Alternative tandem cyclisation pathways in the reaction between imines and enones

Article abstract of DOI:10.1016/j.tet.2016.01.006

Dihydroisoquinoline reacts with Danishefsky's diene under Lewis acidic conditions or neat, to give low to moderate yields of the formal aza-Diels-Alder, [4+2]-cycloadduct. However, using methoxyvinyl methylketone with Lewis acid catalysis does not give the aza-Diels-Alder adduct, rather a formal [2+2+2]-cycloaddition occurs to provide access to a diacetyl dihydropyridine. Increased Lewis acid loading results in reduced dihydropyridine formation, and instead, a trimerisation reaction of the methoxyvinyl methyl ketone occurs, to give 1,3,5-triacetylbenzene from a different formal [2+2+2]-cycloaddition. The formal [4+2]-cycloaddition reaction of methoxyvinyl methylketone requires a cyclic imine in order to form the dihydropyridine because the reaction with acyclic imines produced a dihydropyridine from a formal [1+2+1+2]-cycloaddition. Evidence resulting from the isolation of reaction intermediates and in situ spectroscopic studies, shows that the reaction between 3,4-dihydroisoquinoline and methyl vinyl ketone, catalysed by oxy-philic Lewis acids, proceeds via a Mannich-Michael pathway and an imminium ion species. All reactions occur by one-pot cascade routes.

Full text of DOI:10.1016/j.tet.2016.01.006

Tetrahedron xxx (2016) 1e9
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Alternative tandem cyclisation pathways in the reaction between
imines and enones
P. Ricardo Girling a, Andrei S. Batsanov a, Adam D.J. Calow a, Hong C. Shen b,
,
Andrew Whiting a
*
a Centre for Sustainable Chemical Processes, Department of Chemistry, Durham University, South Road, Durham, DH1 3LE, UK
b Merck & Co., RY800-C114, 126 East Lincoln Avenue, PO Box 2000, Rahway, NJ 07065-0900, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Dihydroisoquinoline reacts with Danishefsky’s diene under Lewis acidic conditions or neat, to give low to
moderate yields of the formal aza-DielseAlder, [4þ2]-cycloadduct. However, using methoxyvinyl
methylketone with Lewis acid catalysis does not give the aza-DielseAlder adduct, rather a formal
[2þ2þ2]-cycloaddition occurs to provide access to a diacetyl dihydropyridine. Increased Lewis acid
loading results in reduced dihydropyridine formation, and instead, a trimerisation reaction of the me-
thoxyvinyl methyl ketone occurs, to give 1,3,5-triacetylbenzene from a different formal [2þ2þ2]-
cycloaddition. The formal [4þ2]-cycloaddition reaction of methoxyvinyl methylketone requires a cyclic
imine in order to form the dihydropyridine because the reaction with acyclic imines produced a dihy-
dropyridine from a formal [1þ2þ1þ2]-cycloaddition. Evidence resulting from the isolation of reaction
intermediates and in situ spectroscopic studies, shows that the reaction between 3,4-
dihydroisoquinoline and methyl vinyl ketone, catalysed by oxy-philic Lewis acids, proceeds via a Man-
nich-Michael pathway and an imminium ion species. All reactions occur by one-pot cascade routes.
Ó 2016 Elsevier Ltd. All rights reserved.
Received 19 November 2015
Received in revised form 24 December 2015
Accepted 5 January 2016
Available online xxx
Keywords:
Aza-DielseAlder
Lewis-acid
Homogeneous catalysis
Cycloaddition
Dihydropyridine
Piperidenone
1. Introduction
The mechanism involved is potentially a concerted [4þ2]-
cycloaddition, however, in most cases it actually proceeds
Natural products containing the piperidine ring system1 are well
known to be biologically active2 and many analogues have been
developed for their potential medicinal properties.3 The synthesis
of these types of six-membered nitrogen heterocycles can be ach-
ieved via a Lewis acid-catalysed, formal aza-DielseAlder reaction
involving an imino dienophile and a conjugated diene4 as outlined
in Scheme 1 (e.g., using electron rich siloxy dienes such as Dani-
shefsky’s diene 1).5
through a step-wise Mannich-Michael reaction.6 Nevertheless, the
formal DielseAlder process is a powerful synthetic strategy for
accessing piperidinones 3, from, which more complex piperidines 4
can be accessed.7
As part of our ongoing development of novel approaches to
piperidinones and their derivatives, typically employing Lewis acid
catalysis8 and associated mechanistic observations,9 we have been
examining the reactions of acyclic electron deficient imines with
various dienes. In addition, we recognised the potential of this
approach for the synthesis of polycyclic nitrogen-containing het-
erocycles of general type 6, starting from cyclic imines. Heterocy-
cles of this type occur naturally in representative structures with
varied biological activity10 and, hence, are worthy of further study
in order to access formal aza-DielseAlder adducts for elaboration to
more substituted targets.
Scheme 1. A general approach to poly-substituted piperidines.
In this paper, we report the synthesis of N-heterocyclic com-
pounds from imines and enones via different tandem cyclisation
pathways. The different modes of reaction that we investigated
include formal [2þ2þ2], [1þ2þ1þ2] and [4þ2] cascade cyclisations
(see Scheme 2).
* Corresponding author. E-mail address: andy.whiting@durham.ac.uk (A.
Whiting).
http://dx.doi.org/10.1016/j.tet.2016.01.006
0040-4020/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006

2
P.R. Girling et al. / Tetrahedron xxx (2016) 1e9
Table 1
Effect of catalyst loading on the reaction of 5 with 7 (Eq. 2)
Entry
Lewis acid (mol %)
Yield 8 (%)a
Yield 9 (%)b
1
2
3
4
5
0
5
10
20
100
0
d
<10b
61
48
0
d
d
Scheme 2. The synthesis of N-heterocyclic compounds from imines and enones via
different tandem cyclisation pathways.
<5
>55
a
Isolated yield after silica gel chromatography.
b
2. Results and discussion
Conversion estimated from the crude 1H NMR spectrum.
The high symmetry of the impurity meant that it was readily
2.1. Synthesis of piperidenone and dihydropyridine analogues
identified12 as 1,3,5-triacetylbenzene 9 (the structure was also
confirmed by X-ray crystallographic studies, see Supplementary
data); a structure, which has been recently reported to readily as-
semble by heating 4-methoxy-3-buten-2-one 7 in water at 150 ꢀC.13
The results in Table 2 also show that the formation of trimer 9 was
favoured over the formation of dihydropyridine 8 with increasing
Lewis acid catalyst loading. Without the addition of Lewis acid, no
reaction occurred (Table 1, Entry 1). With low Lewis acid catalyst
loading (5%), low conversion to adduct 8 was observed (Table 1,
entry 2). Optimal formation of dihydropyridine 8 occurred at
10 mol % catalyst loading, with no trimer 9 being produced (Table 1,
entry 3). With 20 mol % catalyst loading, trimer 9 started to appear
(Table 1, Entry 4), though to obtain complete conversion to the
trimer alone, stoichiometric Lewis acid seemed to be required
(Table 1, Entry 5).
As can be observed from Table 2, relatively low to moderate
yields of the formal [2þ2þ2]-cycloadduct 8 were obtained, i.e., in
the 20e40% range, though in a convenient manner and a pure form,
through direct trituration of the crude reaction product. Despite the
low yields obtained (Table 2), these studies provided some addi-
tional useful information about the catalytic process. The use of
either Sc(OTf)3, Yb(OTf)3, or its hydrate as catalysts made little
difference to the isolated yields (Table 2, entries 1, 3 and 5), with or
Our starting point for the synthesis of piperidinones of type 3
was to examine the racemic formation of tricyclic piperidinone 6 as
a prelude to developing a catalytic asymmetric route. Hence, we
examined the reaction of Danishefsky’s diene 1 under Lewis acid-
catalysed conditions with readily accessible11 imine 5 (Eq. 1).
