A ring-closing metathesis approach to heterocycle-fused azepines

Article abstract of DOI:10.1016/j.tetlet.2012.12.042

Heterocycle-fused azepines, an important class of molecular scaffold, are readily synthesised through the ruthenium-catalysed ring-closing metathesis reaction. Although benzo-fused azepines are well documented, heterocycle-fused examples are poorly developed. Herein, a range of five- and six-membered heterocycle-fused azepines are investigated, allowing access to a series of pharmaceutically interesting products.

Full text of DOI:10.1016/j.tetlet.2012.12.042

Accepted Manuscript
A ring-closing metathesis approach to heterocycle-fused azepines
Thomas A. Moss
PII:
S0040-4039(12)02181-8
DOI:
http://dx.doi.org/10.1016/j.tetlet.2012.12.042
TETL 42287
Reference:
To appear in:
Tetrahedron Letters
Received Date:
Revised Date:
Accepted Date:
19 October 2012
13 November 2012
11 December 2012
Please cite this article as: Moss, T.A., A ring-closing metathesis approach to heterocycle-fused azepines, Tetrahedron
Letters (2012), doi: http://dx.doi.org/10.1016/j.tetlet.2012.12.042
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
A ring-closing metathesis approach to
heterocycle-fused azepines
Leave this area blank for abstract info.
Thomas A. Moss

1
Tetrahedron Letters
journal homepage: www.elsevier.com
A ring-closing metathesis approach to heterocycle-fused azepines
Thomas A. Mossa*
aAstraZeneca Mereside, Alderley Park, Cheshire, SK10 4TG
Tel: (+44) 01865 513452
Email: thomas.moss@astrazeneca.com
ARTICLE INFO
ABSTRACT
Article history:
Received
Received in revised form
Accepted
Heterocycle-fused azepines, an important class of molecular scaffold, are readily synthesized
through the ruthenium-catalyzed ring-closing metathesis reaction. Although benzo-fused
azepines are well documented, heterocycle-fused examples are poorly developed. Herein, a
range of five- and six-membered heterocycle-fused azepines are investigated, allowing access to
a series of pharmaceutically interesting products.
Available online
Keywords:
metathesis; azepine; heterocycle; bicycle
2009 Elsevier Ltd. All rights reserved.
Studies have identified a direct link between the degree of sp3
saturation in a drug candidate molecule and its probability of
clinical success.1 With this in mind, we recently described a one-
pot procedure for the synthesis of fused aza-indolines and aza-
tetrahydroquinolines through an alkylation/ intramolecular
cyclisation sequence with N-protected cyclic sulfamidates.2
Although this allowed a range of partially saturated bicycles to be
synthesized, larger ring systems (7- and above) were not
accessible via this methodology. Medium-sized fused ring
systems such as azepine derivatives are a particularly interesting
structural motif, being found in a number of bioactive natural
products and pharmaceuticals. Within this class, benzazepine
derivatives have found use as vasopressin V2 receptor antagonists
(e.g. tolvaptan),3 and in other disease areas.4 Due to this
widespread utility, benzazepine derivatives have attracted
considerable attention from the synthetic community, with
typical synthetic routes including ring expansion,5 Dieckmann
condensations6 and intramolecular metal-catalyzed coupling
reactions.7
Studies began by investigating the synthesis of the required
pyrimido-fused azepine precursors, as fused pyrimidines are
known to be privileged scaffolds in medicinal chemistry.
Treatment of the parent chloro-pyrimidines with TMPMgCl·LiCl
(TMP = 2,2,6,6-tetramethylpiperidine) at room temperature,
followed by quenching with an allyl bromide gave the 5-allylated
products.11 For less electron rich pyrimidines such as 2,4,6-
trichloropyrimidine and 4,6-dichloropyrimidine, the metallation
o
was conducted at -78 C with CuBr as an additive to minimize
decomposition. Displacement of the chloride with allylamine or
an allylamine derivative under microwave irradiation and
subsequent N-Boc protection gave the required precursors 1-8 in
good overall yields. Alternatively, the chloro-intermediates could
be accessed by cyclisation of diethylallyl malonate with the
appropriate amidine/urea/guanidine followed by chlorination
with POCl3 in MeCN (Scheme 1).
A very useful method for synthesizing medium sized N-
containing heterocycles is the ruthenium-catalyzed ring-closing
metathesis reaction (RCM).8 Since its inception, the RCM
reaction has been widely used in the synthesis of benzo-fused
azepines (7-) and azocines (8-).9 However one area that has not
been well researched is the synthesis of heterocycle-fused
azepine ring systems, particularly those which contain key
molecular recognition sites such as pyrimidines, pyridines and
azoles10 (Figure 1). Herein is described the synthesis of these
useful building blocks through a ring-closing metathesis strategy,
and the synthetic routes towards their RCM precursors.
Scheme 1. Reagents and conditions: a) TMPMgCl·LiCl, THF,
CH2=CHR3CH2Br, RT or -78 oC. b) NH2CH2CR4=CH2, Et3N, MeCN, 140 oC.
c) Boc2O, DMAP, MeCN, reflux.
Figure 1. 5- and 6-membered heterocycle-fused azepine derivatives.

