heterocycles from an Ir(I) bis-amination/ring-closing metathesis sequence

Article abstract of DOI:10.1021/ol302108m

The amination of bis-allylic imidates using an Iridium(I) catalyst leads to the efficient formation of 2,6-divinyl heterocycles. Careful screening of amines, solvents, and conditions has led to the discovery of a system that favors formation of the desired cis products with synthetically useful levels of diastereoselectivity, and these results are further explained by computer based transition state energy calculations. Exposure of the heterocycles to ring-closing metathesis catalysts leads to the desired bridged heterocyclic systems.

Full text of DOI:10.1021/ol302108m

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
An Efficient Synthesis of Bridged
Heterocycles from an Ir(I) Bis-Amination/
Ring-Closing Metathesis Sequence
~
Ryan A. Brawn,* Cristiano R. W. Guimaraes, Kim F. McClure, and Spiros Liras
CVMED Chemistry, Pfizer Worldwide Research and Development, 445 Eastern Point
Road, Groton, Connecticut 06340, United States
ryan.brawn@pfizer.com
Received July 30, 2012
ABSTRACT
The amination of bis-allylic imidates using an Iridium(I) catalyst leads to the efficient formation of 2,6-divinyl heterocycles. Careful screening of
amines, solvents, and conditions has led to the discovery of a system that favors formation of the desired cis products with synthetically useful
levels of diastereoselectivity, and these results are further explained by computer based transition state energy calculations. Exposure of the
heterocycles to ring-closing metathesis catalysts leads to the desired bridged heterocyclic systems.
Bridged heterocycles are popular building blocks for
medicinal chemists, and they have been used as key com-
ponents for a variety of pharmaceutical agents.1 Drug
targets containing bridged heterocycles often have similar
physiochemical properties as their unbridged counter-
parts, while the more rigid bridged structure can impart
additional potency and target selectivity.2 Despite contin-
ued interest in these building blocks, there remain a few
concise synthetic routes to form bridged heterocycles. The
routes reported generally suffer from low product yields
as a consequence of lengthy syntheses.3 Herein we report
a concise synthesis of a number of nitrogen containing
bridged heterocyclic compounds through an iridum cata-
lyzed bis-allylic amination/Grubbs ring-closing metathesis
(RCM) sequence. We also provide a mechanistic hypoth-
esis to explain the stereochemical course of these reactions.
The metal catalyzed amination of allylic acetates was
originally reported using either basic conditions or a
palladium catalyst system.4 More recently iridium5 and
rhodium catalyzed aminations have emerged as reliable
conditions for the selective synthesis of allylic amines
from allylic alcohols, esters, carbonates, or imidates.6
While there are a few examples in the literature of bis-
allylic aminations for the formation of heterocycles, the
diastereoselectivity is typically poor or strongly favors the
trans isomer, which is not suitable to RCM.7 To form the
desired bridged system using an RCM strategy we would
need to first find bis-allylic amination conditions that
favored the formation of cis divinyl azine precursors
(1) (a) Zask, A.; Kaplan, J.; Verheijean, J. C.; Richard, D. J.; Curran,
K.; Brooijmans, N.; Bennett, E. M.; Toral-Barza, L.; Hollander, I.;
Ayral-Kaloustian, S.; Yu, K. J. Med. Chem. 2009, 52, 7942–7945. (b)
Curran, K. J.; Verheijean, J. C.; Kaplan, J.; Richard, D. J.; Toral-Barza,
L.; Hollander, I.; Lucas, J.; Ayral-Kaloustian, S.; Yu, K.; Zask, A.
Bioorg. Med. Chem. Lett. 2010, 20, 1440–1444. (c) Li, H.; Xu, R.; Cole,
D.; Clader, J. W.; Greenlee, W. J.; Nomeir, A. A.; Song, L.; Zhang, L.
Bioorg. Med. Chem. Lett. 2010, 20, 6606–6609.