Danishefsky’s diene reacted on its own with the imine, most
likely initiated via nucleophilic attack of the diene on the imine,
followed by cyclisation, i.e., through a Mannich-Michael mech-
anism.6 After hydrolysis and purification, piperidinone 6 could
be isolated in up to 47% yield. However, none of these reactions
were completely clean according to TLC analysis, and therefore,
an improved methodology to systems of type
required.
3 was still
It was considered that direct reaction of 4-methoxy-3-buten-2-
one 7 with imine 5 in the presence of a suitable catalyst might
induce formation of an enolate equivalent in situ. This might be
expected to cyclise to give the desired cycloadduct, hence, avoid-
ing the use of Danishefsky’s diene. This approach would be a step-
wise addition, cyclisation, elimination, which nonetheless would
accomplish the desired formal cycloaddition. Thus, exposure of 4-
methoxy-3-buten-2-one 7 under a range of Lewis acid and sec-
ondary amine-catalysed reaction conditions was examined as in
Eq. 2, resulting in complex mixtures of products. However, the use
of ytterbium(III) triflate as Lewis acid catalyst did not provide
piperidenone 6. Instead, the diacetyl-dihydropyridine 8 was ob-
tained in 20% yield; its structure was later proved by X-ray crys-
tallography (see Supplementary data). The identification of
product 8, derived from a formal [2þ2þ2]-cycloaddition, through
a cascade process, prompted us to examine this reaction in more
detail.
Table 2
Optimisation studies of the [2þ2þ2]-cycloaddition reaction to give 8, as in Eq. 2
Entry Enone Catalyst (mol %) Solvent Time Additive Yield 8 Yield 9
equiv 7
(days)
(%)a
(%)b
1
2
3
4
5
5
2
2
2
2
Yb(OTf3) (20)
Yb(OTf3) (20)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Yb(OTf3)
CH2Cl2
CDCl3
CHCl3
CHCl3
CHCl3
2
1
2
2
2
d
d
d
40
30
41
<5
<5
0
0
0
ꢀ
4 A MS 42
d
40
hydrate (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
Sc(OTf)3 (10)
6
7
8
9
3
4
5
4
4
4
4
4
4
4
4
4
4
4
CHCl3
CHCl3
CHCl3
EtOAc
MeOH
CH3CN
THF
1e2
2
2
3
3
3
3
3
3
3
3
3
3
d
d
d
d
d
d
d
d
d
d
d
40
50
20
20
15
15
30
21
19
21
20
15
0
0
0
0
0
0
0
0
0
0
0
0
0
0
10
11
12
13
14
15
16
17
18
19
2.2. Formal [2D2D2]-cycloaddition
Et2O
Hexane
CH2Cl2
Toluene
CHCl3
CHCl3
CHCl3
The unexpected observation of the formation of the [2þ2þ2]-
cycloaddition product 8 is almost unprecedented,10 and therefore,
further investigations into this reaction started by varying the
catalyst and its loading. It was found that scandium(III) triflate was
particularly effective for these reactions (see Eq. 2 and Table 1),
however, a side-product 9 was observed, especially at higher cat-
alyst loadings.
H2O
ꢀ
4 A MS 25
3
d
20
a
Isolated yield after work up by trituration.
b
Conversion estimated from the crude 1H NMR spectrum.
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006

P.R. Girling et al. / Tetrahedron xxx (2016) 1e9
3
without molecular sieves (Table 2, entries 3 and 4). Different sol-
vents also had no significant effect, except when using methanol or
acetonitrile, in which cases, the yields tended to be lower (Table 2,
compare entries 9, 12e16 with entries 10 and 11).
Interestingly, this cycloadduct 8 was also isolated in 58% yield
when using butyne-one 10 (Eq. 3) under scandium(III)-catalysed
conditions, demonstrating that this yne-one might be involved in
the reaction mechanism.
Having examined the origin of the unwanted by-product 9 in
order to prevent its formation, optimisation of the synthesis of
dihydropyridine 8 was explored, i.e., the effect of catalyst, catalyst
loading, enone equivalents, solvent, additives and work up pro-
cedure, as outlined in Table 2. Further reaction optimisation (Eq. 2)
was carried out by examining a wider-range of Lewis acids in order
to see if conversion could be improved (Table 3).
2.3. Reaction scope and alternative reaction pathways: a for-
mal [1D2D1D2] pathway to dihydropyridines
Table 3
Effect of different Lewis acids upon the formal [2þ2þ2]-cycloaddition reaction to
In order to further understand the formal [2þ2þ2]-
cycloaddition reaction that provided 8, the use of acyclic imines
was examined. Different acyclic imines 2 were reacted in the
presence of 10 mol % scandium(III) triflate in the expectation of
forming the diacetyl products 13. However, after purification by
silica gel chromatography, dihydropyridine isomers 14 were in-
stead obtained (Eq. 4 and Table 4),14 together with an acyclic pre-
cursor 15.
give 8, as in Eq. 2
Entry
Lewis acid (10 mol %)a
Yield of 8 (%)b
1
2
3
4
5
6
7
8
9
10
Yb(OTf)3
Sc(OTf)3
In(OTf)3
Eu(hfc)3
Eu(hfc)3 (60 ꢀC)
Sc(OTf)3 (þpybox 10 mol %)
Fe(OTf)3
Ga(OTf)3
Yb(OTf)3
61c
46c
49c
d
d
82d
88d
88d
87d
81d
In(OTf)3
a
All reactions carried out in CH2Cl2.
Isolated yield after purification by silica gel chromatography.
Purification eluent EtOAc.
b
c
Dihydropyridines 19 and 20 (Table 4, entries 1, 2) were un-
d
Purification eluent EtOAc:CH2Cl2 (4:1).
expected products, being formed by
a
formal [1þ2þ1þ2]-
cycloaddition and in a four-component reaction.14 Under these
reaction conditions, the starting imine must have hydrolysed,
freeing the amine and aldehyde to undergo the multi-component
reaction;15 related reactions have been reported by us.8,9 Further
evidence for the probable mechanism of this multi-component
assembly comes from the isolation of the Michael adduct 21,16
which results from the reaction between the free amine and
enone 7 (Table 4, entry 3). Acyclic imines appear to be substantially
more prone to hydrolysis to the amine and aldehyde, and can then
undergo the formal [1þ2þ1þ2]-cycloaddition mode. This suggests
that the system needs to be completely anhydrous for the [2þ2þ2]-
cycloaddition to occur. However, when the reaction was performed
under rigorously dry conditions, the major reaction products were
still those derived from the formal [1þ2þ1þ2]-cycloaddition,
rather than the [2þ2þ2]-product, showing that even under these
dryer conditions, the imine was still hydrolysed, though we cannot
rule out a related role for methanol in this reaction.
The results, shown in Table 3, demonstrate that the Lewis acids
that afforded the highest yields were Fe(OTf)3 and Ga(OTf)3 (Table
3, entries 7 and 8), whilst the chiral Lewis acid Eu(hfc)3 was in-
active, even at higher temperatures (Table 3, entries 4 and 5). In
addition, higher yields of 8 were obtained when the reaction
mixture was purified by silica gel chromatography (e.g., Table 4,
entries 1e3). This was further optimised to ensure good solubility
of 8 (Table 3, entries 6e10) to enable efficient elution of the product
off the column during purification. Overall, the highest yields of 8
obtained were 88% for entries 7 and 8 (Table 4), which, considering
the low yields obtained initially and the complexity of the reaction
mixtures, was very satisfactory. In fact, this yield equates to each of
the three new bonds being formed with 96% efficiency during the
formation of formal [2þ2þ2]-cycloadduct 8.
2.4. Developing the formal [4D2]-cycloaddition reaction
Having explored the formal [2þ2þ2]-cycloaddition, which was
Table 4
Reactions between acyclic imines 2 and enone 7, as in Eq. 4
Entry
1
Imine
Enone equiv.
