Synthesis and ring openings of cinnamate-derived N-unfunctionalised aziridines

Article abstract of DOI:10.3762/bjoc.8.199

Tert-Butyl cinnamates are aziridinated with high trans-selectivity by an N-N ylide generated in situ from N-methylmorpholine and O-diphenylphosphinyl hydroxylamine. The resulting N-unfunctionalised aziridines are shown to be versatile synthetic building b

Full text of DOI:10.3762/bjoc.8.199

Synthesis and ring openings of cinnamate-derived
N-unfunctionalised aziridines
Alan Armstrong* and Alexandra Ferguson
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2012, 8, 1747–1752.
Department of Chemistry, Imperial College London, South
Kensington, London, SW7 2AZ, U.K.
doi:10.3762/bjoc.8.199
Received: 24 July 2012
Email:
Accepted: 13 September 2012
Published: 12 October 2012
Alan Armstrong* - A.Armstrong@imperial.ac.uk
* Corresponding author
This article is part of the Thematic Series "Organocatalysis".
Guest Editor: B. List
Keywords:
amino acids; aziridination; organocatalysis; ring opening;
stereoselectivity
© 2012 Armstrong and Ferguson; licensee Beilstein-Institut.
License and terms: see end of document.
Abstract
tert-Butyl cinnamates are aziridinated with high trans-selectivity by an N–N ylide generated in situ from N-methylmorpholine and
O-diphenylphosphinyl hydroxylamine. The resulting N-unfunctionalised aziridines are shown to be versatile synthetic building
blocks that undergo highly selective ring-opening reactions with a wide range of nucleophiles.
Introduction
Amino acids are the fundamental building blocks of life on
earth, and consequently their molecular structure is ubiquitous
in natural and synthetic chemistry [1]. An attractive method of
installing the amino-acid functionality is through the regioselec-
tive ring opening of an aziridine-2-carboxylate (Scheme 1)
[2,3].
Scheme 1: Regioselective ring opening of an aziridine-2-carboxylate.
Such a methodology has been developed for N-functionalised
aziridines [4-9], as N-functionality offers both activation of the
aziridine to ring opening [10], and protection of the nitrogen. been developed, synthesis of the starting NH-aziridine-2-
However, an extra synthetic step may be required to remove the carboxylates is multistep and low-yielding. Synthetic routes
N-functionality, which is often difficult and low yielding [11- include Gabriel–Cromwell addition of ammonia to an enoate
15]. Ring opening of N-unfunctionalised (NH-) aziridines [16,17]; epoxide opening with azide, followed by ring closure
allows direct access to amino-acid derivatives, yet the published [19]; and a sequence of olefin dihydroxylation, conversion to a
methodology is currently limited for this transformation [16- cyclic sulfate, ring opening with azide, and finally ring closure
20]. In literature examples where ring-opening protocols have to afford the NH-aziridine [18].
1747

Beilstein J. Org. Chem. 2012, 8, 1747–1752.
We recently reported [21,22] a nucleophilic aziridination We also report here our studies into the ring opening of aziri-
methodology [23-30] that allows access to NH-aziridines in a dine 1a, derived from tert-butyl cinnamate. As a result, a
single step from α,β-unsaturated carbonyl compounds. Our diverse array of amino-acid derivatives is generated in three
method produces trans-NH-aziridines from α,β-unsaturated steps from commercially available starting materials.
carbonyl compounds in good to excellent yield with high
diastereoselectivity (Figure 1) [21]. An N–N ylide (aminimine) Results and Discussion
is the presumed active intermediate, formed from the reaction of The aziridination of tert-butyl cinnamate was the starting point
a tertiary amine promoter and an aminating agent. Conjugate chosen for our optimisation. A control reaction in the absence
addition of the ylide, followed by a ring-closure step affords the of NMM gave no aziridine product. Despite the low conversion
NH-aziridine. More recently, the methodology has been in the reported procedure [21], the only compounds apparent in
extended to α,β,γ,δ-unsaturated carbonyl compounds, with the 1H NMR spectrum of the crude reaction mixture were the
excellent regio- and diastereoselectivity [31].