2
Tetrahedron Letters
A modified procedure was used for the synthesis of pyrimido-
fused azepinone precursors: diplacement of the chloride with
ammonia, followed by acylation and N-methylation gave
acrylamide 9, whereas Grignard addition to the allylation product
(step d) of a commercially available 5-substituted aldehyde
12
followed by oxidation of alcohol 10 with MnO2 gave vinyl
ketone 11 (Scheme 2).
Scheme 4. Reagents and conditions: a) n-BuLi , allyl bromide, THF, -78 oC.
b) NaH, allyl bromide, DMF, 0 oC. c) NaHDMS, HCO2Et, THF, –78 °C
then BnNHNH2·HCl then d) Boc2O, DMAP, MeCN, RT then b) for
18; NaHDMS, m-ClC6H4CO2Et, THF, –78 °C then NH2OH·HCl then
e) AcCl, pyridine, RT then b) for 19.
Scheme 2. Reagents and conditions: (a) NH3/MeOH (7 M), 100 oC, 30 min.
(b) NaHMDS, acryloyl chloride, THF, -78 oC. (c) NaH, MeI, DMF, -10 oC.
(d) N-allylmethylamine, CH2Cl2. (e) vinylmagnesium chloride, THF, -78 oC.
(f) MnO2, CH2Cl2, reflux.
With the desired azepine precursors in hand, their suitability
in RCM reactions was investigated (Table 1). Treatment of
unprotected pyrimidine 1 with 5 mol% Grubbs (II) catalyst in a
solution of either CH2Cl2 or toluene only gave traces of the RCM
azepine product 20, even at elevated temperatures or in the
presence of an acid co-catalyst. This was slightly surprising since
RCM reactions are known on substrates containing NH
functionality,9a leading us to postulate that the free amino-
pyrimidine may be ligating the catalyst. Pleasingly, the N-Boc
protected precursor 2 did react smoothly in the RCM, giving the
azepine derivative in good yield after stirring for 16 hours at
room temperature in CH2Cl2, or for 1 hour at reflux in CH2Cl2.
Interestingly, a small amount (up to 5%) of the piperidine
product (presumably formed from isomerisation of the C-allyl
prior to cyclisation) had also formed in the reaction. The two
products were inseparable by chromatography, but could be
easily separated once the N-Boc had been cleaved with TFA in
CH2Cl2, allowing access to pyrido-fused azepine 20 in good
overall yield from 2. Separation difficulties as a result of
piperidine formation was sometimes seen in other pyrimidine and
pyridine-fused ring-closed products (but not the azole-fused
products), thus the N-Boc cleavage was routinely telescoped with
the RCM reaction to give the NH products directly. No
intermolecular reaction products were seen in the ring-closing
metathesis, so all the reactions were carried out at moderate
dilution (0.1 M). The N-Boc group could be replaced with an N-
Me 3 giving the azepine product 21 in high yield. This again
points to the NH- aminopyrimidine being the cause of the
sluggish reaction with 1. N- and C-methylallyl precursors 4 and 5
were both well tolerated in the RCM, giving the 7-methylated
and 6-methylated pyridoazepines 22 and 23, respectively, in good
yields. In these cases, the RCM reaction was sluggish at room
temperature, and required heating to reflux in CH2Cl2 to furnish
the products efficiently, with no observable isomerisation of the
alkene. Various orthogonally reactive substituents including
chloro-, amino- and methyl could be appended to the
pyrimidines, with compounds 6-8 all reacting smoothly to give
the azepine derivatives 24-26 in good yields after N-Boc
cleavage.
Next, the synthesis of pyrido-fused azepine precursors was
investigated. As pyridines are generally more stable to
organometallic reagents than pyrimidines, the required precursors
could be synthesized by a directed ortho-metallation/allylation
strategy.13 Treatment of the NHBoc compounds with n-BuLi
o
(2.25 equiv) in THF at -78 C gave the ortho-lithiated species
which were transmetallated with copper(I) bromide before
quenching with allyl bromide. With this strategy and by judicious
choice of the starting material, the relative positions of the C-
allyl and N-allyl groups could be moved around the aryl ring.
Interestingly, in the absence of copper, a significant amount of
the ortho-brominated pyridines were isolated. The N-allyl moiety
could then be introduced by alkylation with allyl bromide in
DMF to give the pyridine substrates 12-14. Double
lithiation/allylation was also attempted, but no significant
amounts of N-allylation were observed under these conditions
(Scheme 3).
Scheme 3. Reagents and conditions: a) n-BuLi, CuBr, allyl bromide, THF, -
78 oC. b) NaH, allyl bromide, DMF, 0 oC.
Finally, viable synthetic routes towards five-membered
heterocycle-fused azepine precursors were investigated. As for
pyridine-fused compounds,
a directed metallation/allylation
strategy could be used in most cases, giving access to thiophene-,
furan- and thiazole-fused azepine precursors 15-17. 1,2-Azoles
such as pyrazoles and isoxazoles were not stable under these
conditions, however the desired starting materials 18-19 could be
constructed by an alternative approach involving acylation of 5-
pentenenitrile, followed by addition of hydroxylamine
hydrochloride or N-benzylhydrazine hydrochloride. The NH2
products were then protected (in the case of isoxazole 19, N-Boc
protection was sluggish, so N-acylation was performed), and
subsequently alkylated with NaH and allyl bromide in DMF
(Scheme 4).
N-Me acrylamide 9 was also successful in the RCM reaction,
giving the azepin-8-one 27 in a pleasing 84% yield. Again in this
case, the free NH precursor was not well tolerated. Alcohol 10
worked well in the RCM reaction, giving hydroxy-azepine 28 in
85% yield. Interestingly however, the vinyl ketone 1114 gave only
traces of the RCM product with 5 mol% of Grubbs (II) catalyst.
Instead, the major product was the Morita Baylis-Hilman type
adduct 29. Aromatic vinyl ketones are known to undergo Baylis-