Product (yield, %)a
possible due to the presence of a
b
-leaving group (e.g., 7), attention
was turned to enones without a leaving group in the
b
-position to
see if the formal [4þ2]-cycloadduct could be accessed, i.e., formerly
through a conjugated enol-ene system. To this end, methyl vinyl
ketone 22 (2 equiv) was reacted with imine 5 in the presence of
Yb(OTF)3 (20 mol %), as shown in Eq. 5.
4
16
19
2
3
4
17
18
20
21
2.5
This reaction proceeded to give a mixture of products, however,
the major products consisted of 23 (37%) and the formal aza-Diel-
seAlder [4þ2]-cycloadduct 24 (25%). It is important to note that
derivatives of type 24 are of considerable pharmacological interest
due to their biological activity.18
a
Isolated yield after purification by silica gel chromatography.
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006

4
P.R. Girling et al. / Tetrahedron xxx (2016) 1e9
reaction outlined in Eq. 6, and particularly, through the enantio-
controlled cyclisation of iminium species 23, either from in situ
generation or after isolation. Advances in asymmetric catalysis and
especially organocatalysis19 provided a number of possible options
worthy of exploration in order to develop the required asymmetric
cyclisation methodology.
Initially we investigated chiral secondary amines as potential
The reaction shown in Eq. 5 is significant in that as well as
catalysts for forming 24 in an enantioselective manner.19a,20
L-
providing easy access to biologically interesting compounds, it also
provides intermediates, which may give mechanistic insights into
this reaction (vide infra). Indeed, Eq. 5 also demonstrates the po-
tential for an efficient formal [4þ2]-cycloaddition reaction between
5 and 22, giving access to the desired product 24. The presence of
23 can be rationalised by a Michael-type reaction, whereby the
Proline (20 mol %) was employed in parallel with In(OTf)3 (20 mol
%) to afford the adduct 24 in 49% yield (see Eq. 7). An ee of 3% was
observed by chiral HPLC. However, this level of asymmetric in-
duction can be considered to be within experimental error, despite
clean product formation and HPLC chromatogram.
nucleophilic imine 5 adds to the electropositive b-carbon of 22. The
simple interpretation of this result is that 23 is an ideal precursor
for an intramolecular 6-endo-trig Mannich-type cyclisation re-
action to derive 24. The question is then whether this is indeed the
case, or whether the reaction is more complex than it appears. To
this end, we next investigated the formal [4þ2]-cycloaddition re-
action of 5 and 22 by screening a series of oxy-philic Lewis acid and
base combinations (Eq. 6), with the aim of optimising the formation
of the cyclic adduct 24. The results are shown in Table 5.
In addition to these results, it should be noted that Jørgensen
et al. developed an efficient asymmetric organocatalytic protocol,
which is analogous to this work and involves an iminium ion-
aldehyde cyclisation.22 Although
L-proline provided cyclisation,
Table 5
Optimising the reaction between imine 5 and enone 22 to give the formal aza-
DielseAlder adduct 24, as shown in Eq. 6
there was no asymmetric induction, and indeed, a C2-symmetric
chiral pyrrolidine was required in order to effect asymmetric
induction.
Entry
Lewis acid (%)
Base
Solvent
23 (%)
24 (%)
25 (%)
Early studies on the synthesis of 24 demonstrated that this could
be achieved by reacting methyl vinyl ketone 22 and imine 5 in the
presence of HCl (reaction heated to reflux), to give the cyclic
product 24.17,18 In addition, chiral Brønsted acids have, in recent
years, proven to be efficient catalysts for a series of asymmetric
transformations, including DielseAlder reactions.23 We therefore
wondered whether, through the addition of chiral Brønsted acids,
the formal [4þ2]-cycloaddition reaction between 5 and 22 could be
achieved. In this context, chiral Brønsted acid catalysis was exam-
ined by using the chiral phosphoric acid (S)-2623 in the formal
[4þ2] cycloaddition (see Table 6, entry 11). Acid (S)-26 catalysed
1
In(OTf)3 (20)
In(OTf)3 (20)
In(OTf)3 (20)
In(OTf)3 (20)
In(OTf)3 (40)
In(OTf)3 (40)
In(OTf)3 (40)
In(OTf)3 (40)
In(OTf)3 (40)
In(OTf)3 (40)
Sc(OTf)3 (10)a
Fe(OTf)2 (20)
Ga(Otf)3 (20)
NaOH
STABc
NaOAc
NaOAc
NaOAc
NaOH
NaOH
NaOH
NaOH
STABc
NaOH
NaOH
NaOH
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
EtOAc
MeOH
MeCN
THF
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
d
d
d
d
d
d
d
d
d
d
d
d
d
62
12
11
45
59
37
37
27
5
11
39b
35
18
d
d
d
d
d
d
d
d
d
34
d
d
d
2d
3e
4
5
6
7
8
9
10d
11
12
13
All reactions were stirred for 24 h at rt.
a
Reaction was conducted in the presence of chiral PyBOX (10%).
b
Enantiomeric excess was found to be 0% by chiral HPLC (see Supplementary
data).
Table 6
Brønsted acid catalysed cyclisation between 5 and 22, to 24, as shown in Eq. 8
c
NaBH(OAc)3.
Reaction was quenched with NaOH.
Reaction was quenched with brine.
Conv. 24 (%)a
d
e
Entry
Catalyst (%)
1
2
-
0
57
(R)-Camphorsulfonic acid (20)
In(OTf)3 (20 mol %), in combination with the addition of NaOH,
gave the formal aza-DielseAlder adduct 24 in 62% yield (Table 5,
entry 1). Further investigations varying the Lewis acid loading, base
additive and solvent (Table 5, entries 1e9) did not improve the
reaction to any significant extent. Even with increased catalyst
loading, formation of 24 could not be increased past 62%. With this
in mind, we turned our attention to alternative Lewis acids. It was
found, however that the most efficient catalyst for forming the
adduct 24 was still In(OTf)3. Interestingly, when Sc(OTf)3 was
employed as the catalyst (Table 5, entry 11) and PyBox (10%) was
used as an additive, with the aim of developing the asymmetric
formal [4þ2]-cycloaddition, the cyclic adduct 24 was obtained in
39% yield, however, with no asymmetric induction.
3
4
5
6
(R)-Camphorsulfonic acid (40)
Benzoic acid (20)
TsOH (20)
60
79
50
0
(R)-BINOL (20)
7
8
9
p-Nitrobenzoic acid (20)
Chloroacetic acid (20)
4-Phenylbutyric acid (20)
(S)-Mandelic acid (20)
64
86 [83]b
57
10
70
11c
Chiral phosphporic acid (20)
[44]b
2.5. Alternative approaches to asymmetric induction in the
formal [4D2]-cycloaddition reaction
aDetermined by 1H NMR spectroscopic analysis. bIsolated yield. cTHF
used as solvent. The reaction was heated to reflux for 16 h.
With the results from Table 5 in hand, attention was turned to
trying to solve the issue of obtaining asymmetric induction in the
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006

P.R. Girling et al. / Tetrahedron xxx (2016) 1e9
5
the transformation, giving the cyclic adduct 19 in 44% isolated
yield; however, chiral HPLC analysis revealed that the product was
racemic.
2.6. Understanding the mechanism of the formal [2D2D2],
[1D2D1D2] and [4D2]-cycloaddition reactions
In addition to (S)-26, several other chiral and achiral Brønsted
acids were examined under these conditions (see Table 6) to see
what types of systems could affect the formal cycloaddition re-
action between 5 and 22.
2.6.1. Proposed mechanism for the formal [2þ2þ2]-cycloaddition. It
is interesting to speculate upon the mode of activation of enone 7
by the Lewis acid, which triggered trimerisation. This might occur
as outlined in Scheme 3, i.e., through dimerisation, followed by
elimination to give electron deficient diene-dione 28, which can
then derive cycloadduct 8 (Pathway A) through a Lewis-acid cata-
lysed DielseAlder reaction. Alternatively, triacetylbenzene 9 can be
explained from the common intermediate 28 undergoing a third
reaction with 7, which leads to the trimer 9 (Pathway A) (see
Scheme 4).