(2) Zask, A.; Verheijen, J. C.; Richard, D. J.; Kaplan, J.; Curran, K.;
Toral-Barza, L.; Lucas, J.; Hollander, I.; Yu, K. Bioorg. Med. Chem.
Lett. 2010, 20, 2644–2647.
(3) (a) Kilonda, A.; Dequeker, E.; Compernolle, F.; Delbeke, P.;
Toppet, S.; Bila, B.; Hoornaert, G. J. Tetrahedron 1995, 51, 849–858. (b)
Neipp, C. E.; Martin, S. F. J. Org. Chem. 2003, 68, 8867–8878. (c) Chen,
Z.; Venkatesan, A. M.; Dos Santos, O.; Delos Santos, E.; Dehnhardt,
C. M.; Ayral-Kaloustian, S.; Ashcroft, J.; McDonald, L. A.; Mansour,
T. M. J. Org. Chem. 2010, 75, 1643–1651. (d) Wishka, D. G.; Beagley, P.;
Lyon, J.; Farley, K. A.; Walker, D. P. Synthesis 2011, 16, 2619–2624.
(4) (a) Takagi, M.; Yamamoto, K. Chem. Lett. 1989, 2123–2124. (b)
Kim, J. D.; Lee, M. H.; Lee, M. J.; Jung, Y. H. Tetrahedron Lett. 2000,
41, 5073–5076. (c) Kim, J. D.; Lee, M. H.; Han, G.; Park, H.; Zee, O. P.;
Jung, Y. H. Tetrahedron 2001, 57, 8257–8266. (d) You, S.-L.; Zhu, X.-Z.;
Luo, Y.-M.; Hou, X.-L.; Dai, L.-X. J. Am. Chem. Soc. 2001, 123, 7471–
7472. (e) Shekhar, S.; Trantow, B.; Leitner, A.; Hartwig, J. F. J. Am.
Chem. Soc. 2006, 128, 11770–11771. (f) Nagano, T.; Kobayashi, S.
J. Am. Chem. Soc. 2009, 131, 4200–4201.
(5) For additions of other nucleophiles to Ir(I) π-allyl systems, see:
Stanley, L. M.; Bai, C.; Ueda, M.; Hartwig, J. F. J. Am. Chem. Soc. 2010,
132, 8918–8920.
r
10.1021/ol302108m
XXXX American Chemical Society

(4ꢀ6).8 These heterocycles would be derived from acyclic
bis-allylic alcohols, protected to form a more reactive
allylic system.
temperatures, resulting in low levels of diastereoselectivity.
Nguyen reported that an allylic imidate is a highly reactive
leaving group for allylic amination, allowing the addition
of secondary amines which were not reactive with acetates
or carbonates.6i Diols 2 were converted to the desired bis-
imidates 3 in good yields using trichloroacetonitrile and
DBU (Scheme 2).
Scheme 1. Formation of Bis-allylic Alcohols 2 from Cyclic
Diolsa
Scheme 2. Formation of Bis-imidates 3 from Diols 2a
a Yields refer to purified material after isolation by column chroma-
tography over silica gel.
The synthetic sequence began with cyclic diols 1, which
were commercially available in the cases of 1a (X = O) and
1b (X = CH2), or could be formed in a single step by a
dihydroxylation of commercially available N-Boc 2,5-
dihydropyrrole in the case of 1c (X = N-Boc).9 Oxidative
cleavage of the diol with di(acetoxy)iodobenzene in DCM,
followed by addition of vinylmagnesium bromide to the
crude dialdehyde at low temperature, cleanly afforded the
desired bis-allylic alcohols 2 (Scheme 1).10 Attempts to
purify the dialdehyde were unsuccessful due to issues with
stability and volatility, favoring the one-pot procedure.