3
Hilman type dimerisation reactions with a range of nucleophilic
catalysts including DABCO, phosphines and phosphine ligated
metal complexes.14 Anticipating that a dissociated tricyclohexyl
phosphine ligand present in the Grubbs (II) catalyst might be
promoting this unwanted pathway, the same reaction was
attempted using the Hoveyda-Grubbs (II) catalyst, which does
not contain any phosphine ligands. Gratifyingly, treatment of
vinyl ketone 11 with 5 mol% of Hoveyda-Grubbs (II) catalyst in
CH2Cl2 gave solely the azepin-5-one product 30 in 82% yield
after heating to reflux for 4 hours. Pyridines 12-14 worked well
in the RCM reaction, giving pyrido-fused azepines 31-33 in good
yields (77-85%). Pleasingly, the five-membered heterocycles
were also good substrates in the RCM reaction. Thiophene 15
reacted smoothly at RT to give the azepine 34 in good yield,
however the NH compound, after N-Boc cleavage, was found to
be somewhat unstable as the free-base, so the product was
stored as the N-Boc protected form. Furan-fused azepine 35,
formed rapidly at RT (85%, H NMR yield of the crude product
after stirring for 1 hour), but was found to be unstable. Even the
N-Boc protected product 35 was prone to isomerisation and
decomposition on silica gel. Thiazole 17 did not react at room
temperature, but formed azepine 36 in good yield after heating to
1
o
70 C in toluene for 3 hours. Further heating the RCM reaction
with thiazole 17 to reflux in toluene gave the isomerised alkene
37 in 83% yield after 4 hours, allowing both isomers of the
product to be readily obtained from the same starting material
simply by conducting the reaction at different temperatures. 1,2-
Azoles 18 and 19 both worked well in the RCM reaction giving
the pyrazole- and isoxazole-fused azepine derivatives 38 and 39
respectively, in high yields.
Table 1 Scope of the RCM reaction
Entry
1
Starting Material
Conditions
CH2Cl2, RT
Time (h)
16
Product
Yield (%)a
<5
1
2
20
CH2Cl2, RT
then
TFA, CH2Cl2, RT
20: 85
16
1
20b: <5
2
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
20: 83
89
20
20b
CH2Cl2, 45 oC
3
4
1
1
3
4
21
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
83
22
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
5
6
2
2
1
1
3
1
91
88
86
92
84
85
23
5
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
24
6
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
7
25
7
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
8
8
26
toluene
70 oC
9
27
28
9
CH2Cl2
45 oC
10
10