starting cinnamate and the aziridine. A time delay of 20 minutes
between the addition of NaOH and the cinnamate to the reac-
Herein, we report the optimisation of our nucleophilic aziridina- tion mixture led to increased conversion (Table 1, entry 1
tion methodology to access NH-aziridine-2-carboxylates in a versus entry 2), with longer times leading to lower conversions
single step from enoates. To the best of our knowledge, there (Table 1, entry 3 and entry 4). Using 4 equiv of NaOH
exist only three examples of this transformation. Furukawa used increased the conversion to 50% (Table 1, entry 5), although a
diphenyl sulfilimine to form NH-aziridines from fumarate esters large excess (20 equiv) did not increase the conversion any
[32], and O-mesitylenesulfonylhydroxylamine (MSH) was used further (Table 1, entry 5 versus entry 6). Increasing the concen-
to aziridinate electron-poor alkenes by Hamelin [33]. We tration of the reaction mixture was found to have the most
described the aziridination of two tert-butyl enoate substrates in significant effect; performing the reaction at a concentration
our initial disclosure, but arguably the more synthetically useful of 0.24 M provided a 63% conversion to aziridine (Table 1,
of these, tert-butyl cinnamate, gave a modest yield of 32% [21]. entry 8). Problems with mechanical stirring were observed at
Figure 1: Amine-promoted alkene aziridination.
1748

Beilstein J. Org. Chem. 2012, 8, 1747–1752.
concentrations above 0.24 M, due to the heterogeneous nature and para-halo-substituted cinnamates provided optimum yields
of the reaction mixture. The isolated yield of this reaction when subjected to conditions (A), using fewer equiv of reagents
(51%) was slightly lower than the conversion, but was repro- (Table 2, entries 4–6, 8, 10). Poor isolated yields were observed
ducible on a 12 mmol scale (Table 1, entry 9). Use of an when conditions (B) were used on electron-poor substrates; this
increased number of equiv of the aminating agent, tertiary was attributed to the increased susceptibility to hydrolysis of the
amine and NaOH provided the aziridine in 65% isolated yield aziridine products. The increased steric demand posed by ortho
on a 3 mmol scale (Table 1, entry 10). The increased equiv of substitution was found to have a detrimental effect on the
the heterogeneous reagents meant that the concentration of the aziridination (Table 2, entry 6 versus entry 8). A 5-chloro-furyl-
reaction was reduced to 0.12 M for these conditions. The reac- substituted enoate was also subjected to the aziridination
tion was found to be specific for tert-butyl esters, with γ- and (Table 2, entry 10); a shorter reaction time of 20 h under the
δ-benzylidene lactones, methyl, ethyl, phenyl, and benzyl cinna- moderate set of aminating conditions (A) was optimal for this
mates all suffering from degradation, presumably via hydroly- substrate. In all cases, only starting material and product were
sis pathways. β-Alkyl-substituted enoates were not explored, observed in the 1H NMR spectrum of the crude reaction mix-
and will be the subject of future investigation. A preliminary ture. However, a loss in mass recovery was often observed
screen of alternative solvents, tertiary amines and bases showed during the isolation/purification process (see Table 2, isolated
that NMM and NaOH in CH2Cl2 was the optimum reagent yield and recovered starting material columns), meaning that
combination for the aziridination.
many of the isolated yields of aziridine are moderate. Nonethe-
less, it is worth noting that this single-step transformation
Having optimised our aziridination protocol for tert-butyl provides NH-aziridine-2-carboxylates in competitive or superi-
cinnamate, our attention then turned to examining the scope for or yields to alternative multistep methodologies.
aromatic substitution in the aziridination. A range of ortho-,
meta- and para-substituted cinnamates were synthesised via Having explored the scope of the aziridination, we were keen to
Wittig, Heck or esterification routes (for details, see Supporting investigate the susceptibility of our unactivated NH-aziridines
Information File 1). Each substrate was subjected to two sets of to ring-opening chemistry. To the best of our knowledge, only
aziridination conditions, with the optimum yields reported.