While there are examples of allylic amination reactions
on allylic alcohols, this reactivity typically favors mono-
amination and does not provide the desired heterocycles in
our system.6 The desired product could be formed using a
bis-allylic acetate or bis-allylic carbonate, but the reactivity
was poor and the reactions only went tocompletion at high
a Yields refer to purified material after isolation by column chroma-
tography over silica gel.
With the desired bis-imidates in hand, a variety of
reaction conditions were screened to determine which
combination of catalyst, amine, and solvent would lead
to selective formation of the desired 2,6-cis heterocycles.
When bis-imidate 3a was exposed to a primary benzyla-
mine derivative in the presence of an iridium(I) catalyst the
major product was the desired divinyl morpholine 4.
Palladium(0) gave the product in lower yields. Amides,
carbamates, and sulphonamides proved to be poor reac-
tion partners, giving little to no desired products under a
variety of conditions. Benzylamine gave the product in
acceptable yield, but the diastereoselectivity was poor and
slightly favored the undesiredtrans diastereomer (Table 1).
Reactions withtritylamine asthe nitrogen sourcegavehigh
yields but poor selectivity under a variety of conditions.
Cumylamine proved to be the best nitrogen source for the
allylic amination reactions, providing the desired morpho-
line product in high yields and favoring the desired cis
(6) (a) Takeuchi, R.; Ue, N.; Tanabe, K.; Yamashita, K.; Shiga, N.
J. Am. Chem. Soc. 2001, 123, 9525–9534. (b) Matunas, R.; Lai, A. J.; Lee,
C. Tetrahedron 2005, 61, 6298–6308. (c) Defieber, C.; Ariger, M. A.;
Moriel, P.; Carreira, E. M. Angew. Chem., Int. Ed. 2007, 46, 3139–3143.
(d) Pouy, M. J.; Leitner, A.; Weix, D. J.; Ueno, S.; Hartwig, J. F. Org.
Lett. 2007, 9, 3949–3952. (e) Singh, O. V.; Han, H. Org. Lett. 2007, 9,
4801–4804. (f) Evans, P. A.; Clizbe, E. A. J. Am. Chem. Soc. 2009, 131,
8722–8723. (g) Vrieze, D. C.; Hoge, G. S.; Hoerter, P. Z.; Van Haitsma,
J. T.; Samas, B. M. Org. Lett. 2009, 11, 3140–3142. (h) Roggen, M.;
Carreira, E. M. J. Am. Chem. Soc. 2010, 132, 11917–11919. (i) Arnold,
J. S.; Stone, R. F.; Nguyen, H. M. Org. Lett. 2010, 12, 4580–4583. (j)
Tosatti, P.; Horn, J.; Campbell, A. J.; House, D.; Nelson, A.; Marsden,
S. P. Acc. Chem. Res. 2010, 352, 3153–3157. (k) Lafrance, M.; Roggen,
M.; Carreira, E. M. Angew. Chem., Int. Ed. 2012, 51, 3470–3473. For an
example using Au(III), see: Guo, S.; Song, F.; Liu, Y. Synlett 2007, 6,
964–968.
1
isomer, which was assigned based on H NMR analysis.
After screening a number of solvents it was found that
dichloroethane and toluene both gave moderate diaste-
reoselectivity, favoring the desired 2,6-cis isomer of mor-
pholine 4. More polar solvents typically gave good yields
but poor diastereoselectivity.11
Further reaction optimization found that the best cata-
lyst for the bis-allylic amination reaction was the commer-
cially available [Ir(cod)Cl]2, and the optimum conditions
were found when the reaction was run in dichloroethane
starting at 0 °C and allowed to warm slowly to room
temperature over 18 h (2 h of warming followed by 16 h at
ambient temperature; see Table 2). Reactions in toluene
also gave good levels of selectivity but typically gave
slightly lower yields. Reactions at lower temperature gave
lower yields due toincomplete conversion. Careful workup
was necessary to remove the catalyst, since it is highly
soluble in a variety of organic solvents. Prolonged exposure
(7) (a) Miyabe, H.; Yoshida, K.; Kobayashi, Y.; Matsumura, A.;
Takemoto, Y. Synlett 2003, 1031–1033. (b) Welter, C.; Dahnz, A.;
Brunner, B.; Streiff, S.; Dubon, P.; Helmchen, G. Org. Lett. 2005, 7,
1239–1242.