4
Tetrahedron Letters
CH2Cl2
RT
11
16
46
82
11
11
29
CH2Cl2
45 oC
12b
4
30
31
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
13
24
77
12
13
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
14
15
1
1
82
85
32
33
CH2Cl2, 45 oC
then
TFA, CH2Cl2, RT
14
CH2Cl2
RT
16
2
80
34
15
CH2Cl2
RT
17
18
19
20
2
3
4
2
(85) <5c
35
16
toluene
70 oC
81
36
17
toluene
110 oC
83
80
37
17
CH2Cl2, RT
then
TFA, CH2Cl2, RT
38
18
CH2Cl2
RT
21
2
93
19
39
a Isolated yields after column chromatography. b 5 mol% Hoveyda-Grubbs (II)
used as the catalyst. c Value in parenthesis is the crude 1H NMR yield. The
product could not be readily isolated on silica gel.
X.; Morris, S. W.; Webb, T. R. Bioorg. Med. Chem. 2009, 17,
3308.
5. Grunewald, G. L.; Dahanukar, V. H.; Ching, P.; Criscione, K. R.
J. Med. Chem. 1996, 39, 3539.
In conclusion, the ruthenium-catalysed ring-closing metathesis
is shown to be an extremely useful reaction in the synthesis of
heterocycle-fused azepine and azepinone derivatives.15 The
products formed have multiple handles for further elaboration
and constitute exceptionally versatile fragment-sized building
blocks. As such, this transformation and the synthetic routes
towards the required starting materials should be of interest to the
synthetic community in general, and particularly in medicinal
chemistry campaigns.
6. Stukenbrock, H.; Mussmann, R.; Geese, M.; Ferandin, Y.; Lozach,
O.; Lemcke, T.; Kegel, S.; Lomow, A.; Burk, U.; Dohrmann, C.;
Meijer, L.; Austen, M.; Kunick, C. J. Med. Chem. 2008, 51, 2196.
7. (a) Bendorf, H. D.; Ruhl, K. E.; Shurer, A. J.; Shaffer, J. B.;
Duffin, T. O.; LaBarte, T. L.; Maddock, M. L.; Wheeler, O. W.
Tetrahedron Lett. 2012, 53, 1275. (b) Fujita, K.; Yamamoto, K,;
Yamaguchi, R. Org. Lett. 2002, 4, 2691. (c) Cropper, E. L.; White,
A. J. P.; Ford, A.; Hii, K. K. J. Org. Chem. 2006, 71, 1732. (d)
Qadir, M.; Priestley, R. E.; Rising, T. W. D. F.; Gelbrich, T.;
Coles, S. J.; Hursthouse, M. B.; Sheldrake, P. W.; Whitall, N.; Hii,
K. K. Tetrahedron Lett. 2003, 44, 3675.
8. For reviews, see: (a) Nolan, S. P.; Clavier, H. Chem. Soc. Rev.
2010, 39, 3305. (b) Chattopadhyay, S. K.; Karmakar, S.; Biswas,
T.; Majumdar, K. C. Rahaman, H.; Roy, B. Tetrahedron 2007, 63,
3919. (c) Hasan, H. M. A.; Chem. Commun. 2010, 46, 9100.
9. For recent examples, see: (a) Ramachary, D. B.; Narayana, V. V.
Eur. J. Org. Chem. 2011, 3514. (b) Ghosh, D.; Thander, L.;
Ghosh, S. K.; Chattopadhyay, S. K. Synlett 2008, 3011. (c) Kotha,
S.; Shah, V. R. Eur. J. Org. Chem. 2008, 6, 1054.
References and notes
1. (a) Lovering, F.; Bikker, J.; Humblet, C. J. Med. Chem. 2009, 52,
6752. (b) Lachance, H.; Wetzel, S.; Kumar, K.; Waldmann, H. J.
Med. Chem. 2012, 55, 5989.
2. Moss, T. A.; Hayter, B. R.; Hollingsworth, I, A.; Nowak, T.
Synlett 2012, 23, 2408.
10. For examples of pyrimido-fused azepines prepared via RCM, see:
(a) Kaïm, L. E.; Grimaud, L.; Oble, J. J. Org. Chem. 2007, 72,
5835. (b) Majumdar, K. C.; Mondal, S.; Ghosh, D. Synthesis 2010,
1176.
11. Mosrin, M.; Knochel, P. Chem. Eur. J. 2009, 15, 1468.
12. Li, S-R.; Chen, H-M.; Chen, L-Y.; Tsai, J-C.; Chen, P-Y.; Hsu, S.
C-N.; Wang, E-C. Arkivoc, 2008, (ii), 172.
3.
(a) Ogawa, H.; Yamashita, H.; Kondo, K.; Yamamura, Y.;
Miyamoto, H.; Kan, K.; Kitano, K.; Tanaka, M.; Nakaya, K.;
Nakamura, S.; Mori, T.; Tominaga, M.; Yabuuchi, Y. J. Med.
Chem. 1996, 39, 3547. (b) Torisawa, Y.; Furuta, T.; Nishi, T.;
Aki, S.; Minamikawa, J. Bioorg. Med. Chem. Lett. 2007, 17, 6455.
(c) Kondo, K.; Ogawa, H.; Yamashita, H.; Miyamoto, H.; Hisashi,
T.; Tanaka, M.; Nakaya, K.; Kitano, K.; Yamamura, Y.;
Nakamura, S.; Onogawa, T.; Mori, T.; Tominaga, M. Bioorg.
Med. Chem. 1999, 7, 1743.
13. Lane, C.; Snieckus, V. Synlett 2000, 1294.
14. (a) Shi, M.; Li, C-Q.; Jiang, J-K. Helv. Chim. Acta, 2002, 85,
1051. (b) Sato, S.; Matsuda, I.; Shibata, S. J. Organomet. Chem.
1989, 347.
4. (a) Annedi, S. C.; Ramnauth, J.; Cossette, M.; Maddaford, S. P.;
Dove, P.; Rakhit, S.; Andrews, J. S.; Porreca, F. Bioorg. Med.
Chem. Lett. 2012, 22, 2510. (b) Jake Slavish, P.; Jiang, Q.; Cui.