four papers exist detailing the regioselective ring opening of
NH-aziridine-2-carboxylates to provide amino-acid derivatives
Electron-rich substrates and the ortho-chloro cinnamate [16-19]. The substrate scope has been limited to NH-aziridine-
favoured conditions (B), in which increased equiv of reagents 2-carboxylates derived from isopropyl [16], menthyl [17],
were used (Table 2, entries 2, 3, 7, 9). Electron-poor substrates benzyl [18], and methyl [19] cinnamates. Nucleophiles are
Table 1: Enoate aziridination optimisation.a
Entry
NaOH (equiv)
Concentrationb (M)
Time for step ii (min)
Conversion (%)c,d
1
2
0.06
0.06
0.06
0.06
0.06
0.06
0.12
0.24
0.24
0.12
0
32 (30)
47
2
2
20
30
160
20
20
20
20
20
20
3
2
41
4
2
23
5
4
50
6
20
4
47
7
58
8
4
63 (51)
63 (50)
(65)
9e
10f
4
11
aAll reactions performed on 0.11 mmol scale unless stated otherwise. bWith respect to tert-butyl cinnamate. c Estimated by 1H NMR: relative integra-
tion of tert-butyl peaks in starting material and product. dIsolated yield in parentheses. e12 mmol scale. fNMM (3 equiv), DppONH2 (3 equiv), 3 mmol
scale.
1749

Beilstein J. Org. Chem. 2012, 8, 1747–1752.
Table 2: Substrate scope of enoate aziridination.
Entry
Ar
Conditionsa
Isolated yield (%)
R.S.Mb (%)
1c
2
1a
1b
1c
1d
1e
1f
Ph
B
B
B
A
A
A
B
A
B
Af
65
39
36
33
53
76
47
40
32
35
6
(4-OMe)C6H4
(4-Me)C6H4
(4-I)C6H4
34
12
40
12
6
3
4d
5e
6
(4-Cl)C6H4
(4-NO2)C6H4
(2-Cl)C6H4
(2-NO2)C6H4
(3-OMe)C6H4
(5-Cl)furyl
7
1g
1h
1i
5
8
24
14
44
9
10
1j
aConditions A: NMM (1.04 equiv), DppONH2 (1.04 equiv), NaOH (4 equiv), 0.24 M; Conditions B: NMM (3 equiv), DppONH2 (3 equiv), NaOH
(11 equiv), 0.12 M. All reactions performed on 0.20–0.23 mmol scale. bRecovered starting material (cinnamate ester) following purification.
c2.90 mmol scale. d0.15 mmol scale at 0.08 M. eReaction performed at 0.06 M. f20 h reaction.
limited to thiophenols [17-19], fluoride [16], indole [18,19] and The regiochemistry of 2a was established through correlation
acetic acid [18,19] and a hydrogenation protocol also exists spectroscopy. A 2J correlation between the ipso-carbons of both
[18]. We were keen to test these nucleophiles in the ring aromatic systems and the C3 proton was observed by HMBC
opening of our NH-aziridines, in order to provide products with NMR. Comparison of the 1H NMR chemical shifts of the minor
alternative ester functionality in the form of a tert-butyl group. product with 2a and literature compounds led to the tentative
Additionally, we wanted to expand the range of ring-opening assignment of the minor product as the C3 ring-opened dia-
nucleophiles, to provide access to a wider variety of amino-acid stereomer of 2a. This reaction provides an attractive example of
derivatives.
how our ring-opening chemistry can be used to form new
carbon–carbon bonds.