(8) For some reviews of ring closing metathesis, see: (a) Grubbs,
R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28, 446–452. (b)
Grubbs, R. H. Tetrahedron 2004, 60, 7117–7140. (c) Deiters, A.; Martin,
S. F. Chem. Rev. 2004, 104, 2199–2238. (d) Hoveyda, A. H.; Zhugralin,
A. R. Nature 2007, 450, 243–251. For examples of RCM to form bridged
systems, see: (e) Aggarwal, V. K.; Astle, C. J.; Rogers-Evans, M. Org.
Lett. 2004, 6, 1469–1471. (f) Kuznetsov, N. Y.; Khrustalev, V. N.;
Godovikov, I. A.; Bobnov, Y. N. Eur. J. Org. Chem. 2006, 113–120.
(g) Cheng, G.-L.; Wang, X.-Y.; Zhu, R.; Shao, C.-W.; Xu, J.-M.; Hu,
Y.-F. J. Org. Chem. 2011, 76, 2694–2700.
(9) N-Boc dihydropyrrole was dihydroxylated using OsO4/NMO
conditions; see Supporting Information for details.
(10) Nicolaou, K. C.; Adsool, V. A.; Hale, C. R. H. Org. Lett. 2010,
12, 1552–1555.
(11) Reactions in propionitrile, DMSO, diglyme, and diethyl ether
also gave moderate yields and low levels of diastereoselectivity. Reac-
tions did not proceed in heptane, presumably due to poor solubility
B
Org. Lett., Vol. XX, No. XX, XXXX

Table 1. Diastereoselective Bis-allylic Amination To Form 4a
Table 2. Optimization of Bis-allylic Amination Conditionsa
amine
catalyst
solvent temp yieldb drc product
BnNH2
BnNH2
BnNH2
[Ir(cod)Cl]2
Pd(PPh3)4
[Ir(cod)Cl]2
MeCN
MeCN
DCE
60 °C 42
1:1.3
1:1
4a
4a
4a
4b
4b
4b
4b
4c
4c
4c
4c
4c
4c
4c
4c
4c
4c
4c
rt
32
solvent
catalyst
temp
time (h) yieldb
drc
rt
trace N/A
tritylNH2 [Ir(cod)Cl]2
tritylNH2 [Ir(cod)Cl]2
tritylNH2 [Ir(cod)Cl]2
tritylNH2 [Ir(cod)Cl]2
cumylNH2 [Ir(cod)Cl]2
cumylNH2 [Ir(cod)Cl]2
MeCN
MeCN
THF
60 °C 64
1.1:1
1.1:1
1.1:1
1.1:1
1.4:1
1.3:1
1.5:1
DCE
DCE
DCE
DCE
DCE
DCE
[Ir(cod)Cl]2
[Ir(cod)Cl]2
rt
0 °C
18
3
53
68
73
79
74
70
39
66
66
45
47
59
3.5:1
5.1:1
4.4:1
5.1:1
4.7:1
4.9:1
4.2:1
3.7:1
4.2:1
4.3:1
2.4:1
5.1:1
rt
rt
rt
92
76
62
[Ir(cod)OMe]2 0 °C
[Ir(cod)Cl]2
3
DCE
0 °C to rt
18
18
18
18
18
18
3
MeCN
MeCN
60 °C 52
[Ir(cod)OMe]2 0 °C to rt
[Ir(cod)Cl]2
rt
rt
rt
rt
61
ꢀ15 °C to rt
ꢀ20 °C
rt
cumylNH2 [Ir(cod)OMe]2 MeCN
65
toluene [Ir(cod)Cl]2
toluene [Ir(cod)Cl]2
cumylNH2 Pd(PPh3)4
cumylNH2 [Ir(cod)Cl]2
cumylNH2 [Ir(cod)Cl]2
MeCN
THF
trace N/A
58
54
53
57
55
66
75
2.0:1
2.3:1
3.5:1
2.2:1
1.3:1
3.7:1
1.2:1
toluene [Ir(cod)OMe]2 rt
dioxane rt
toluene [Ir(cod)Cl]2
0 °C
cumylNH2 [Ir(cod)Cl]2 DCE
rt
rt
rt
toluene [Ir(cod)OMe]2 0 °C
toluene [Ir(cod)Cl]2
3
cumylNH2 [Ir(cod)Cl]2
cumylNH2 [Ir(cod)Cl]2
DCM
0 °C to rt
18
CHCl3
a All reactions run at 0.2 M in the indicated solvent with 5 mol %
catalyst. b Yields refer to purified material after isolation by column
chromatography over silica gel. All yields refer to a mixture of diaste-