No asymmetric induction was achieved in any cases. It should be
noted that a relationship between pKa and conversion was ob-
served, whereby the most efficient acid catalyst was found to be the
achiral chloroacetic acid, leading to the formation of (rac)-24 in 83%
isolated yield.
Next, we examined the potential of thiourea catalysts [e.g., (S)-
27] to catalyse the formal [4þ2]-cycloaddition reaction. It was an-
ticipated that the thiourea (S)-27 would hydrogen-bond to the
carbonyl of methyl vinyl ketone 22, thus further activating the Cb to
nucleophilic attack. Indeed, subsequent Michael-addition (of 5)
would, presumably, yield the analogous O-hydrogen bonded eno-
late; after suitable isomerisation to the external enolate. The
resulting enolate would then be ideally placed for the required
asymmetric 6-endo-trig cyclisation. The results of these studies are
shown in Eq. 9 and Table 7.
Scheme 3. Proposed mechanism for the formation of the dihydropyridine 8 or the
trimer 9.
Table 7
Thiourea-catalysed reaction between imine 5 and enone 22 to give the formal aza-
DielseAlder adduct 24, as shown in Eq. 8
Entry Temp ꢀC (S)-27 (%) Additive (%)
Solvent Conv. 24 (%)a
1
2
3
rt.
rt.
60
20
10
10
d
CH2Cl2
Chloroacetic acid (3) CH2Cl2
Toluene
0
60
0
d
a
Measured by 1H NMR analysis. Product 24 was found to be racemic by chiral
HPLC analysis.
The room-temperature reaction between 5 and 22, in the
presence of (S)-27, failed to give the desired product 24 (Table 7,
entry 1). The addition of catalytic amounts (3 mol %) of chloroacetic
acid (Table 7, entry 2), in parallel with (S)-27, afforded the cyclic
adduct (rac)-24 in 60% conversion (by 1H NMR analysis). Heating
the reaction to 60 ꢀC for 24 h had no influence and, indeed, the
formation of 24 was not achieved.
Scheme 4. Rationalisation of the observed formal [1þ2þ1þ2]-cycloaddition reaction.
The asymmetric formal [4þ2]-cycloaddition reaction between 5
and 22 was proving to be elusive. Regardless, the formation of 24 by
general Lewis acid catalysis or, indeed, Brønsted acid catalysis, did
efficiently produce access to formal cycloadduct 24 with varying
degrees of efficiency. In order to try and find a solution to the
asymmetric induction issue, the development of a more detailed
mechanistic understanding of the reaction was necessary. Along
this line, a study into the mechanism of the formal [4þ2]-
cycloaddition was undertaken through in situ spectroscopic tech-
niques (this will be discussed below).
2.6.2. Proposed mechanism for the formal [1þ2þ1þ2]-cyclo-
addition. It is possible to speculate that the formal [1þ2þ1þ2]-
cycloaddition reaction can be rationalised by assuming initial
Michael-addition of imine 16 to the activated enone 7. Methoxide
elimination and hemi-aminal formation would result in species 29.
Addition of a further enone 7 to 29, elimination of methoxide and
trapping the activated benzaldehyde equivalent 30 via diene-amine
31, would lead to imminium-diene 32, following further methoxide
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006

6
P.R. Girling et al. / Tetrahedron xxx (2016) 1e9
elimination. A [3,3]-sigmatropic shift cyclisation of 32, followed by
iminium-enamine conversion, would provide the formal
[1þ2þ1þ2]-cycloaddition product 19.
22 at 1682 cmꢁ1. This is matched with the tandem appearance of
a new C]O stretch at 1722 cmꢁ1, which is additionally supported
by characterisation data of the pure iminium compound 23 (see
Fig. 2).
2.6.3. Proposed mechanism for the formal [4þ2]-cycloaddition. We
have recently studied the direct versus conjugate addition reaction
of primary amines to enones and enals using in situ IR spectroscopy
(ReactIRÔ).24 In situ IR spectroscopy was particularly useful here
because it allowed the real-time monitoring of the reactions, pro-
viding instant feedback with regards to reaction progression, and
was especially useful when reactions were conducted under air-
and moisture-sensitive conditions.25 It was, therefore, ideal for the
investigation of the formal [4þ2]-cycloaddition reaction as out-
lined in Eq. 9, in order to gain mechanistic insight into this reaction.
Hence, the investigation was started using in situ IR spectroscopy
and by adding three equal portions of In(OTf)3 (/35 mol %, i.e.,
stoichiometric overall, with respect to triflate) to a stirred CH2Cl2
solution of imine 5 and methyl vinyl ketone 22 (1.1 equiv), and
following the reaction over time. The results are shown in Fig. 1.
Figs. 1 and 2 clearly show that upon the addition of the oxy-philic
In(OTf)3, Lewis acid activation of methyl vinyl ketone 22 occurs,
though coordination to the carbonyl. This results in a rapid
Michael-addition by the nitrogen of imine 5, as judged by the loss of
the C]N stretch (of 5) at 1630 cmꢁ1 in parallel with the C]O of
The real-time reaction monitoring suggests that the presumed
Michael-Mannich pathway predominates in the formal [4þ2] cy-
cloaddition (see Scheme 5), and presumably proceeds through an
indium-catalysed mechanism and enolate species 33. Moreover,
the formation of 23 appears to be facile and, indeed, the formation
of 24, via the Mannich-step (Scheme 5), appears to be slow and
could well be rate-determining (as shown by in situ IR spectro-
scopic analysis).
Scheme 5. Proposed mechanism for the formation of the dihydropyridinone 24.
The in situ IR study was also compared with 1H NMR analysis,
and in particularly the Mannich-step of the reaction (see
Supplementary data for NMR spectra) using preformed iminium
triflate 23 (formed under the same reaction conditions as in the in
situ IR spectroscopic study, except that the solvent was switched to
CDCl3 to allow 1H NMR monitoring). Once 23 had formed (ratio of
23:24 was 1.0:0.7, respectively), several drops of NaOD (in D2O)
were added to the NMR tube and the sample vigorously shaken.
The resulting 1H NMR spectrum (acquired within 15 min), showed,
to our surprise, the instantaneous loss of the iminium ion 23, but
this was not mirrored by the formation of the cycloadduct 24. In-
stead, the reaction proceeded to give a mixture of imine 5, enone 22
and cycloadduct 24. The regeneration of the starting imine 5 pre-
sumably occurs through a facile b-elimination.
The lack of success in the attempted asymmetric cyclisation of
the iminium ion 23 led us to examine the potential reactivity of the
analogous iodide towards 23, i.e., where the triflate was exchanged
for iodide using InI3 instead of In(OTf)3 (vide supra). Initially, the
iminium iodide was generated in situ by reaction of enone 22 with
imine 5 (evidence of iminium ion formation from 1H NMR analysis).
Subsequent attempts at direct cyclisation were consistent with
those of the analogous triflate 23.
Fig. 1. ReactIRdReaction progression: I addition of imine 5 and methyl vinyl ketone
22; II / IV addition of three equal portions of In(OTf)3, which shows the loss of 5 and
22 and the rise of the conjugate addition product 23.
The lack of a successful asymmetric entry to cycloadduct 24
through the in situ formation of 23 led us to check that iminium ion
23 was indeed being formed. Compounds of type 23 are rather
unusual. For example, 23, despite being an iminium ion salt, was
obtained after column chromatography. This may cast doubt on its
structural assignment, however, all characterisation data suggests
this to be the case, as opposed to the corresponding triflate adduct,
i.e.,
the literature,26 the fact that the iminium ion CH was observed at
low field in the 1H NMR spectrum (
9.12 for 23), compared to 7.95
for an -amido triflate seems to agree with our structural assign-
a-amino triflate. By comparison with the few reports found in
d
d
a
ment being the corresponding iminium ion. Hence, in order to
further corroborate these observations, we prepared the iodide 34
(see Scheme 6), with the hope of producing the crystalline material
(suitable for crystallographic studies) to confirm, beyond doubt, the
structure of 23 (through comparison with 34). However, the anal-
ogous iodide 34 was obtained, after chromatography, as an oil.