Aziridine 1a was selected for initial investigations. Pleasingly,
aziridine 1a was found to undergo ring opening under Lewis Hruby reported the hydrogenation of the benzyl ester analogue
and Brønsted acid catalysed conditions with a range of nucle- of aziridine 1a, resulting in the concomitant removal of the
ophiles to provide amino-acid derivatives in good to excellent benzyl group to provide free phenylalanine [18]. Pleasingly,
yield (Figure 2). The ring openings proved to be completely treatment of aziridine 1a with H2 in the presence of Pd/C
regioselective for the C3-benzylic position of aziridine 1a. This provides phenylalanine tert-butyl ester (2b) in an excellent yield
regioselectivity was confirmed through atom correlations of 95%, without the need for purification. The orthogonal
observed in 2D NMR experiments and by comparison with protection provided by the tert-butyl ester yields a product that
literature compounds, including C2-substituted analogues.
is easier to handle than the free amino acid.
Indole was the first nucleophile investigated, as regioselective The ring opening of NH-aziridine-2-carboxylates with alco-
ring opening of the analogous benzyl [18] and methyl [19] ester holic nucleophiles is almost unexplored. To the best of our
protected NH-aziridines with indole in the presence of knowledge, only one literature example exists, in which the
BF3·OEt2 had previously been reported, in 48 and 53% yields, NH-aziridine derived from dimethyl fumarate is subjected to
respectively. On subjecting the orthogonally protected aziridine ring opening with MeOH in 50% yield [20]. We report that the
1a to the reported conditions, only a 7% isolated yield of ring- ring opening of aziridine 1a with alcohols in the presence of
opened product was observed. Pleasingly, the use of BF3·THF camphorsulfonic acid (CSA), provides a facile synthesis of 1,2-
resulted in a considerably cleaner reaction. Separation of a aminoalcohols. The alcohol nucleophiles were used as solvents,
major and minor product was achieved, with the major product as stoichiometric quantities resulted in decreased conversions.
identified as the tryptophan derivative 2a, isolated in 63% yield. Single ring-opened products were observed in the 1H NMR
1750

Beilstein J. Org. Chem. 2012, 8, 1747–1752.
Figure 2: Ring-opened products of aziridine 1a.
spectra of the crude products in all cases. In MeOH, amino-acid Finally, β-chlorophenylalanine derivatives were accessed in
derivative 2c was obtained in quantitative yield, without the three different forms, depending on the choice of HCl source
need for purification. This represents the first regioselective and work-up conditions employed. Treatment of aziridine 1a
example of the ring opening of an NH-aziridine-2-carboxylate with HCl (35% aq) in dichloromethane led to isolation of 2g,
with methanol. Allyl alcohol gave access to amino-acid deriva- i.e., the HCl salt of β-chlorophenylalanine, in 45% yield. It is
tive 2d (57%). The terminal olefin provides a useful handle for notable that this is the only one of our examples in which the
further functionalisation of this substrate. Reaction of benzyl tert-butyl ester is cleaved to provide the amino-acid motif.
alcohol yielded the β-benzyloxy amino-acid derivative 2e in Altering the HCl source to 4 M HCl in dioxane provided pro-
46%. The orthogonal benzyl protection could allow the free duct 2h as the HCl salt, with the tert-butyl ester intact. Treat-
alcohol to be accessed selectively by hydrogenation protocols, ment with aqueous NaHCO3 provided pure β-chloroamine 2i in
leaving the tert-butyl ester intact. Products 2d and 2e are the 90% yield over two steps. The chlorine atom provides a useful
first examples of the ring opening of NH-aziridine-2-carboxy- handle for further functionalisation, via substitution or coupling
lates with alcoholic nucleophiles other than MeOH.
reactions, for example [34,35]. This is the first reported regio-
and diastereoselective synthesis of NH-β-chlorophenylalanine
derivatives from NH-aziridine-2-carboxylates.