5
15. Representative procedure for the synthesis of 20: Grubbs 2nd
generation catalyst (63.6 mg, 0.075 mmol, 5 mol%) was added to
a
solution of tert-butyl allyl[5-allyl-6-chloro-2-(methylthio)
pyrimidin-4-yl]carbamate (2) (532.7 mg, 1.50 mmol) in anhydrous
CH2Cl2 (15 mL) at room temperature. The mixture was heated at
reflux for 2 h. After cooling to room temperature, TFA (5 mL)
was added and the mixture was stirred for a further 30 min before
the volatiles were removed under vacuum. The residue was
dissolved in CH2Cl2 (15 mL) and washed with sat. NaHCO3
solution (15 mL), then dried over Na2SO4 and concentrated. The
crude product was purified by silica gel chromatography
(heptane/EtOAc, 9:1 to 3:1) to give 4-chloro-2-(methylthio)-8,9-
dihydro-5H-pyrimido[4,5-b]azepine 20 as
a colourless solid
(283.2 mg, 83%). M.p. 129-130 oC; νmax (cm-1) 3425 (br), 3287
(m), 1580 (s), 1525 (s), 1386 (m), 1335 (m), 1302 (m), 1260 (m),
1
1100 (w); H NMR [400 MHz, CDCl3] 6.16 (dt, 1 H, J = 6.9, 9.6
Hz, CH), 6.00 (dt, 1 H, J = 6.8, 9.7 Hz, CH), 6.00 (br s, 1 H, NH),
4.00 (dd, 2 H, J = 5.0, 6.7 Hz, CH2), 3.57 (d, 2 H, J = 6.9 Hz,
CH2), 2.44 (s, 3 H, CH3); 13C NMR [100 MHz, CDCl3] 168.24,
163.25, 157.58, 131.64, 127.33, 106.58, 39.67, 25.22, 13.95; m/z
(ES) 228 ([M+H]+, 100); HRMS (EI) m/z [M+H]+ calcd for
C9H11N3ClS: 228.03567; found; 228.03555,