reomers. c Diastereomeric ratio determined by 1H NMR analysis of the
crude reaction mixture.
cumylNH2 [Ir(cod)Cl]2 toluene rt
cumylNH2 [Ir(cod)Cl]2 DMF rt
a All reactions run at 0.2 M in the indicated solvent with 5 mol %
catalyst. b Yields refer to purified material after isolation by column
chromatography over silica gel. All yields refer to a mixture of diaste-
reomers. c Diastereomeric ratio determined by 1H NMR analysis of the
crude reaction mixture.
phase of 0.33 kcal/mol. Incorporation of dichloroethane
solvent effects provided a final value of 0.98 kcal/mol or
a diastereomeric ratio of 6.1:1, very close to the most
selective conditions in Table 2. In the case of the less
selective conditions at room temperature, it is plausible
that the diastereoselectivity is thermodynamically con-
trolled. Therefore, models for the cis and trans products
were obtainedin the gas phase. The energydifference at the
B3LYP/6-311G**þ level between the two products was
0.25 kcal/mol favoring the trans isomer. Addition of zero-
point vibrational energies and thermal contributions at
25 °C to enthalpy and entropy led to a free-energy differ-
ence of 0.48 kcal/mol now favoring the cis isomer. Incor-
poration of DCE solvent effects brought the free-energy
difference to a final value of 0.71 kcal/mol, or a 3.3:1
diastereomeric ratio, in excellent agreement with the ex-
perimental value in Table 2.
Bis-imidates 3b and 3c also underwent the desired
iridium(I) catalyzed cyclization with cumylamine to form
piperidine 5 and piperazine 6 respectively (Scheme 3).
Piperidine 5 was formed in 67% yield as a 3.5:1 mixture
of diastereomers when the reaction was run at 0 °C for 3 h.
Allowing this reaction to warm to room temperature
resulted in a loss of diastereoselectivity. The bis-allylic
amination to form piperazine 6 gave the desired product
in 84% yield as a 12:1 mixture of diastereomers. Both
reactions favored the desired 2,6-cis isomer of the
heterocycle.
can lead to reinsertion into the allylic system of the product
and a complete loss of diastereoselectivity. When the
purified product was re-exposed to the catalyst system
in a polar organic solvent, a loss of diastereoselectivity
was observed over time. This effect was not observed in
dichloroethane. Direct purification of the crude reaction
mixture by silica gel chromatography was found to be the
best way to remove the catalyst. The two diastereomers
could not be separated using standard chromatographic
techniques, so all yields refer to the mixture of diastereo-
mers. These diastereomeric mixtures were taken on to the
subsequent RCM step without further purification.