Acetic acid has previously been used as a nucleophile to
achieve ring opening of benzyl [18] and methyl-ester [19]
substituted NH-aziridines in 88% and 69% yields, respectively, Conclusion
with O- to N-acetate transfer observed. Ring opening of 1a in In summary, we have reported the synthesis of a range of
acetic acid at 70 °C provides pure product 2f (79%), with amino-acid derivatives from the ring opening of an NH-aziri-
orthogonal tert-butyl ester protection in contrast to the reported dine in excellent regio- and diastereoselectivity. NH-aziridines
examples. As detailed in the literature, the acetate protecting were synthesised in a single step from tert-butyl cinnamates,
group was found to transfer to the nitrogen of the ring-opened following optimisation of our nucleophilic aziridination
aziridine. Product 2f can be considered complementary to 2e; methodology. This allows access to amino-acid derivatives in
O- or N-protected amino-acid derivatives can be accessed three steps from commercially available starting materials.
easily, depending on the choice of conditions.
Efforts are currently underway to develop an enantioselective
1751

Beilstein J. Org. Chem. 2012, 8, 1747–1752.
19.Legters, J.; Thijs, L.; Zwanenburg, B. Recl. Trav. Chim. Pays-Bas
1992, 111, 16–21. doi:10.1002/recl.19921110102
20.Antolini, L.; Bucciarelli, M.; Caselli, E.; Davoli, P.; Forni, A.; Moretti, I.;
Prati, F.; Torre, G. J. Org. Chem. 1997, 62, 8784–8789.
doi:10.1021/jo971285+
version of the aziridination that is applicable to tert-butyl
enoates.
Supporting Information
21.Armstrong, A.; Baxter, C. A.; Lamont, S. G.; Pape, A. R.; Wincewicz, R.
Org. Lett. 2007, 9, 351–353. doi:10.1021/ol062852v
22.Shen, Y.-M.; Zhao, M.-X.; Xu, J.; Shi, Y. Angew. Chem., Int. Ed. 2006,
45, 8005–8008. doi:10.1002/anie.200602586
Supporting Information File 1
Experimental procedures and characterisation data for all
new NH-aziridines, including 1H and 13C NMR spectra for
all new compounds.
23.Xu, J.; Jiao, P. J. Chem. Soc., Perkin Trans. 1 2002, 1491–1493.
doi:10.1039/B200046F
24.Pesciaioli, F.; De Vincentiis, F.; Galzerano, P.; Bencivenni, G.;
Bartoli, G.; Mazzanti, A.; Melchiorre, P. Angew. Chem., Int. Ed. 2008,
47, 8703–8706. doi:10.1002/anie.200803647
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-8-199-S1.pdf]
25.Sasaki, M.; Dalili, S.; Yudin, A. K. J. Org. Chem. 2003, 68, 2045–2047.
doi:10.1021/jo020696+
Acknowledgements
We thank the EPSRC for their support of this work.
26.Cromwell, N. H.; Barker, N. G.; Wankel, R. A.; Vanderhorst, P. J.;
Olson, F. W.; Anglin, J. H. J. Am. Chem. Soc. 1951, 73, 1044–1051.
doi:10.1021/ja01147a048
27.Vesely, J.; Ibrahem, I.; Zhao, G.-L.; Rios, R.; Córdova, A.
Angew. Chem., Int. Ed. 2007, 46, 778–781.
References
1. Cardillo, G.; Tomasini, C. Chem. Soc. Rev. 1996, 25, 117–128.
doi:10.1002/anie.200603810
doi:10.1039/cs9962500117
28.Armstrong, A.; Carbery, D. R.; Lamont, S. G.; Pape, A. R.;
Wincewicz, R. Synlett 2006, 2504–2506. doi:10.1055/s-2006-950425
29.Chen, G.; Sasaki, M.; Yudin, A. K. Tetrahedron Lett. 2006, 47,
255–259. doi:10.1016/j.tetlet.2005.11.039
2. Yudin, A. K., Ed. Aziridines and Epoxides in Organic Synthesis;
Wiley-VCH: Weinheim, Germany, 2006.
3. Hu, X. E. Tetrahedron 2004, 60, 2701–2743.
doi:10.1016/j.tet.2004.01.042
30.Wang, X.; Ding, K. Chem.–Eur. J. 2006, 12, 4568–4575.
doi:10.1002/chem.200501109
4. Schäfer, A.; Henkensmeier, D.; Kröger, L.; Thiem, J.
Tetrahedron: Asymmetry 2009, 20, 902–909.
31.Armstrong, A.; Pullin, R. D. C.; Jenner, C. R.; Scutt, J. N. J. Org. Chem.
2010, 75, 3499–3502. doi:10.1021/jo100407s
doi:10.1016/j.tetasy.2009.02.038
5. Forbeck, E. M.; Evans, C. D.; Gilleran, J. A.; Li, P.; Joullié, M. M.
J. Am. Chem. Soc. 2007, 129, 14463–14469. doi:10.1021/ja0758077
6. Kelley, B. T.; Joullié, M. M. Org. Lett. 2010, 12, 4244–4247.
doi:10.1021/ol101584z
32.Furukawa, N.; Oae, S.; Yoshimura, T. Synthesis 1976, 30–32.
doi:10.1055/s-1976-23945
33.Métra, P.; Hamelin, J. J. Chem. Soc., Chem. Commun. 1980,
1038–1039. doi:10.1039/C39800001038
7. Legters, J.; Willems, J. G. H.; Thijs, L.; Zwanenburg, B.
Recl. Trav. Chim. Pays-Bas 1992, 111, 59–68.
34.Hendriks, W.; Woolhouse, A.; Tarttelin, M.; Moughan, P. Bioorg. Chem.
1995, 23, 89–100. doi:10.1006/bioo.1995.1007
doi:10.1002/recl.19921110201
35.Chuang, T.-H.; Sharpless, K. B. Org. Lett. 2000, 2, 3555–3557.
doi:10.1021/ol000221+
8. Ammetto, I.; Gasperi, T.; Loreto, M. A.; Migliorini, A.; Palmarelli, F.;
Tardella, P. A. Eur. J. Org. Chem. 2009, 6189–6197.
doi:10.1002/ejoc.200900891
9. Taylor, A. M.; Schreiber, S. L. Tetrahedron Lett. 2009, 50, 3230–3233.
doi:10.1016/j.tetlet.2009.02.034
10.Sweeney, J. B. Chem. Soc. Rev. 2002, 31, 247–258.
doi:10.1039/b006015l
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
11.Vedejs, E.; Klapars, A.; Warner, D. L.; Weiss, A. H. J. Org. Chem.
2001, 66, 7542–7546. doi:10.1021/jo0106243
12.Patwardhan, A. P.; Lu, Z.; Pulgam, V. R.; Wulff, W. D. Org. Lett. 2005,
7, 2201–2204. doi:10.1021/ol050597t
(http://creativecommons.org/licenses/by/2.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
13.Fioravanti, S.; Morreale, A.; Pellacani, L.; Tardella, P. A. C. R. Chim.
2005, 8, 845–847. doi:10.1016/j.crci.2005.02.010
14.Fioravanti, S.; Massari, D.; Morreale, A.; Pellacani, L.; Tardella, P. A.
Tetrahedron 2008, 64, 3204–3211. doi:10.1016/j.tet.2008.01.098
15.Tarrade-Matha, A.; Valle, M. S.; Tercinier, P.; Dauban, P.; Dodd, R. H.
Eur. J. Org. Chem. 2009, 673–686. doi:10.1002/ejoc.200801000
16.Wade, T. N.; Gaymard, F.; Guedj, R. Tetrahedron Lett. 1979, 20,
2681–2682. doi:10.1016/S0040-4039(01)86385-1
17.Ploux, O.; Caruso, M.; Chassaing, G.; Marquet, A. J. Org. Chem. 1988,
53, 3154–3158. doi:10.1021/jo00249a006
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(http://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
18.Xiong, C.; Wang, W.; Cai, C.; Hruby, V. J. J. Org. Chem. 2002, 67,
1399–1402. doi:10.1021/jo010860d
doi:10.3762/bjoc.8.199
1752