Fig. 2. ReactIR graphical outputdshift of C]O stretch at 1682 cmꢁ1 to C]O stretch at
1722 cmꢁ1 is consistent with conjugate addition of imine 5 to enone 22.
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006

P.R. Girling et al. / Tetrahedron xxx (2016) 1e9
7
4.2. 6,7-Dihydro-1H-pyrido[2,1-a]isoquinolin-2(11bH)-one (6)
ꢀ
To 5 (0.131 g, 1.0 mmol) and 3 A molecular sieves (1 g) under
argon was added Danishefsky’s diene 1 (0.240 mL, 1.2 mmol)
dropwise. The reaction mixture was stirred at rt for 48 h, diluted
with CH2Cl2 (2 5 mL), filtered and concentrated in vacuo. Purifi-
cation by silica gel chromatography (3:1, EtOAc:hexane, to 100%,
EtOAc, as eluent) afforded 6 as a pale orange solid (0.093 g, 47%): Rf.
0.05 (2:1, EtOAc:hexane as eluent); 1H NMR (400 MHz, CDCl3)
Scheme 6. Formation of iodide 34 in situ; isolated after column chromatography in
14% yield.
Regardless, the iodide 34 had consistent characterisation data with
that of the triflate 23 (similar structures exist in the literature23
)
d
7.29e7.17 (m, 4H), 7.16 (dd, J¼7.4, 0.8 Hz), 5.08 (dd, J¼7.4, 1.0 Hz,
with the corresponding iminium CH being observed at a similar low
field 1H NMR chemical shift of
9.81.
The facile -elimination of imine 5 from either adducts 23 or 34
1H), 4.76 (dd, J¼16.3, 4.5 Hz, 1H), 3.64 (ddd, J¼12.2, 5.1, 2.3 Hz, 1H),
3.45 (td, J¼12.2, 3.2 Hz, 1H), 3.16 (ddd, J¼15.8, 5.1, 0.9 Hz, 1H),
2.87e2.85 (m, 1H), 2.84e2.81 (m, 1H), 2.53 (t, J¼16.3 Hz, 1H) ppm;
d
b
was also found to occur under other conditions. For example, when
13C NMR (100 MHz, CDCl3)
d 192.8, 154.2, 135.0, 133.5, 129.5, 127.3,
adduct 23 was formed in situ, addition of pyrrolidine was found to
127.2, 125.7, 98.7, 56.7, 49.8, 44.1, 30.4; IR (thin film) 1630 (s), 1586,
1581 cmꢁ1; LRMS (TOF ESþ), 222.2 (100%) [MþNa]þ, 200.2 (40%)
[MþH]þ; HRMS (TOF ESþ), calculated for C13H13NOþHþ, 200.1075;
found 200.1079. All spectroscopic and analytical properties were
identical to those reported in the literature.17
catalyse
b-elimination, leading to the formation of imine 5 (see
Supplementary data and Eq. 10).
4.3. 1,10-(7,11b-Dihydro-6H-pyrido[2,1-a]isoquinoline-1,3-
diyl)diethanone (8)
This can be rationalised by initial iminium ion formation,
through condensation, with subsequent tautomerisation to the
internal enamine, leading to facile elimination. The same generic
outcome was found when using the respective Jørgensen and
Hayashi catalysts,21 which, in addition to the observed conjugate
addition of amines to methyl vinyl ketone 22,24 explains the lack of
ee and poor catalytic performance (see above).
Method A. To 5 (0.131 g, 1.0 mmol) and Fe(OTf)2 (0.05 g,
0.1 mmol) in CH2Cl2 (1 mL) under argon was added 4-methoxy-3-
buten-2-one 7 (0.204 mL, 2 mmol). The reaction mixture was
stirred at rt for 24 h. Purification by silica gel chromatography (4:1,
EtOAc:CH2Cl2, as eluent) afforded 8 as a yellow solid (0.234 g, 88%):
mp 225e230 ꢀC; Rf. 0.1 (2:1, EtOAc:hexane as eluent); 1H NMR
(700 MHz, CDCl3)
d
7.83 (br s, 1H), 7.58 (br s, 1H), 7.18 (t, J¼7.2 Hz,
1H), 7.13 (t, J¼7.2 Hz, 1H), 7.10 (d, J¼0.4 Hz, 1H), 6.70 (d, J¼0.4 Hz,
1H), 5.91 (br s, 1H), 3.94 (dt, 13.1, 7.7 Hz, 1H), 3.79 (ddd, 13.1, 8.1,
4.3 Hz, 1H), 3.19 (dt, J¼16.4, 8.1 Hz, 1H), 3.09 (ddd, J¼16.4, 7.7,
4.3 Hz, 1H), 2.52 (br s, 3H), 2.17 (br s, 3H) ppm; 13C NMR (176 MHz,
3. Conclusions
In summary, a new class of reaction has been found when
reacting a conformationally locked cyclic imine 5 with 4-methoxy-
3-buten-2-one 7, in order to access diacetyl dihydropyridine 8 in
the presence of a Lewis acid. This is thought to go through a formal
[2þ2þ2]-cycloaddition reaction pathway. Using higher Lewis acid
loadings, however, trimer 9 predominantly forms. The use of acyclic
imines failed to provide these types of formal [2þ2þ2]-
cycloadducts, rather formal [1þ2þ1þ2]-cycloaddition products
were obtained. Using Danishefsky’s diene 1 instead of the unac-
tivated enone 7 afforded the aza-DielseAlder adduct 6, i.e., through
a formal [4þ2]-cycloaddition that was most likely a Mannich-
Michael process. In addition, the Lewis acid catalysed Michael-
Mannich reaction was examined, leading to the development of
a formal [4þ2]-cycloaddition pathway towards the synthesis of the
aza-DielseAlder adduct 24. In situ spectroscopic investigations
were conducted, which shed light on this process and, indeed,
support the Michael-Mannich pathway. Further applications of
these one-pot assembly processes to derive novel heterocycles are
under examination.
CDCl3)
d 196.8, 191.0, 151.0, 138.3, 135.4, 132.6, 128.8, 127.5, 126.7,
124.7, 109.0, 55.0, 51.7, 31.0, 28.4, 25.0, 24.7; IR (neat) 1644 (s), 1591,
1538 cmꢁ1; UV (MeOH nm) 409 (S 5841), 315 (S 20,076), 228 (S
11,361), 212 (S 13,139); LRMS (TOF ESþ), 290.3 (100%) [MþNa]þ,
268.3 [MþH]þ; HRMS (TOF ESþ), calculated for C17H17NO2þHþ,
268.1338; found 268.1335. Anal. calcd: C, 76.38, H, 6.41, N, 5.24,
found: C, 75.85, H, 6.38, N, 5.13.
Method B. Compound 8 was also isolated and purified by tritu-
ration as follows: To 5 (0.131 g, 1.0 mmol) and Sc(OTf)3 (0.049 g,
0.1 mmol) in CH2Cl2 (1.5 mL) under argon was added 4-methoxy-3-
buten-2-one 7 (0.408 mL, 4 mmol). The reaction mixture was stir-
red at rt for 48 h, filtered and washed with Et2O, followed by EtOAc
drop by drop whilst filtering, in order to give 8 as a yellow solid
(0.133 g, 50%). All spectroscopic and analytical properties were
identical to those reported in Method A.
4.4. 1,3,5-Triacetylbenzene (9)
4. Experimental section
To 4-methoxy-3-buten-2-one 7 (0.102 mL, 1.0 mmol) in CHCl3
(0.5 mL) was added Yb(OTf)3 (0.124 g, 0.2 mmol) and the reaction
mixture was stirred at rt for 7 days. Purification by silica gel chro-
matography (3:2 diethyl ether:hexane as eluent) afforded 9 as
a white solid (0.030 g, 45%): mp 158e159 ꢀC (lit. 158e160 ꢀC);12b Rf.
0.16 (2:1, diethyl ether:hexane as eluent); 1H NMR (400 MHz,
4.1. 3,4-Dihydroisoquinoline (5)
Prepared following standard literature procedure:11 1H NMR
(700 MHz, CDCl3)
d
8.3 (br s, 1H), 7.31 (t, J¼7.3 Hz, 1H), 7.26 (t,
J¼7.3 Hz, 1H), 7.22 (d, J¼7.3 Hz, 1H), 7.11 (d, J¼7.3 Hz, 1H), 3.73 (t,
CDCl3)
d
8.70 (br s, 3ꢂ1H), 2.71 (br s, 3ꢂ3H) ppm; 13C NMR
J¼7.8 Hz, 2H), 2.71 (t, J¼7.8 Hz, 2H) ppm; 13C NMR (101 MHz, CDCl3)
(400 MHz, CDCl3)
d 196.7, 138.1, 131.9, 27.0; IR (thin film) 1687 (s),
d
160.4, 136.4, 131.1, 128.5, 127.4, 127.2, 127.1, 47.4, 25.0; IR nmax
1361, 1225 cmꢁ1; LRMS (FTMS NESþ), 222.1 (100%) [MþNH4]þ,
(neat) 1626 cmꢁ1; LRMS (TOF ESþ), 132.2 (100%) [MþH]þ; HRMS
(TOF ESþ), calculated for C9H9NþHþ, 132.08078; found 132.08092.
All spectroscopic and analytical properties were found to be iden-
tical to those reported in the literature.11
205.1 (16%), [MþHþ]; HRMS (FTMS ESþ), calculated for
C
12H12O3þNHþ4 , 222.1125; found 222.1127; Anal. Calcd: C, 70.57, H,
5.92, found: C, 69.19, H, 5.89. All spectroscopic and analytical
properties were identical to those reported in the literature.12b
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006

8
P.R. Girling et al. / Tetrahedron xxx (2016) 1e9
4.5. 1,10-(1-Allyl-4-phenyl-1,4-dihydropyridine-3,5-diyl)dieth-
anone (19)
4.8. 2-(3-Oxobutyl)-3,4-dihydroisoquinolin-2-ium tri-
fluoromethanesulfonate (23)
To 16 (0.145 g, 1.0 mmol) and Sc(OTf)3 (0.049 g, 0.1 mmol) in
CHCl3 (1.5 mL) under argon was added 4-methoxy-3-buten-2-one 7
(0.408 mL, 4.0 mmol). The reaction mixture was stirred at rt for 2
days, washed with satd (aq) NaHCO3 (5 mL) and extracted with
CH2Cl2 (3ꢂ7 mL). The combined organic extracts were dried
(MgSO4) and concentrated in vacuo. Purification by silica gel
chromatography (1:9, EtOAc:diethyl ether, as eluent) afforded 19 as
a yellow oil (0.089 g, 31%): mp 118e119 ꢀC; Rf. 0.18 (1:9, EtOAc:-
To 5 (0.131 g, 1.0 mmol) and Yb(OTf)3 (0.124 g, 0.20 mmol) in
CH2Cl2 (1.5 mL) under argon was added methyl vinyl ketone 22
(0.162 mL, 2 mmol). The reaction mixture was stirred at rt overnight
and concentrated in vacuo. Purification by silica gel chromatogra-
phy (1:1, EtOAc:hexane, to 100%, EtOAc, as eluent) afforded 24 as
a white solid (0.051 g, 25%) and 23 as a yellow oil (0.128 g, 37%): Rf.
0.23 (1:1, EtOAc:methanol, as eluent); 1H NMR (700 MHz, CDCl3)
d
9.12 (br s, 1H), 7.87 (dd, J¼7.6, 1.1 Hz, 1H), 7.72 (td, J¼7.6, 1.1 Hz,
diethyl ether, as eluent); 1H NMR (700 MHz, CDCl3)
d
7.30 (d,
1H), 7.47 (td, J¼7.6, 1.1 Hz, 1H), 7.35 (dd, J¼7.6, 1.1 Hz, 1H), 4.32 (t,
J¼6.0 Hz, 2H), 4.14 (t, J¼8.1 Hz, 2H), 3.29 (t, J¼6.0 Hz, 2H), 3.27 (t,
J¼7.7 Hz, 2H), 7.22 (t, J¼7.7 Hz, 2H), 7.14e7.11 (m, 1H), 7.13 (s, 2H),
5.94 (ddt, J¼17.0, 10.5, 5.6 Hz, 1H), 5.38 (dtd, J¼10.5, 1.6, 0.9 Hz, 1H),
5.36 (dtd, J¼17.0, 1.6, 0.9 Hz, 1H), 5.18 (s, 1H), 4.10 (dt, J¼5.6, 1.6 Hz,
J¼8.1 Hz, 2H), 2.23 (s, 3H) ppm; 13C NMR (176 MHz, CDCl3)
d 206.2,
168.1, 138.6, 136.4, 134.7, 128.7 (q, J¼285 Hz), 124.7, 121.9, 119.4,
116.8, 55.6, 49.2, 40.1, 30.1, 25.5; IR nmax (thin film) 1714 (C]O),
1661 (C]N) cmꢁ1; LRMS (TOF ESþ), 203.5 (100%) [MþH]þ, 201.7
(70%), 132.1 (25%); LRMS (TOF ES-), 149.0 (100%) [OTf]; HRMS
2H), 2.15 (br s, 6H) ppm; 13C NMR (176 MHz, CDCl3)
d 195.2, 145.9,
138.0, 132.5, 128.3, 128.3, 126.6, 119.6, 119.5, 57.3, 35.9, 25.7; IR (thin
film) 1633 (C]O), 1566 cmꢁ1; LRMS (TOF ESþ), 304.3 (100%)
[MþNa]þ, 282.3 (35%) [MþH]þ, 176.2 (20%), 146.2 (20%); HRMS
(FTMS ESþ) calculated for
C
13H15NOþHþ, 202.12264; found
(TOF ESþ), calculated for
C
18H19NO2þHþ, 282.1494; found
202.12262; HRMS (FTMS ES-), calculated for CF3O3Se, 148.95257;
found 148.95217.
282.1495; Anal. Calcd: C, 76.84, H, 6.81, N, 4.98, found: C, 76.65, H,
6.84, N, 4.94. All spectroscopic and analytical data were identical to
those reported in the literature.14
4.9. 2-(3-Oxobutyl)-3,4-dihydroisoquinolin-2-ium iodide (34)
To 5 (0.178 g, 1.35 mmol) and InI3 (0.234 g, 0.47 mmol) in CH2Cl2
(4 mL) under argon was added methyl vinyl ketone 22 (0.125 mL,
1.5 mmol). The reaction mixture was stirred for 4 h and concen-
trated in vacuo. Purification by silica gel chromatography (1:1,
EtOAc:hexane, to 100%, EtOAc, as eluent) gave 34 as a pure yellow
4.6. 1,10-(1-Allyl-4-(dimethoxymethyl)-1,4-dihydropyridine-
3,5-diyl)diethanone (20)
To 17 (0.143 g, 1.0 mmol) and Sc(OTf)3 (0.049 g, 0.1 mmol) in
CHCl3 (1.5 mL) under argon was added 4-methoxy-3-buten-2-one 7
(0.408 mL, 4.0 mmol). The reaction mixture was stirred at rt for 2
days, washed with satd (aq) NaHCO3 (5 mL) and extracted with
CH2Cl2 (3ꢂ7 mL). The combined organics were dried (MgSO4) and
concentrated in vacuo. Purification by silica gel chromatography
(1:9, EtOAc:diethyl ether, to EtOAc, as eluent) afforded 20 as a dark
orange oil (0.045 g, 20%): Rf. 0.15 (EtOAc, as eluent); 1H NMR
oil (62 mg, 14%): 1H NMR (400 MHz, CDCl3)
d 9.81 (s, 1H), 8.04 (d,
J¼7.4 Hz, 1H), 7.70 (t, J¼7.6 Hz, 1H), 7.45 (t, J¼7.5 Hz, 1H), 7.35 (d,
J¼7.6 Hz, 1H), 4.43 (t, J¼5.8 Hz, 2H), 4.22 (t, J¼8.0 Hz, 2H), 3.50 (t,
J¼5.7 Hz, 2H), 3.33 (t, J¼7.9 Hz, 2H), 2.27 (s, 3H) ppm; 13C NMR
(176 MHz, CDCl3)
d 205.9, 166.9, 138.2, 136.2, 134.7, 128.7, 128.3,
124.5, 55.7, 50.2, 41.0, 30.7, 25.6; IR nmax (thin film) 3001, 2943, 1706
(C]O), 1660 (C]N), 1575, 1358, 1172 cmꢁ1; LRMS (TOF ESþ), 202.6
(100%) [MþH]þ. HRMS (FTMS ESþ) calculated for C13H15NOþHþ,
202.1232; found 202.1218.
(400 MHz, CDCl3)
d
7.09 (s, 2H), 5.86 (ddt, J¼16.7, 10.2, 5.0 Hz, 1H),
5.33 (dtd, J¼16.7, 1.6, 0.9 Hz, 1H), 5.29 (dtd, J¼10.2, 1.6, 0.9 Hz, 1H),
4.45 (d, J¼4.0 Hz, 1H), 4.03 (dt, J¼5.0, 1.6 Hz, 2H), 3.97 (d, J¼4.0 Hz,
1H), 3.22 (br s, 6H), 2.26 (br s, 6H) ppm; 13C NMR (101 MHz, CDCl3)
4.10. 3,4,6,7-Tetrahydro-1H-pyrido[2,1-a]isoquinolin-2(11bH)-
d
195.9, 139.9, 132.7, 118.8, 114.9, 107.3, 57.3, 55.9, 33.3, 25.5; IR (thin
one (24)
film) 1639 (C]O), 1567 cmꢁ1; LRMS (TOF ESþ), 302.3 (100%)
[MþNa]þ, 280.3 (60%) [MþH]þ, 176.2 (20%), 248.2 (30%); HRMS
Method A. To 5 (0.131 g, 1.0 mmol) and Yb(OTf)3 (0.124 g,
0.20 mmol) in CH2Cl2 (1.5 mL) under argon was added methyl vinyl
ketone 22 (0.162 mL, 2 mmol). The reaction mixture was stirred at
rt overnight and concentrated in vacuo. Purification by silica gel
chromatography (1:1, EtOAc:hexane, to 100% EtOAc, as eluent)
afforded 23 as a yellow oil (0.128 g, 37%) and 24 as a white solid
(0.051 g, 25%).
(TOF ESþ), calculated for
C
15H21NO4þNaþ, 302.1368; found
302.1382. All spectroscopic and analytical data were identical to
those reported in the literature.14
4.7. p-Methoxyphenylamino-4-butene-3-one-2 (21)
Method B. Compound 5 (0.39 g, 3 mmol) and methyl vinyl ke-
To 16 (0.211 g, 1.0 mmol) in CHCl3 (1 mL) was added 4-methoxy-
3-buten-2-one (0.255 mL, 2.5 mmol) and Sc(OTf)3 (0.049 g,
0.1 mmol). The reaction mixture was flushed with argon and stirred
at rt for 2 days. The crude was concentrated in vacuo and purified
by silica gel chromatography (1:9, EtOAc:diethyl ether, to EtOAc, as
eluent) to afford 21 as an orange oil (0.048 g, 25%): Rf. 0.53 (EtOAc,
tone 22 (275 mL, 3.3 mmol) were added to a stirring solution of
tBME (12 mL) under argon. Chloroacetic acid (56.7 mg, 0.6 mmol)
was added to the solution and the mixture was stirred overnight
(16 h). The resulting solution was partitioned between EtOAc and
washed with K2CO3 (1ꢂ20 mL), brine (2ꢂ20 mL), after, which the
organic layer was dried over MgSO4. After filtration, the organic
layer was concentrated to yield a dark orange oil. Purification by
silica gel chromatography (1:1, EtOAc:hexane, to 100% EtOAc, as
eluent) afforded 22 as a white solid (0.501 g, 83%): mp 75e77 ꢀC (lit.
76e77 ꢀC);18 Rf. 0.18 (EtOAc, as eluent); 1H NMR (600 MHz, CDCl3)
as eluent); 1H NMR (400 MHz, CDCl3)
d
11.63 (br d, J¼12.0 Hz, 1H),
7.15 (dd, J¼12.0, 7.6 Hz, 1H), 7.01e6.97 (m, 2H), 6.90e6.86 (m, 2H),
5.26, (d, J¼7.6 Hz,1H), 3.80 (s, 3H), 2.15 (s, 3H) ppm (addition of D2O
caused the signal at
d
11.63 to disappear, and the signal at
d
7.15 to
change to a d, J¼7.6 Hz); 13C NMR (101 MHz, CDCl3)
d
198.5, 156.3,
d
7.20e7.13 (m, 3H), 7.10e7.06 (m, 1H), 3.59 (dd, J¼11.9, 3.0 Hz, 1H),
144.2, 134.2, 117.8, 115.1, 96.7, 55.7, 29.5; IR (thin film) 1636 (C]O),
1597, 1569, 1513, 1479 cmꢁ1; LRMS (TOF ES-), 190.2 (100%) [MꢁH]e,
3.28 (ddd, J¼10.8, 5.2, 3.0 Hz, 1H), 3.22e3.12 (m, 2H), 2.96 (ddd,
J¼14.6, 3.4, 2.4 Hz, 1H), 2.84e2.80 (m, 1H), 2.75e2.67 (m, 2H), 2.63
(td, J¼10.6, 3.9 Hz, 1H), 2.50 (ddd, J¼14.6, 12.0, 0.9 Hz, 1H), 2.43
(ddd, J¼12.0, 3.4, 1.6 Hz, 1H) ppm; 13C NMR (151 MHz, CDCl3)
175.1 (25%); HRMS (TOF ES-), calculated for
C
11H13NO2eHþ,
190.0868; found 190.0871. All spectroscopic and analytical prop-
erties were identical to those reported in the literature.16
d
208.7, 136.8, 134.1, 129.1, 126.7, 126.3, 124.9, 61.9, 54.9, 50.8, 47.4,
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006

P.R. Girling et al. / Tetrahedron xxx (2016) 1e9
9
41.2, 29.9; IR nmax (thin film) 1714 (C]O), 1360 cmꢁ1; LRMS (TOF
ESþ), 202.2 (100%) [MþH]þ; HRMS (FTMS ESþ), calculated for
8. (a) Di Bari, L.; Guillarme, S.; Hermitage, S.; Jay, D. A.; Pescitelli, G.; Whiting, A.
Chirality 2005, 17, 323e331; (b) Bundu, A.; Guillarme, S.; Hannan, J.; Wan, H.;
Whiting, A. Tetrahedron Lett. 2003, 44, 7849e7850; (c) McFarlane, A. K.;
Thomas, G.; Whiting, A. Tetrahedron Lett. 1993, 34, 2379e2382; (d) McFarlane,
A. K.; Thomas, G.; Whiting, A. J. Chem. Soc., Perkin Trans. I 1995, 2803e2808.
9. (a) Hermitage, S.; Howard, J. A. K.; Jay, D.; Pritchard, R. G.; Probert, M. R.;
Whiting, A. Org. Biomol. Chem. 2004, 2451e2460; (b) Guillarme, S.; Hermitage,
S.; Howard, J. A. K.; Jay, D. A.; Whiting, A.; Yufit, D. S. Synlett 2004, 708e710; (c)
Guillarme, S.; Whiting, A. Synlett 2004, 711e713; (d) Hermitage, S.; Jay, D.;
Whiting, A. Tetrahedron Lett. 2002, 43, 9633e9636.
C
13H15NOþHþ, 202.12264; found 202.12264. The ee was de-
termined by chiral HPLC using OJ-H-CHIRALCELÒ column
(250ꢂ4.6 mm), 35 ꢀC, 1 mL/min, 215 nm, hexane:IPA (9:1),
tR1¼10.4 min; tR2¼16.7 min. All spectroscopic and analytical
properties were identical to those reported in the literature.27
10. (a) Zheng, P. G.; Lieberman, B. P.; Choi, S. R.; Ploessl, K.; Kung, H. F. Bioorg. Med.
Chem. Lett. 2011, 21, 3435e3438; (b) Zorn, C.; Anichini, B.; Goti, A.; Brandi, A.;
Kozhushkov, S. I.; de Meijere, A.; Citti, L. J. Org. Chem. 1999, 64, 7846e7855.
11. Elliott, M. C.; Williams, E. Org. Biomol. Chem. 2003, 1, 3038e3047.
12. Murahashi, S. I.; Mitsue, Y.; Tsumiyama, T. Bull. Chem. Soc. Jpn. 1987, 60,
3285e3290; (b) Kumbaraci, V.; Ergunes, D.; Midilli, M.; Begen, S.; Talinli, N. J.
Heterocycl. Chem. 2009, 46, 226e230.
Acknowledgements
We thank Merck & Co. supporting this work, the EPSRC for a DTA
studentship (to PRG), and the EPSRC National Mass Spectrometry
Service, Swansea.
13. Iwado, T.; Hasegawa, K.; Sato, T.; Okada, M.; Sue, K.; Iwamura, H.; Hiaki, T. J. Org.
Chem. 2013, 78, 1949e1954.
14. Girling, P. R.; Batsanov, A. S.; Shen, H. C.; Whiting, A. Chem. Commun. 2012,
4893e4895.
Supplementary data
15. Di Bari, L.; Guillarme, S.; Hanan, J.; Henderson, A. P.; Howard, J. A. K.; Pescitelli,
G.; Probert, M. R.; Salvadori, P.; Whiting, A. Eur. J. Org. Chem. 2007, 5771e5779.
16. Duguay, G.; Metayer, C.; Quiniou, H. Bull. Soc. Chim. Fr. 1974, 2507e2512.
17. (a) Cicchi, S.; Revuelta, J.; Zanobini, A.; Betti, M.; Brandi, A. Synlett 2003,
2305e2308; (b) Revuelta, J.; Cicchi, S.; Brandi, A. J. Org. Chem. 2005, 70,
5636e5642.
Supplementary data associated with this article can be found in
the online version, at http://dx.doi.org/10.1016/j.tet.2016.01.006.
References and notes
18. Caroon, J. M.; Clark, R. D.; Kluge, A. F.; Lee, C.-H.; Strosberg, A. M. J. Med. Chem.
1983, 26, 1426e1433.
19. For reviews on asymmetric organocatalysis: (a) Bertelsen, S.; Jørgensen, K. A.
Chem. Soc. Rev. 2009, 38, 2178e2189; (b) Taylor, M. S.; Jacobsen, E. N. Angew.
Chem., Int. Ed. 2006, 45, 1520e1543.
1. Bailey, P. D.; Millwood, P. A.; Smith, P. D. Chem. Commun. 1998, 633e640.
2. Asano, N.; Nash, R. J.; Molyneux, R. J.; Fleet, G. W. J. Tetrahedron: Asymmetry
2000, 11, 1645e1680.
3. (a) Huang, X. G.; Zhang, A. Q.; Chen, D. L.; Jia, Z. H.; Li, X. S. Bioorg. Med. Chem.
Lett. 2010, 20, 2859e2863; (b) Maryanoff, B. E.; McComsey, D. F.; Taylor, R. J.;
Gardocki, J. F. J. Med. Chem. 1981, 24, 79e88.
20. Itoh, T.; Yokoya, M.; Miyauchi, K.; Nagata, K.; Ohsawa, A. Org. Lett. 2003, 5,
4301e4304.
ꢁ
4. (a) Jurcík, V.; Arai, K.; Salter, M. M.; Yamashita, Y.; Kobayashi, S. Adv. Synth. Catal.
ꢂ
21. For example: (a) Bertelsen, S.; Marigo, M.; Brandes, S.; Diner, P.; Jørgensen, K. A.
ꢂ
ꢂ
2008, 350, 647e651; (b) Sunden, H.; Ibrahem, I.; Eriksson, L.; Cordova, A. An-
gew. Chem. 2005, 44, 4877e4880; (c) Yao, S. L.; Saaby, S.; Hazell, R. G.;
Jørgensen, K. A. Chem.dEur. J. 2000, 6, 2435e2448; (d) Ishihara, K.; Miyata, M.;
Hattori, K.; Tada, T.; Yamamoto, H. J. Am. Chem. Soc. 1994, 116, 10520e10524; (e)
Hattori, K.; Yamamoto, H. J. Org. Chem. 1992, 57, 3264e3265.
J. Am. Chem. Soc. 2006, 128, 12973e12980; (b) Hayashi, Y.; Gotoh, H.; Masui, R.;
Ishikawa, H. Angew. Chem., Int. Ed. 2008, 47, 4012e4015.
22. Frisch, K.; Landa, A.; Saaby, S.; Jørgensen, K. A. Angew. Chem., Int. Ed. 2005, 44,
6058e6063.
23. (a) Rueping, M.; Kuenkel, A.; Atodiresei, I. Chem. Soc. Rev. 2011, 40, 4539e4549;
(b) Nakashima, D.; Yamamoto, H. J. Am. Chem. Soc. 2006, 128, 9626e9627.
5. Danishefsky, S.; Kitahara, T. J. Am. Chem. Soc. 1974, 96, 7807e7808.
6. (a) Girling, P. R.; Kiyoi, T.; Whiting, A. Org. Biomol. Chem. 2011, 9, 3105e3121; (b)
Buonora, P.; Olsen, J.-C.; Oh, T. Tetrahedron 2001, 57, 6099e6138; (c) Heintzel-
man, G. R.; Meigh, I. R.; Mahajan, Y. R.; Weinreb, S. M. Org. React. 2005, 65,
141e599; (d) Rowland, G. B.; Rowland, E. B.; Zhang, Q.; Antilla, J. C. Curr. Org.
Chem. 2006, 10, 981e1005; (h) Kouznetsov, V. V. Tetrahedron 2009, 65,
2721e2750; (i) Masson, G.; Lalli, C.; Benohoud, M.; Dagousset, G. Chem. Soc. Rev.
2013, 42, 902e923.
ꢂ
ꢂ
24. Calow, A. D. J.; Carbo, J. J.; Cid, J.; Fernandez, E.; Whiting, A. J. Org. Chem. 2014,
11, 5163e5172.
25. Stead, D.; Carbone, G.; O’Brien, P.; Campos, K. R.; Coldham, I.; Sanderson, A. J.
Am. Chem. Soc. 2010, 132, 7260e7261.
ꢃ
26. De Meester, F.; Frere, J. M.; Piette, J. L.; Jacquemin, P.; Grooters, L.; Llabres, G.;
Defays, S. J. Heterocycl. Chem. 1992, 29, 535e541.
27. Cui, L.; Peng, Y.; Zhang, L. M. J. Am. Chem. Soc. 2009, 131, 8394e8395.
7. (a) Weintraub, P. M.; Sabol, J. S.; Kane, J. M.; Borcherding, D. R. Tetrahedron
2003, 59, 2953e2989; (b) Sahu, S. K.; Dubey, B. K.; Tripathi, A. C.; Koshy, M.;
Saraf, S. K. Mini Rev. Med. Chem. 2013, 13, 565e583.
Please cite this article in press as: Girling, P. R.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.01.006