In order to investigate the impact of temperature on the
diastereoselectivity observed in dichloroethane, computa-
tional models for the cis and trans transition states and
products were obtained at a high level of theory. As shown
in Figure 1 the computations modeled the transition states
leading to the cis and trans products for the second allylic
amination step. The energy difference at the B3LYP/
LACV3P**þ//B3LYP/LACV3P* level between the two
TS models is 0.73 kcal/mol favoring the cis isomer con-
sistent with a kinetically controlled formation. Addition of
zero-point vibrational energies and thermal contributions
at 0 °C to enthalpy provided a cis TS more enthalpically
favorable by 0.74 kcal/mol. Addition of entropic contribu-
tions led to a reduced free-energy difference in the gas
Org. Lett., Vol. XX, No. XX, XXXX
C

presumably due to the greater basicity of the piperidine
nitrogen. Additions of acid led to decomposition of the
starting materials, while increasing the temperature by
using xylenes as a solvent gave a modest increase in yield.
Piperazine 6 gave clean conversion to bridged ring system
7c. The cumyl group of product 7a could be easily removed
using TFA, to afford the secondary amine salt.12
Table 3. Grubbs Metathesis To Form Bridged Heterocyclesa
Figure 1. Model for the cis and trans transition states and
products. See Supporting Information for a full size model
and more details on the calculations and assumptions built into
this transition state model.
X
catalyst
solvent
temp
rt
yieldb product
O
O
O
O
Hoveyda-Grubbs 2 DCM
3
17
25
79
29
34
82
7a
7a
7a
7a
7b
7b
7c
Grubbs 2
Grubbs 2
Grubbs 2
Grubbs 2c
Grubbs 2c
DCM
DCE
rt
90 °C
toluene 120 °C
toluene 120 °C
xylenes 150 °C
toluene 120 °C
CH2
CH2
Scheme 3. Formation of Piperidine 5 and Piperizine 6a
NBoc Grubbs 2
a All reactions run with 5 mol % catalyst in the specified solvent
(0.08 M). b Yields refer to purified material after isolation by column
chromatography over silica gel. c Reactions run with 10 mol % catalyst.
In conclusion, a straightforward, efficient sequence was
developed to prepare bridged heterocycles. A one-pot
oxidative cleavage/Grignard addition gave direct access
to the key bis-allylic alcohols. Iridium(I) catalyzed bis-
allylic amination gave the desired heterocycles, with the cis
isomer predominating. Computations suggest that under
kinetically controlled conditions coordination to the irid-
ium catalyst gives rise to a more stable cis transition state.
Grubbs’ ring closing metathesis finished the sequence,
forming the desired bridged heterocycles in good yield.
The products will be used as intermediates for potential
pharmaceutical development.
a Yield refers to purified material after isolation by column chroma-
tography over silica gel. All yields refer to a mixture of diastereomers.
Diastereomeric ratio determined by 1H NMR analysis of crude reaction
mixtures.
With the divinyl azine heterocycles in hand, conditions
were screened for the Grubbs ring-closing metathesis to
form the desired bridged heterocyclic products. Treatment
of diene 4 (as a mixture of diastereomers) with Grubbs’
second generation catalyst in toluene at 120 °C gave clean
conversion of the cis isomer to desired bridged heterocycle
7a (Table 3). The structure of 7a was confirmed by X-ray
crystallography. This reaction also worked for the piper-
idine 5 leading to product 7b, albeit in lower yield,
Acknowledgment. R.A.B. would like to thank Pfizer for
the postdoctoral fellowship and Dr. Allyn Londregan and
Mr. Ben Rocke (Pfizer CVMED) for helpful discussions.
The authors would like to thank Dr. Brian Samas (Pfizer)
for obtaining X-ray data of 7a.
Supporting Information Available. Preparation and
characterization of all compounds is included. This ma-
terial is available free of charge via the Internet at http://
pubs.acs.org.
(12) For structure of the RCM catalysts and details on the selective
removal of the cumyl protecting group after cyclization, see Supporting
Information.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX