Stereoselective synthesis of the rat selective toxicant norbormide

Article abstract of DOI:10.1016/j.tet.2016.07.014

Norbormide [5-(α-hydroxy-α-2-pyridylbenzyl)-7-(α-2-pyridylbenzylidene)-5-norbornene-2,3-dicarboximide] (NRB, 1), an existing but infrequently used rodenticide, is known to be uniquely toxic to rats but relatively harmless to other rodents/mammals. However, as an acute vasoactive, NRB has a rapid onset of action which makes it relatively unpalatable to rats, often leading to sub-lethal uptake/accompanying bait shyness. It is recognized that the pharmacological profile of NRB is highly sensitive to its stereoisomeric composition, yet no comprehensive endeavor has been made to impart any stereochemical control in its manufacture. The effect of temperature/concentration/reaction solvent/Lewis acid on the Diels–Alder synthesis of NRB is investigated. An attempted stereoselective synthesis of key NRB precursor 2-fulvenylmethanol 4 is also described. Palatability/efficacy data in rats is reported for novel NRB batches 1a–1e. Disappointingly, given the level of stereocontrol now available in the synthesis of NRB as a result of this research, no clear relationship between stereoisomeric composition and palatability was observed.

Full text of DOI:10.1016/j.tet.2016.07.014

Tetrahedron xxx (2016) 1e12
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Stereoselective synthesis of the rat selective toxicant norbormide
,
,
,e
Morgan Jay-Smith a, Elaine C. Murphy b c, Lee Shapiro b d, Charles T. Eason b
,
Margaret A. Brimble a , David Rennison
,
a,
*
*
a School of Chemical Sciences, University of Auckland, Auckland, New Zealand
b Centre for Wildlife Management and Conservation, Lincoln University, Lincoln, New Zealand
c Department of Conservation, Christchurch, New Zealand
d Connovation Ltd., Manukau, New Zealand
e Cawthron Institute, Nelson, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history:
Norbormide [5-(a-hydroxy-a-2-pyridylbenzyl)-7-(a-2-pyridylbenzylidene)-5-norbornene-2,3-dicarboximide]
Received 25 April 2016
Received in revised form 22 June 2016
Accepted 4 July 2016
(NRB, 1), an existing but infrequently used rodenticide, is known to be uniquely toxic to rats but rel-
atively harmless to other rodents/mammals. However, as an acute vasoactive, NRB has a rapid onset of
action which makes it relatively unpalatable to rats, often leading to sub-lethal uptake/accompanying
bait shyness. It is recognized that the pharmacological profile of NRB is highly sensitive to its stereo-
isomeric composition, yet no comprehensive endeavor has been made to impart any stereochemical
control in its manufacture. The effect of temperature/concentration/reaction solvent/Lewis acid on the
DielseAlder synthesis of NRB is investigated. An attempted stereoselective synthesis of key NRB pre-
cursor 2-fulvenylmethanol 4 is also described. Palatability/efficacy data in rats is reported for novel
NRB batches 1ae1e. Disappointingly, given the level of stereocontrol now available in the synthesis of
NRB as a result of this research, no clear relationship between stereoisomeric composition and pal-
atability was observed.
Available online xxx
Keywords:
Norbormide
Rodenticide
Rat toxicant
DielseAlder
Stereoselective
Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction
only the endo isomers of NRB were toxic in rats.5 The mechanisms
involved in these divergent effects of NRB have yet to be clarified.
Discovered in the 1960s and first introduced to the market 50
years ago, Norbormide(NRB 1) (Fig.1), avasoactivealsoknownunder
the trade names ShoxinÒ and RaticateÒ, was found to be uniquely
toxic to rats but relatively harmless to other rodents and mammals.1,2
NRB displays unique species-specific contractile activity that is re-
stricted to the peripheral arteries of the rat. In arteries from all other
species tested, as well as in rat aorta and extravascular smooth
muscle tissue, NRB exhibits vasorelaxant properties at concentra-
tions that induce vasoconstriction in rat peripheral arteries.3 Fur-
thermore, detailed studies conducted upon the individual
stereoisomers of NRB, isolated from the endo rich stereoisomeric
mixture, found the parent compound’s physiological effects to be
strongly stereospecific. In rat peripheral arteries only the endo iso-
mers of NRB retained the contractile activity elicited by the stereo-
isomeric mixture, with both the endo and exo isomers exhibiting
vasorelaxant activity in rat aorta.4 In vivo evaluation established that
Disappointingly, to date, efforts to establish NRB as a viable
rodenticide have been largely unsuccessful. Over time, rats as
a species have developed an evolutionary trait relating to how they
sample food, particularly novel food, such that they often do not
ingest a potentially toxic dose at the first encounter. This survival
strategy is most likely linked to their lack of an emetic centre, and
thus their incapacity, as a species, to vomit. As an acute poison, NRB
has a rapid onset of action in rats with toxic symptoms being
recorded almost immediately, and evidence suggests that rats de-
velop a learnt aversion to this poison following the consumption of
sub-lethal doses during sampling, a phenomenon referred to as
bait-shyness.6,7 NRB is also known to be relatively unpalatable to
rats.8e10 Although efforts to address this palatability problem using
microencapsulation technologies have had varying degrees of
success, the onset of symptoms has not been sufficiently delayed to
significantly improve palatability,11,12 and the potential for sub-
lethal dosing remains a major problem. More recently, a prodrug
approach13e15 revealed several NRB derivatives which were dem-
onstrated to be more palatable to rats.
* Corresponding authors. Tel.: þ64 9 373 7599x83881 (D.R.); tel.:þ64 9 373
7599x88259 (M.A.B.); fax: þ64 9 373 7422 (school); e-mail address: d.rennison@
auckland.ac.nz (D. Rennison).
A survey of literature NRB syntheses1,4,16 unequivocally de-
scribed a two-step procedure culminating in the DielseAlder
http://dx.doi.org/10.1016/j.tet.2016.07.014
0040-4020/Ó 2016 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

2
M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
Fig. 1. Naming and numbering of individual stereoisomers of NRB (1).4
reaction of 2-fulvenylmethanol 4 and commercially available mal-
eimide (6); in all cases, precursor 2-fulvenylmethanol 4 was pre-
pared through the base-promoted reaction of 2-benzoylpyridine
(2) and cyclopentadiene (3) (Scheme 1). It is recognized that the
pharmacological profile of NRB is highly sensitive to its stereoiso-
meric composition,5 yet to our knowledge no comprehensive en-
deavor has been made to impart any stereochemical control in its
manufacture.
in combination with other Lewis acids, including Li(I),20 Ti(IV),20
Ni(II),20 Cu(I),20 Cu(II),20 Nb(V),20 In(III),22 Yb(III)20 and Bi(III),20
along with silicon-based Lewis acids such as TMSOTf20 and
TMSCl.20
Alternatively, Scettri and co-workers described aza-Michael-
type reactions employing cinchona alkaloids as chiral ligands not
requiring the participation of a Lewis acid.23 In the absence of
a Lewis acid the catalytic properties of cinchona alkaloids can be
Scheme 1. a) See Tables 1e6 for conditions; b) see Tables 7e9 for conditions.
With a view to devising a stereoselective synthesis of key NRB
precursor 2-fulvenylmethanol 4, Ojida and co-workers reported
that high enantioselectivity could be achieved in the Reformatsky
reaction of 2-pyridyl ketones, including 2-benzoylpyridine (2),
when carried out in the presence of chiral ligands such as cinchona
alkaloids (Fig. 2);17 cinchona alkaloids being among the most
prominent and readily available chiral 1,2-amino alcohols.
reasonably explained through the proposed transition state model
of Jørgensen,24 in which the hydroxyl group of, for example, cin-
chonine (8) establishes a hydrogen bond with the carbonyl group of
the electrophile, while the quinuclidine group coordinates to the
accompanying nucleophile.
Examples of non-cinchona alkaloid-type chelation-assisted
enantioselective reactions involving ketones, where an sp2-nitro-
gen again acts as a directing group, utilize chiral bidentate bisox-
azoline box-type ligands (Fig. 2) in combination with Lewis acids
such as Mg(II), Cu(II), Zn(II) and Yb(III).21,25 More recently, Pedro
and co-workers reported an asymmetric Henry reaction describing
the enantioselective addition of nitromethane to acylpyridine N-
oxides in the presence of box-Cu(II) complexes.26 The chelating
properties of 2-acylpyridine N-oxides in asymmetric metal-cata-
lysed reactions are well known.27e30
It was postulated that chelation of the pyridine sp2-nitrogen
adjacent to the reactive carbonyl centre of the ketone contributed
to such enantiotopic face discrimination by serving as a co-
ordination site for zinc during chelate formation.17 The stereo-
selective addition of dialkyl zinc reagents to aromatic aldehydes,
catalysed by cinchona alkaloids, has also been described.18 Cin-
chona alkaloids have similarly been studied in enantioselective
Michael19 and aza-Michael-type reactions.20 For examples not in-
volving organozinc nucleophiles, a stoichiometric amount of a zin-
c(II) salt, such as Zn(OTf)2, may be introduced to the reaction as an
additive.21 Alternatively, cinchona alkaloid ligands have been used
In terms of inducing stereocontrol in the reaction of 2-
fulvenylmethanol 4 and maleimide (6); cycloadditions per se are
known to be highly sensitive to the conditions employed in their
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
3
Fig. 2. Cinchonidine (7), cinchonine (8), and box ligands (9) and (10).
synthesis. The effect of solvent on intermolecular DielseAlder re-
actions has been well documented, though many reports, despite
featuring a broad range of reaction media, have only been con-
cerned with overall reaction rate.31,32 Extensive work has been
done on modelling the solvent effect mechanisms involved in
DielseAlder reactions. Three main types of solute-solvent in-
teractions have been proposed in an attempt to explain the ob-
served variations in rate through modification of the reaction
medium, namely; solvophobic effects, hydrogen bond donor (HBD)
interactions and dipolarity-polarizability.33 The increase in reaction
rate observed in aqueous solvents has largely been explained by the
high solvophobic power of such media.34 Likewise, the use of sol-
vents with high HBD ability, namely fluorinated alcohols, also lead
to increases in reaction rate.35 It has been speculated that fluori-
nated alcohols can act not only as mild Lewis acids, but as true
Brønsted acids, giving rise to proton transfer catalysis mecha-
nisms.36 In terms of stereochemical outcome, fewer studies have
focused exclusively on the influence of solvent on endo/exo
selectivity.31,36e38 The HBD power of the solvent and its polarity,
together with its ability to induce hydrophobic interactions, have
been considered as the main factors favouring the formation of the
endo-adduct.39 Hydrophobic effects are assumed to stabilize the
more compact endo transition state (TS), versus the extended exo
TS.39
By far the most common method of catalysis in the promotion of
DielseAlder cycloadditions uses Lewis acids; in organic solvents,
rate accelerations in the order of 104e106 have been observed, often
accompanied by considerable increases in selectivity.40,41 As a re-
sult, Lewis acid catalysis allows DielseAlder reactions to be carried
out under mild conditions. The most commonly used traditional
Lewis acids are halides of boron, aluminium, titanium, copper, zinc
and tin, along with an array of lanthanide complexes. In summary,
the effect of temperature (conventional heating and microwave
irradiation), concentration, reaction solvent (including ionic liq-
uids) and Lewis acid on the DielseAlder synthesis of NRB will be
investigated.
mixture (original studies16 described isolated products), largely
confirmed such observations (Table 1, Scheme 1).
Built on the literature platform outlined above, in an attempt to
influence the stereochemical outcome, the synthesis of 2-
fulvenylmethanol 4 was conducted in the presence of stoichio-
metric equivalents of cinchonidine (7) (chiral ligand) and Zn(OTf)2
(Lewis acid) (Table 2, Scheme 1); in conjunction with reaction
conditions earlier optimized for the preparation of compound 4.
Disappointingly, irrespective of such conditions imparting the de-
sired stereochemical bias, the condensation reaction itself failed.
On repeating the experiment in the absence of Zn(OTf)2, 2-
fulvenylmethanol 4 was again furnished in good yield, though
without stereochemical induction (thus confirming that cinchoni-
dine (7) was not operating as a chiral ligand in the absence of
a Lewis acid). Verification that the Lewis acid, rather than the Lewis
acid/ligand complex, was exclusively responsible for reaction sup-
pression was achieved by performing the reaction in the absence of
the ligand; again no product formation was observed. In order to
confirm that Zn(OTf)2 was not interfering with the in situ genera-
tion of the cyclopentadienyl species, pre-formed cyclopentadienyl
sodium (as a commercially available solution in THF) was added to
a mixture of ketone 2, chiral ligand and Lewis acid, again resulting
in no reaction. Given that many literature examples17,18 feature an
organozinc species as the nucleophile, the use of cyclopentadienyl
zinc chloride43 in combination with cinchonidine (7) was briefly
explored, but again without success (only starting materials re-
covered). On repeating the reaction of ketone 2 with cyclo-
pentadienyl sodium in the presence of cinchonidine (7) and
Zn(OTf)2, but on this occasion allowing the reaction to warm to
room temperature, 2-fulvenylmethanol 4 was formed, albeit in low
yield (20%; no stereoselectivity was observed). Given such findings,
all subsequent reactions were now allowed to warm to room
temperature (if no product was observed following 4 h at 10 ꢀC),
and the reaction monitored until no further progress was noted.
A brief study was now conducted using cinchonidine (7) in
combination with other Lewis acids (known literature partners to
cinchona alkaloids), namely Ti(OiPr)4,20 LiClO4,20 Yb(OTf)3,20
InCl3,22 ZnCl2, CuCl2,20 CuCl20 and TMSCl20 (Table 3, Scheme 1).
Performing the reaction in the presence of hard Lewis acids44,45
2. Results and discussion
2.1. Attempted stereoselective synthesis of 2-
fulvenylmethanol 4
Table 1
Influence of concentration, temperature and time on the formation of 2-
fulvenylmethanol 4 (see Scheme 1, step a)
Original work conducted by Mohrbacher and Poos studying the
reaction of 2-benzoylpyridine (2) and cyclopentadienyl sodium
(generated in situ from freshly prepared cyclopentadiene (3) using
sodium ethoxide) reported both product distribution and yield to
be highly sensitive to the reaction conditions employed (Scheme
1).16,42 It was revealed that lower reaction temperatures and
a concentrated reaction medium favoured a second condensation
step with ketone 2, to predominantly afford 2-fulvenylmethanol 4
over the corresponding mono-condensation product, fulvene 5
(formation of fulvene 5 was confirmed through 1H NMR analysis of
its corresponding dimerization product). A brief in-house study,
based on contemporary RP-HPLC analysis of the crude reaction
b
Entrya
Conc. (M)
Temp (ꢀC)
Time (h)
%
4
5
2
1
2
3
4
5
0.5
0.5
0.25
0.25
0.1
10
25
10
10
10
4
4
4
8
4
81
75
75
76
48
5
10
7
19
9
14
15
18
5
43
a
Cyclopentadiene (3) (1 equiv) added to a solution of ketone 2 (2 equiv) and
sodium ethoxide (2 equiv) in EtOH.
b
As determined by achiral RP-HPLC analysis of the crude reaction mixture (rel-
ative percentage based on area under the curve at
SEꢁꢂ1%).
l
¼254 nm, mean of n¼2,
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

4
M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
Table 2
Table 4
Influence of cinchonidine (7) (chiral ligand), Zn(OTf)2 (Lewis acid) and the cyclo-
pentadienyl (Cp) counterion on the formation of 2-fulvenylmethanol 4 (see Scheme
1, step a)
Stereochemical influence of chiral ligand and Lewis acid on the formation of 2-
fulvenylmethanol 4 and 2-fulvenylmethanol N-oxide 13 (see Scheme 1, step a)
b,c
b,d
Entrya
Ligand
Lewis acid
%
%
Entrya Ligand
Lewis acid Cp counterion Solvent
%
b
Cis
Trans
4a
4b
4c
4d
4
5
2
1e
2e
3e
4e
5e
6e
7f
d
d
50
62
58
44
20
42
42
40
50
38
42
56
80
58
58
60
25
26
30
22
10
20
d
25
36
28
22
10
22
d
25
20
22
30
40
28
d
25
18
20
26
40
30
d
1
d
d
Na
EtOH
75
0
76
0
0
0
7
0
6
0
0
0
18
100
18
100
100
100
Cinchonidine
Cinchonidine
Cinchonine
Box 9
Box 10
Box 9
ZnCl2
CuCl
CuCl
Zn(OTf)2
Zn(OTf)2
Mg(OTf)2
Mg(OTf)2
2
Cinchonidine Zn(OTf)2
Na
EtOH
3
Cinchonidine
d
d
Na
Na
Na
ZnCl
EtOH
EtOH
THF/EtOH
THF/EtOH
4
Zn(OTf)2
5c
6c
Cinchonidine Zn(OTf)2
Cinchonidine
d
8f
Box 10
d
d
d
d
a
Cyclopentadiene (3) (1 equiv) was added to ketone 2 (2 equiv), sodium ethoxide
(2 equiv), chiral ligand (0 or 2 equiv) and Lewis acid (0 or 2 equiv) in EtOH (unless
stated otherwise), overall reaction concentration¼0.25 M, 10 ꢀC, 4 h.
a
See Tables 3 and 5 for reaction conditions.
Relative percentage based on area under the curve at
b
l¼254 nm, mean of n¼2,
b
As determined by achiral RP-HPLC analysis of the crude reaction mixture (rel-
SEꢁꢂ1%.
c
ative percentage based on area under the curve at
SEꢁꢂ1%).
l¼254 nm, mean of n¼2,
As determined by achiral RP-HPLC analysis of the crude reaction mixture.
As determined by chiral NP-HPLC analysis of the crude reaction mixture.
Data for 2-fulvenylmethanol 4.
d
e
f
c
Pre-formed cyclopentadienyl species (1 equiv, 2 M in THF) was added to ketone
2 (2 equiv) and chiral ligand (2 equiv) in THF/EtOH (1:1, v/v), overall reaction
concentration¼0.25 M, 10 ꢀC, 4 h.
Data for 2-fulvenylmethanol N-oxide 13.
trans component constituted a 1:1 mixture of erythro and threo
isomers (as determined by chiral NP-HPLC). Selectivity for the cis
isomers was also demonstrated in reactions conducted in the
presence of CuCl (ca. 1.4:1, cis/trans); again, no apparent selectivity
for either the erythro or threo configuration was observed. Mar-
ginally better yields were also noted in the formation of 2-
fulvenylmethanol 4 (49%) using CuCl. In an effort to demonstrate
that the observed stereochemical control (putatively accredited to
the influence of the chiral ligand in partnership with the Lewis acid)
could be inverted, reactions were repeated in the presence of cin-
chonidine’s pseudoenantiomer, cinchonine (8). Significantly, re-
actions performed in the presence of cinchonine (8) and CuCl (our
highest yielding chiral ligand/Lewis acid system) successfully
afforded 2-fulvenylmethanol 4 exhibiting inverted cis/trans selec-
tivity (ca. 1:1.3); compared to when using cinchonidine (7) in
combination with CuCl. Again, no discernable differences were
observed in the ratio of erythro/threo products. Given that the
selectivites reported for other cinchona alkaloids such as quinine
and quinidine are known to be comparable to those of cinchonine
(8) and cinchonidine (7), the application of 1,2-amino alcohols as
chiral ligands was taken no further.17,18
Table 3
Influence of chiral ligand (cinchona alkaloid) and Lewis acid on the formation of 2-
fulvenylmethanol 4 (see Scheme 1, step a)
Entrya
Ligand
Lewis acidb
%
c
4
5
2
1d
2e
Cinchonidine
Cinchonidine
Cinchonidine
Cinchonidine
Cinchonidine
Cinchonidine
Cinchonidine
Cinchonidine
Cinchonidine
Cinchonidine
Cinchonine
d
64g
0
21
0
1
7
0
5
0
0
5
0
36
70
42
97
87
80
55
100
51
95
62
Ti(OiPr)4
LiClO4
Yb(OTf)3
InCl3
9g
3d
4d
5d
6d
7f
58g
2g
6g
Zn(OTf)2
ZnCl2
CuCl2
CuCl
TMSCl
CuCl
20g
40h
0
8d
9d
10d
11d
49h
0
38h
a
Cyclopentadienyl sodium (1 equiv) in THF (2 M) was added to ketone 2
(2 equiv), chiral ligand (2 equiv) and Lewis acid (2 equiv) in THF/EtOH (1:1, v/v),
overall reaction concentration¼0.25 M, 10 ꢀC, 4 h, then rt, 18e120 h. Reactions were
monitored until the point at which the reaction progressed no further.
b
Lewis acids (excl. TMSCl) ordered from hard to soft downwards using Pearson’s
classification,44 as according to Zhang.45
c
As determined by achiral RP-HPLC analysis of the crude reaction mixture (rel-
Instead, we now proceeded to study the influence of box-type
ative percentage based on area under the curve at
l
¼254 nm, mean of n¼2,
ligands, namely box ligand 9, together with Mg(OTf)2,25
SEꢁꢂ1%).
d
25
25
Cu(OTf)2,21 Zn(OTf)2 and Yb(OTf)3 (known literature partners
to box ligands), on the formation of 2-fulvenylmethanol 4 (Table 5,
Scheme 1). Note, in terms of trying to reveal a commercially viable
rat toxicant, given the relative cost associated with box ligands, this
component of the study should be considered a largely academic
exercise. Analogous to the cinchona alkaloids, box ligand 9 did not
appear to be acting as a chiral ligand (a racemic mixture of 2-
fulvenylmethanol 4 was observed in the absence of a Lewis acid,
in 54% yield). Of the Lewis acids trialed in combination with box
ligand 9, the presence of Mg(OTf)2, Cu(OTf)2 and Yb(OTf)3 all
appeared to suppress the formation of 2-fulvenylmethanol 4, with
the exception of Zn(OTf)2 (16% after 3 days). Encouragingly, under
these conditions, a trans-enriched mixture of 2-fulvenylmethanol 4
(ca. 1:4, cis/trans) was observed, albeit in low yield. In terms of
erythro/threo configuration, no selectivity was noted. With a posi-
tive result in hand we proceeded to repeat the reaction, this time in
the presence of box ligand 10, in an attempt to reverse the ste-
reochemical outcome. Curiously, the cis/trans selectivity was not
inverted, with the cis isomers again revealed as the predominant
product; moreover, the level of stereoselectivity was found to be
significantly lower (ca. 1:1.5, cis/trans).
t¼4 h.
e
t¼18 h.
f
t¼120 h.
g
No stereoselectivity observed.
h
Stereoselectivity observed (see Table 4).
such as Ti(OiPr)4, Yb(OTf)3 and InCl3 led only to the formation of
small amounts of 2-fulvenylmethanol 4 (<10%). Similarly, the
comparatively soft Lewis acids, CuCl2 and TMSCl, also failed to
furnish any fulvenylmethanol. LiClO4 was found to have minimal
influence on the reaction; when compared directly to the reaction
profile carried out in the absence of both ligand and Lewis acid.
Encouragingly, in the presence of ZnCl2 (in an effort to further
probe the importance of the zinc(II) salt counterion), 2-
fulvenylmethanol 4 was formed in ca. 20% yield following 16 h at
room temperature, which increased to 40% after 5 days; performing
the reaction at 50 ꢀC had no discernable effect on either yield,
product distribution or stereochemical outcome. More gratifyingly,
a cis/trans ratio of 1.6:1, respectively, was observed (as determined
using a combination of achiral RP-HPLC and 1H NMR4 analysis)
(Table 4, Scheme 1).
Next, building on the findings of Pedro and co-workers,26 an
investigation into the reaction of cyclopentadienyl sodium with
Extended analysis revealed the cis products to be present as an
unassigned 1.4:1 mixture of erythro and threo isomers, whereas the
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
5
Table 5
the substrate. Instead, Mg(OTf)2 was shown to be a more effective
Lewis acid partner for box ligand 9, effecting a ca. 1.4:1 cis/trans
ratio, albeit in very low yield (7%); again, substantial fulvene gen-
eration was observed through the formation of N-oxide 14. Re-
peating the experiment using box ligand 10 (in combination with
Mg(OTf)2) resulted in a similar cis/trans ratio (1.5:1), in improved
yield (24%), when again stereochemical inversion was expected.
Unfortunately, despite our best efforts, we were unable to separate
2-fulvenylmethanol N-oxide 13 into its respective isomers (using
chiral NP-HPLC methods). However, given that the results displayed
no improvement over those observed working with 2-
benzoylpyridine, such an approach was discontinued.
Finally, stereobiased samples of 2-fulvenylmethanol 4, obtained
from conducting the reaction of ketone 2 and cyclopentadienyl
sodium in the presence of either cinchonidine (7)/ZnCl2 or box li-
gand 9/Zn(OTf)2 systems, were subsequently reacted with mal-
eimide (6) in the DielseAlder synthesis of NRB, and will be
discussed in Section 2.2.
Influence of chiral ligand (box ligand) and Lewis acid on the formation of 2-
fulvenylmethanol 4 (see Scheme 1, step a)
Entrya
Ligand
Lewis acidb
%
c
4
5
2
1d
2
3
4
5
Box 9
Box 9
Box 9
Box 9
Box 9
Box 10
d
54e
8
24
0
0
1
0
22
92
94
83
100
85
Mg(OTf)2
Yb(OTf)3
Zn(OTf)2
Cu(OTf)2
Zn(OTf)2
6
16f
0
6
15f
0
a
Cyclopentadienyl sodium (1 equiv) in THF (2 M) was added to ketone 2
(2 equiv), chiral ligand (2 equiv) and Lewis acid (2 equiv) in THF/EtOH (1:1, v/v),
overall reaction concentration¼0.25 M, 10 ꢀC, 4 h, then rt, 18 h (unless stated
otherwise). Reactions were monitored until the point at which the reaction pro-
gressed no further.
b
Lewis acids ordered from hard to soft downwards using Pearson’s classifica-
tion,44 as according to Zhang.45
c
As determined by achiral RP-HPLC analysis of the crude reaction mixture (rel-
ative percentage based on area under the curve at
l
¼254 nm, mean of n¼2,
SEꢁꢂ1%).
d
t¼2 h.
2.2. Effect of temperature, concentration and reaction sol-
vent on the DielseAlder reaction of 2-fulvenylmethanol 4 and
maleimide (6)
e
No stereoselectivity observed.
f
Stereoselectivity observed (see Table 4).
benzoylpyridine N-oxide 1226 in the presence of different metal-
ligand systems was now carried out. Correspondingly, several
complexes of di- and trivalent metal triflates (Mg(II), Cu(II), Zn(II)
and Yb(III))26 in combination with box ligands 9 and 10 were
screened; using a procedure analogous to that developed for the
asymmetric preparation of 2-fulvenylmethanol 4 (Table 6, Scheme
1). In keeping with earlier observations, box ligand 9 did not appear
to function as a chiral ligand since no stereocontrol was observed in
the formation of 2-fulvenylmethanol N-oxide 13. However, an im-
proved yield of 94% was recorded (vs 77% in the absence of the
chiral ligand); this would appear to suggest that box ligand 9 may
be activating the ketone towards nucleophilic attack (possibly
through coordination), thus facilitating the reaction. Of the Lewis
acids trialed in combination with box ligand 9, Yb(OTf)3 and
Cu(OTf)2 both appeared to suppress the formation of 2-
fulvenylmethanol N-oxide 13; in the example of Yb(OTf)3, signifi-
cant amounts of the corresponding fulvene 14 were observed. In
contrast to experiments using 2-benzoylpyridine (2) in combina-
tion with box ligand 9 and Zn(OTf)2, no apparent stereocontrol was
observed when 2-benzoylpyridine N-oxide (12) was employed as
Mohrbacher and Poos first accessed NRB through the Diel-
seAlder reaction of 2-fulvenylmethanol 4 and maleimide (6) in
refluxing PhH, to reveal an endo rich mixture of isomers (Table 7,
Scheme 1).16 Of the other independent NRB syntheses reported in
the literature, Brimble and co-workers did not report crude ste-
reoisomeric composition.4 A brief in-house study was conducted to
ascertain the effect of temperature and concentration on the re-
action of 2-fulvenylmethanol 4 and maleimide (6) in mixed xy-
lenes; selected as a reaction solvent over PhH on the basis of its
greater thermal operating range (Table 7, Scheme 1). Similar ob-
servations in stereoisomeric composition were observed at 50 ꢀC
and 80 ꢀC (at 25 ꢀC the reaction proceeded very slowly), while in-
creasing the reaction temperature to 110 ꢀC led to greater pro-
portions of exo-adducts. This was significantly more pronounced at
140 ꢀC where exo-isomers were found to contribute to >90% of the
crude reaction mixture, thus confirming exo-NRB as the thermo-
dynamic product. The proportion of erythro/threo isomers was also
found to be strongly influenced by temperature. Curiously, it was at
Table 7
Influence of temperature and concentration on the DielseAlder reaction between 2-
fulvenylmethanol 4 and maleimide (6) (see Scheme 1, step b)
Table 6
Entrya
Temp (ꢀC)
Stereochemical compositionb
% (1)b
Influence of chiral ligand (box ligand) and Lewis acid on the formation of 2-
fulvenylmethanol N-oxide 13 (see Scheme 1, step a)
Endo/Exo
Cis/Trans
Threo/Erythro
1c
2
3
d
5.7
d
1.1
d
1.4
d
d
c
Entrya
Ligand
Lewis acidb
%
25
50
80
80
80
80
110
140
<5
13
14
12
4.0
3.8
100
4.6
4.0
1.8
0.08
1.0
1.0
1.0
1.7
0.3
0.9
1.0
1.9
1.8
2.5
1.6
1.3
1.4
0.8
100
100
d
1
2
3
4
5
6
7
d
d
77d
94d
7e
0
0
24
29
0
23
6
4d
5e
6f
7g
8
Box 9
Box 9
Box 9
Box 9
Box 9
Box 10
d
Mg(OTf)2
Yb(OTf)3
Zn(OTf)2
Cu(OTf)2
Mg(OTf)2
69
71
91
100
50
100
100
100
100
0
9d
0
0
26
9
24e
a
2-Fulvenylmethanol 4 (1 equiv) and maleimide (6) (1 equiv) in mixed xylenes
a
Cyclopentadienyl sodium (1 equiv) in THF (2 M) was added to ketone 12
(2 equiv), chiral ligand (2 equiv) and Lewis acid (2 equiv) in THF/EtOH (1:1, v/v),
overall reaction concentration¼0.25 M, 10 ꢀC, 4 h, then rt, 18 h. Reactions were
monitored until the point at which the reaction progressed no further.
(0.1 M), 16 h, unless stated otherwise.
b
As determined by achiral RP-HPLC analysis of the crude reaction mixture (rel-
ative percentage based on area under the curve at
l¼254 nm, mean of n¼2,
SEꢁꢂ1%).
b
c
Lewis acids ordered from hard to soft downwards using Pearson’s classifica-
Isolated stereoisomeric composition as reported by Mohrbacher et al.;16 re-
tion,44 as according to Zhang.45
action solvent¼PhH, temp¼80 ꢀC.
c
d
As determined by achiral RP-HPLC analysis of the crude reaction mixture (rel-
Similar results were observed at reaction concentrations of 0.05 and 0.2 M.
Precipitated reaction product collected by filtration and washed with mixed
e
ative percentage based on area under the curve at
l
¼254 nm, mean of n¼2,
SEꢁꢂ1%).
xylenes.
d
f
No stereoselectivity observed.
Stereoselectivity observed (see Table 4).
Starting with stereobiased 2-fulvenylmethanol 4 (see Table 4, entry 2).
Starting with stereobiased 2-fulvenylmethanol 4 (see Table 4, entry 5).
e
g
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

6
M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
elevated temperatures that the erythro/threo ratio was found to be
more reflective of that of the starting 2-fulvenylmethanol 4, while
at both 50 ꢀC and 80 ꢀC significant migration towards the threo
isomers was noted. Conversely, the cis/trans isomeric ratio
remained constant throughout. Concentration was found to have
little effect on the reaction. On investigating standard literature
methods for the isolation of NRB (namely filtration of the pre-
cipitated product from the cooled reaction medium), it soon be-
came apparent that the solubility profiles of the individual isomers
were not equal; likely providing further explanation for why
commercially sourced NRB mixtures varied so radically in terms of
stereoisomeric composition. On the basis that an endo-exclusive
mixture was revealed post isolation (via filtration), it was plainly
evident that the exo-isomers of NRB were significantly more solu-
ble in the reaction solvent (xylenes) than their endo counterparts.
In terms of other notable variations in stereochemical composition,
the relative proportion of threo and erythro isomers was more
pronounced in favour of the former, whereas the cis/trans ratio was
unaffected.
As expected, the reaction of maleimide (6) with stereobiased
samples of 2-fulvenylmethanol 4 (generated as described in Section
2.1) in xylenes at 80 ꢀC led to endo/exo compositions similar to
those observed using racemic starting materials. In terms of cis/
trans ratio, products emanating from reactions using 2-
fulvenylmethanol 4 with an approximate cis/trans ratio of 1.6:1
(Table 4, entry 2) led to cis-enriched DielseAlder adducts, whereas
trans-adducts were the predominant product of reactions starting
with 2-fulvenylmethanols rich in trans isomers (ca. 1:4, cis/trans;
Table 4, entry 5); in essence, the cis/trans ratio of the starting ful-
venylmethanol was transferred to that of the DielseAlder product.
With regards to erythro/threo configuration, given that it was not
possible to assign the stereoisomers of 2-fulvenylmethanol 4 as
either erythro or threo, it is difficult to draw any firm conclusions
from this result other than to comment that a threo-enriched
product was observed, albeit slightly less pronounced than in ex-
amples using racemic starting materials.
component was raised; a similar stereochemical outcome to that of
reactions performed solely in water was noted.
While the DielseAlder reaction has been studied in a wide range
of molecular solvents, ionic liquids offer a distinctly unique solvent
environment, one constituted entirely of ions;46 notable advan-
tages including their large thermal operating range and the op-
portunity to conduct reactions under substantial pressure. Ionic
liquids are also recyclable, thus generating minimal waste prod-
uct.47 Performing the DielseAlder reaction of 2-fulvenylmethanol 4
and maleimide (6) in 1-butyl-3-methylimidazolium hexa-
fluorophosphate ([bmim][PF6]) led to stereoisomeric ratios similar
to those observed in water.
Under classical heating, cycloaddition reactions often require
higher temperatures, longer reaction times, and/or Lewis acid ca-
talysis, occasionally resulting in the partial decomposition of sen-
sitive compounds. Remarkable rate accelerations and improved
yields have been observed through performing DielseAlder re-
actions under microwave irradiation conditions.48,49 Significantly,
it has been reported that no kinetic differences should be expected
between reactions promoted by microwaves and those heated us-
ing conventional methods; so long as careful temperature control is
employed.48 Moreover, studies by Gedye and co-workers reported
that microwave radiation does not influence the endo/exo selec-
tivity in DielseAlder reactions.50 To study the effect of microwave
energy on the DielseAlder reaction of 2-fulvenylmethanol 4 and
maleimide (6), three different reaction media were selected based
on their physical properties (non-polar, polar protic and polar
aprotic), namely PhMe, EtOH and DMF, respectively, observed ac-
celerations are known to be more significant in apolar solvents that
show weak dielectric losses.48 To ensure that the type of energy
employed remained the focus of the experiment, and to safeguard
that overheating was not a factor, reaction temperatures were
carefully controlled and limited to 80 ꢀC. In terms of conversion,
there was no discernible difference between conventional and
microwave heating, though it was observed that reactions con-
ducted in either PhMe or EtOH required shorter reaction times (ca.
6e8 h) than those using DMF (ca. 16 h), whether conventionally
heated or using microwaves. As largely anticipated based on the
literature, no variations in stereoisomeric composition were noted.
With regards to plausible reasoning for the observed differences
in endo/exo-selectivity in terms of performing the DielseAlder re-
action in different reaction media, it is tentatively postulated that
under certain conditions the kinetic endo-product of the cycloaddi-
tion interconverts to the more thermodynamically stable exo-prod-
uct, presumably via a retro-DielseAlder process. It has been reported
that the DielseAlder products formed in the reaction between
maleic anhydride and a range of fulvenes readily dissociate in so-
lution either at ambient temperatures, or upon heating.51 The
diphenylfulvene adducts were revealed to rapidly dissociate in both
glacial acetic acid and ethyl acetate solutions, whereas cyclo-
reversion was observed to be considerably slower in benzene. Sim-
ilarly, the reaction of 6,6-pentamethylenefulvene with maleic
anhydride at room temperature in benzene led to the exclusive
generation of the endo-adduct; however, upon allowing the reaction
mixture to stand for several weeks, only the exo-adduct was re-
covered.52 More pertinent to this study, it is known that the endo
isomers of NRB can be thermally converted to their corresponding
exo configurations in benzene.16 Given that polar solvents have been
demonstrated to increase the rate of both the DielseAlder53 and
retro-DielseAlder reactions51 of fulvene adducts, it is reasonable to
suggest that solvent polarity could influence reactions involving 2-
fulvenylmethanol 4 in similar fashion. As such, it is plausible that
polar solvents could similarly facilitate the retro-DielseAlder re-
action of endo-NRB and, as a consequence, increase the relative
quantity of the more thermodynamically stable exo-adducts.
To probe the effect of reaction medium on the DielseAlder re-
action of 2-fulvenylmethanol 4 and maleimide (6), an exhaustive
range of common laboratory solvents were now screened; listed in
order of solvent dielectric constant as an approximate measure of
polarity. Gratifyingly, a remarkable variation in stereochemical
outcome was observed (Table 8, Scheme 1). While there did not
appear to be a strict correlation with either solvent dielectric con-
stant (
r) or normalized polarity (ENT ), it was clearly apparent that
3
performing the reaction in strongly polar solvents (e.g., DMSO,
DMF, ethylene glycol, alcohols, water) led to a higher proportion of
exo-adducts; regardless of the solvent being protic or aprotic in
nature (Fig. 3). Notable exceptions included the chlorinated sol-
vents DCE and DCB, which afforded the most endo-enriched mix-
tures of all the media evaluated, while both n-hexane and
cyclohexane led to endo/exo proportions more akin to those found
using more polar solvents (e.g., DME, THF, MEK) than aromatic
hydrocarbons such as PhH, PhMe and mixed xylenes. Brønsted
acidic solvents, particularly AcOH, also gave higher proportions of
exo-adducts, comparable to those found using water; reactions
using AcOH and formic acid were performed at room temperature
owing to appreciable side-product formation at elevated temper-
atures. Similar observations in endo/exo stereoselectivity were
likewise noted using fluorinated alcohols such as TFE and HFIP.
Given the limited solubility of 2-fulvenylmethanol 4 in water,
aqueous solvent mixtures were also investigated to further probe
the well documented high solvophobic potential of such reaction
media. DielseAlder reactions were subsequently performed in
three different aqueous THF mixtures (25%, 50% and 75% v/v),
leading to higher conversions as the proportion of the organic
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
7
Table 8
Influence of reaction solvent on the DielseAlder reaction between 2-fulvenylmethanol 4 and maleimide (6) (see Scheme 1, step b)
63
64
Entrya
Solventb
Dielectric constant (
3
)
Normalized polarity ðETN
Þ
Stereochemical compositionc
r
Endo/Exo
Cis/Trans
Threo/Erythro
% (1)c
1
2
3
4
n-Hexane
Cyclohexane
Dioxane
PhH
PhMe
PhMe
Xylenes
MTBE
EtOAc
AcOH
AcOH
DME
THF
THF
DCB
DCE
DCE
Pyridine
HFIP
MEK
1.965
2.0
2.2
2.3
2.4
2.4
2.3e2.6
4.065
6.0
6.2
6.2
7.2
7.6
7.6
966
1065,67
1065,67
12
0.009
0.006
0.164
0.111
0.099
0.099
0.074
0.124
0.228
0.648
0.648
0.231
0.207
0.207
0.225
0.327
0.327
0.302
0.969
0.327
0.654
0.654
0.898
0.762
0.762
0.324
0.481
0.460
0.79
2.6
2.5
2.0
3.8
3.8
4.0
3.8
3.4
2.9
d
0.9
1.0
1.0
1.0
1.1
1.0
1.0
0.5
1.0
d
1.5
1.6
1.4
1.8
1.8
1.8
1.8
1.5
1.6
d
85
80
95
100
100
98h
99
48
97
5
6d
7
8
9
10
11e
12
13
14f
15
16
17f
18
19
20
21
22d
23
24
25f
26
27
28
29
30
31d
32
33f
34
35e
36
37
38
39
40
NR
85
89
1.6
2.5
2.6
3.2
4.3
4.3
4.6
1.6
1.7
2.7
1.3
1.3
1.5
1.1
1.1
2.7
2.5
1.9
0.9
1.0
1.0
0.5
0.5
d
0.9
1.0
0.9
0.8
1.1
1.0
1.0
1.0
1.1
0.9
1.0
1.0
1.1
0.9
1.0
1.1
1.1
1.0
1.0
0.9
1.0
0.9
1.0
d
1.4
1.5
1.6
1.6
1.8
2.0
2.0
1.5
1.7
1.6
1.4
1.4
1.8
1.4
1.4
1.8
1.8
1.6
1.3
1.5
1.4
1.3
1.2
d
90
70g
80
100
87g
87
92
93
100
96i
98
1767
18
EtOH
EtOH
TFE
MeOH
24
24
2668
32
90
MeOH
32
34
35
37
37
37
37
46
46
58
58
80
d
87g
98
PhNO2
MeNO2
MeCN
EG
DMF
DMF
DMSO
DMSO
Formic acid
Formic acid
H2O
H2OeTHF (1:3 v/v)
H2OeTHF (1:1 v/v)
H2OeTHF (3:1 v/v)
[bmim][PF6]
98
94
93
0.386
0.386
0.444
0.444
0.728
0.728
1
94
95j
90
69g
NR
50
1.9
1.5
1.5
1.6
1.8
1.6
0.8
0.8
1.0
1.1
0.8
1.0
1.2
1.4
1.4
1.4
1.4
1.6
52
83
83
60
d
d
d
d
11.469
d
0.66970
93
DCB¼1,2-dichlorobenzene; DCE¼1,2-dichloroethane; Dioxane¼1,4-dioxane; DME¼1,2-dimethoxyethane; DMF¼N,N-dimethylformamide; DMSO¼dimethyl sulfoxide;
EG¼ethylene glycol; HFIP¼1,1,1,3,3,3-hexafluoro-2-propanol; MEK¼methyl ethyl ketone; MTBE¼tert-butyl methyl ether; TFE¼2,2,2-trifluoroethanol; THF¼tetrahydrofuran;
[bmim][PF6]¼1-butyl-3-methylimidazolium hexafluorophosphate. NR¼no result, reaction led to multiple decomposition/side-products. ENT is an empirical measure of polarity
derived from the solvatochromism of a pyridinium N-phenolate betaine based on a scale of 0 (trimethylsilane) to 1 (water).
a
2-Fulvenylmethanol 4 (1 equiv) and maleimide (6) (1 equiv) in solvent (0.1 M), 16 h.
b
Only solvents with boiling points above ca. 55 ꢀC were considered due to the perceived thermal requirements of the reaction; reactions were performed at either 80 ꢀC, or
at reflux if solvent bp below 80 ꢀC (using conventional heating), unless stated otherwise.
c
As determined by achiral RP-HPLC analysis of the crude reaction mixture (relative percentage based on area under the curve at
l¼254 nm, mean of n¼2, SEꢁꢂ1%).
d
e
f
Microwave heating, not exceeding 80 ꢀC.
Performed at 25 ꢀC.
2.5 mmol scale.
Isolated yield post silica gel chromatography.
t¼6 h; reaction 100% complete at same time-point using conventional heating.
t¼8 h; reaction 92% complete at same time-point using conventional heating.
t¼16 h; reaction 100% complete at same time-point using conventional heating.
g
h
i
j
In terms of other notable variations in stereochemical outcome,
solubility of trans 2-fulvenylmethanol 4 in the reaction solvent,16
leading to greater proportions of the trans-NRB products.
the proportion of erythro/threo isomers again appeared to be sig-
nificantly influenced by reaction solvent, as it was by temperature.
DielseAlder reactions performed in aromatic hydrocarbons (e.g.,
PhH, PhMe, mixed xylenes), chlorinated solvents (e.g., DCB, DCE)
and nitrated solvents (e.g., PhNO2, MeNO2) led to the greatest bias
towards the threo isomers, whereas DMSO and formic acid gave
erythro/threo ratios closer to parity. In the main, the cis/trans iso-
meric ratio remained constant throughout; a significant outlier was
noted when the DielseAlder reaction was performed in MTBE. It
was observed that 2-fulvenylmethanol 4 was not entirely soluble in
the reaction medium, even under reflux, and as a consequence the
reaction failed to go to completion in the allotted time (18 h); it is
postulated that the distorted cis/trans ratio is due to the greater
2.3. Effect of Lewis acid on the DielseAlder reaction of 2-
fulvenylmethanol 4 and maleimide (6)
To study the role of Lewis acid catalysis in the synthesis of NRB,
the DielseAlder reaction of 2-fulvenylmethanol 4 and maleimide
(6) was carried out in the presence of an array of Lewis acids (Table
9, Scheme 1). Given that Zn(II) salts have literature precedence in
DielseAlder cycloadditions featuring maleimides as the dien-
ophile,54 preliminary experiments were conducted using stoichio-
metric equivalents of ZnCl2. A brief study to determine the optimal
reaction solvent revealed DCM to give the best conversions (vs THF
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

8
M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
Fig. 3. Influence of solvent on the DielseAlder reaction of 2-fulvenylmethanol 4 and maleimide (6) (endo/exo ratio above bar) (in conjunction with Table 8).
and PhMe); from hereon in all Lewis acid catalysed reactions were
performed in DCM unless stated otherwise. Although little im-
provement in yield was noted (vs the Lewis acid-free control), it
was immediately apparent that ZnCl2 had a notable influence on
endo/exo selectivity, affording a significantly endo-enriched prod-
uct. Perhaps even more strikingly, the erythro/threo ratio was also
found to be strongly affected by the Lewis acid, with the erythro
isomer predominating. To further probe this finding, and in an ef-
fort to determine at which stage of the synthesis such erythro/threo
epimerization was taking place, the reaction was repeated but on
this occasion in the absence of the dienophile (maleimide (6)).
Chiral NP-HPLC subsequently revealed the stereochemical com-
position of 2-fulvenylmethanol 4, post exposure to the Lewis acid,
to be unaffected. Similarly, in parallel, a sample of purified NRB
(NRB 1a) demonstrated to be rich in threo isomers, was stirred in
the presence of ZnCl2 to determine whether the erythro/threo
configuration was generated pre- or post-NRB formation. No
change in stereoisomeric composition was noted. In another ex-
periment, the reaction was repeated in the presence of excess
ZnCl2, with no discernable improvement in yield; better conver-
sions could be achieved through extended reaction times. It has
also been reported that ultrasonic irradiation (sonochemistry)
provides efficient promotion in DielseAlder cycloadditions, leading
to increased yields and shorter reaction times.55 Sonochemistry has
also been shown to be amenable to large-scale synthesis.56 Per-
forming the DielseAlder reaction of 2-fulvenylmethanol 4 and
maleimide (6) in the presence of ZnCl2 under ultrasonic irradiation
not only led to higher conversions (88%), but also had a significant
impact on stereochemical outcome. As an alternative source of
zinc(II), Zn(OTf)2 was evaluated to determine what influence the
counter-ion had on the reaction. In terms of reaction yield, greater
quantities of NRB were formed using Zn(OTf)2 (73%) compared to
ZnCl2 (46%); on the other hand more side-products were also ob-
served. With regards to stereochemical outcome, reactions carried
out in the presence of Zn(OTf)2 led to the exclusive formation of the
endo-adduct.
On the back of these promising results we now proceeded to
trial an extensive array of common Lewis acids, ranging from strong
Lewis acids (e.g., BF3.(OEt)2, AlCl3 and TiCl4), through to milder
examples such as SnCl2, CuCl2, and CuCl, along with a selection of
lanthanide complexes and silicon-based Lewis acids. Lewis acids
(excl. TMSCl, TMSOTf and montmorillonite) were ordered from
hard to soft using Pearson’s classification,44 as according to Zhang45
(Table 8). With regards to reaction rate enhancement, LiClO4, MgCl2
and YbCl3 undoubtedly gave the best conversions (>90%), while
InCl3 (69%), SnCl4 (67%), CuCl (60%), Sc(OTf)3 (51%), CuCl2 (50%) and
LaCl3 (50%), to varying degrees, also appeared to catalyse the
DielseAlder reaction of 2-fulvenylmethanol 4 and maleimide (6);
whereas FeCl2 (46%), ZrCl4 (40%) and FeCl3 (35%) gave yields com-
parable to the Lewis acid-free control (39%). Montmorillonite K-10
(42%), a mineral clay known to have Lewis acid properties,57 sim-
ilarly failed to promote the reaction. Conversely, BF3$(OEt)2 (15%),
TiCl4 (15%), SnCl2 (26%) and AlCl3 (31%) all seemed to suppress the
reaction, as did the silicon based Lewis acids TMSCl (21%) and
Table 9
Influence of Lewis acid on the DielseAlder reaction between 2-fulvenylmethanol 4
and maleimide (6) (see Scheme 1, step b)
Entrya
Lewis acidb
Stereochemical compositionc
% (1)c
Endo/Exo
Cis/Trans
Erythro/Threo
1
d
8.1
6.1
d
0.8
0.9
d
0.5
0.5
d
39
82
15
15
31
40
92
46
51
4
67
92
35
94
69
66k
50
19
46, 60l
30
10
88
73
46
26
50
60
21
2
2d
3
d
BF3$(OEt)2
TiCl4
AlCl3
4
5
6
d
d
d
d
d
d
ZrCl4
7
100
11
24
d
1.2
1
0.6
0.4
d
4
7
LiClO4
LiClO4
Sc(OTf)3
Sc(OTf)3
SnCl4
MgCl2
FeCl3
YbCl3
InCl3
InCl3
LaCl3
LaCl3
ZnCl2
ZnCl2
ZnCl2
ZnCl2
Zn(OTf)2
FeCl2
SnCl2
CuCl2
CuCl
1.4
0.4
0.5
d
8e
9
10f
11
12
13
14
15
16g
17
18h
19
20i
21j
22d
23
24
25
26
27
28
29
30
100
100
d
1.9
1.2
d
13
16
d
100
100
100
5.7
d
1.2
0.8
0.7
0.7
d
4.3
100
100
0.5
d
49
d
0.5
d
3.6
d
d
d
d
10
1.0
0.6
0.6
d
1.8
5.3
2.9
d
100
8.6
d
100
20
0.4
0.5
d
10
0.3
d
TMSCl
TMSOTf
Montmorillonite
d
d
d
d
4.6
0.8
0.5
42
a
2-Fulvenylmethanol 4 (1 equiv), maleimide (6) (1 equiv) and Lewis acid (0 or
1 equiv) in DCM (0.1 M), 25 ꢀC, 24 h, unless stated otherwise.
b
Lewis acids (excl. TMSCl, TMSOTf and montmorillonite) ordered from hard to
soft downwards using Pearson’s classification,44 as according to Zhang.45
c
As determined by achiral RP-HPLC analysis of the crude reaction mixture (rel-
ative percentage based on area under the curve at
l
¼254 nm, mean of n¼2,
SEꢁꢂ1%); only reactions which generated >40% NRB were considered for stereo-
isomeric composition analysis.
d
With sonication.
5 M lithium perchlorate-diethyl ether (LPDE).
e
f
Solvent¼[bmim][PF6].
g
2.5 mmol scale.
h
Solvent¼H2OeTHF (1:3 v/v).
i
Solvent¼PhMe.
j
Solvent¼THF.
k
Isolated yield post silica gel chromatography.
l
Higher conversions were observed at t¼3 d, whereas adding a second equivalent
of Lewis acid had no effect.
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
9
TMSOTf (2%). In terms of AlCl3, InCl3, FeCl2 and CuCl2, appreciable
side-product formation (>20%) was also noted.
lowest conversions through to NRB were recorded; contrastingly,
reactions which went near to completion, such as those using ei-
ther MgCl2 or YbCl3, had cis/trans product ratios closer to 1:1. It is
thus postulated that a kinetic effect may be operation, with trans 2-
fulvenylmethanol 4 reacting faster than its cis counterpart (po-
tentially under the influence of Lewis acid), therefore subsequently
giving rise to greater proportions of trans-NRB products in the in-
complete reactions.
Grieco and co-workers reported that DielseAlder cycloadditions
carried out in 5 M lithium-perchlorate diethyl ether (LPDE,
LiClO4$Et2O) led to faster reactions and improved stereo-
selectivities.58 Originally, it was assumed that the effect of the LPDE
medium was to create a high internal solvent pressure,58 but more
recently such rate enhancements have been attributed to Lewis
acid catalysis.59 Performing the DielseAlder reaction of 2-
fulvenylmethanol 4 and maleimide (6) in LPDE gave poorer con-
versions compared to those observed when using LiClO4 in DCM
(attributed to the low solubility of compound 4 in diethyl ether);
potentially affecting the stereochemical outcome as a consequence.
While most Lewis acids decompose in the presence of moisture,
it was found that rare earth triflates (e.g., Sc(OTf)3, La(OTf)3,
Yb(OTf)3) could be employed as Lewis acid catalysts in water or
water-containing solvents.60 Not only were these Lewis acids found
to be compatible with water, they were also activated by water,
possibly due to dissociation of the counterions from the Lewis
acidic metal.61 Additionally, such catalysts could be recovered and
reused. Kobayashi and co-workers61 also revealed that it was not
only the lanthanides which could be used in aqueous solvent
mixtures, but also Fe(II), Cu(II) and Zn(II), amongst others. As an
example of a lanthanide Lewis acid, LaCl3 was selected ahead of
Yb(OTf)3 in this study on the basis of cost. Conducting the Diel-
seAlder reaction of 2-fulvenylmethanol 4 and maleimide (6) in the
presence of stoichiometric amounts of LaCl3 in aqueous THF suc-
cessfully afforded the product, but in low yield. Ionic liquids can
also be used in conjunction with Lewis acids such as Sc(OTf)3,
leading to both rate acceleration and improved selectivity, while
also facilitating catalyst recovery.62 Disappointingly, the reaction of
2-fulvenylmethanol 4 and maleimide (6) in Sc(OTf)-[bmim][PF6]
mixtures similarly led to low conversions.
In summary, in terms of the degree of stereocontrol available in
the synthesis of NRB, it has been demonstrated that markedly
varied stereoisomeric compositions could be attained under dif-
ferent conditions. Of the reaction media evaluated, DielseAlder
reactions performed in DCE led to the most endo-rich mixtures
(NRB V, Y, U, W), whereas cycloadditions carried out in DMSO
generated the greatest proportions of accompanying exo-adduct
(NRB R, T, X, S). With regards to erythro/threo configuration, re-
actions employing DCE and DMSO again led to products displaying
the greatest contrast, affording mixtures predominant in erythro
(NRB U, W, X, S) and threo (NRB V, Y, R, T) isomers, respectively. In
the main, the ratio of cis (NRB V, U, R, X) and trans (NRB Y, W, T, S)
products remained largely unchanged regardless of solvent. With
respect to the effect of Lewis acid on the reaction of 2-
fulvenylmethanol 4 and maleimide (6), endo-exclusive mixtures
were revealed for cycloadditions conducted in the presence of
LiClO4, SnCl4, MgCl2, YbCl3, InCl3 and Zn(OTf)2. Equally pronounced
was the influence of Lewis acid on erythro/threo configuration, with
SnCl4, MgCl2 and InCl3 all affording erythro-enriched products. In
the case of the latter erythro isomers were observed to be the ex-
clusive product of the reaction, which in combination with being
free of any exo-adduct revealed a product comprised only of NRB U
and W, and with it the highest level of stereocontrol observed
within this study.
In terms of stereocontrol, it was immediately apparent that
Lewis acid catalysis had a major influence on endo/exo selectivity.
Performing the DielseAlder reaction in the presence of stoichio-
metric amounts LiClO4, SnCl4, MgCl2, YbCl3, InCl3, Zn(OTf)2 and
CuCl2 led exclusively to the endo-adduct, while Sc(OTf)3, ZnCl2 and
CuCl also afforded endo-rich mixtures. Likewise, both the pro-
portion of cis/trans and erythro/threo isomers were significantly
influenced by the Lewis acid (Fig. 4). To varying degrees, reactions
employing ZrCl4, LiClO4, SnCl4, MgCl2, YbCl3, InCl3, ZnCl2, Zn(OTf)2,
FeCl2 and CuCl2 all led to mixtures predominant in erythro isomers.
Remarkably, in the case of InCl3, erythro isomers were the exclusive
product of the reaction. With regards to cis/trans ratios, a deviation
from parity was most noticeable in those examples where the
2.4. Bait-palatability and efficacy trial in SpragueeDawley
rats
In order to provide experimental evidence that NRB batches of
dissimilar stereochemical composition have their own unique
palatability and efficacy profiles in rats, a no-choice bait trial was
commissioned. In terms of constitution, it is tentatively suggested
that increasing proportions of NRB V (cis-endo-threo), the most
toxic isomer of NRB,5 present in each mixture could be responsible,
at least in part, for the low palatability of certain NRB batches.
Earlier, we revealed that by using different reaction solvents in the
DielseAlder synthesis of NRB we were able to prepare mixtures
comprising markedly varied stereoisomeric compositions. Condi-
tions which generated mixtures containing the most NRB V (29%,
NRB 1a), the least NRB V (8%, NRB 1d) and two examples midway
between (20%, NRB 1b and 14%, NRB 1c) were put forward for
evaluation, as was an example completely free of NRB V (NRB 1e,
prepared under the influence of a Lewis acid) (Table 10). Following
1 day ‘pre-baiting’ with a highly palatable bait formulation free of
toxicant, rats were fed bait containing either NRB 1a,1b,1c,1d or 1e
at a concentration of 1.0% w/w (Table 11). Commercially sourced
NRB #1489 was run concurrently as a control. Upon exchanging the
non-toxic bait for bait containing one of the NRB mixtures, con-
sumption levels (based on % body mass over a 2 h feeding period,
n¼5) for NRB 1a (0.21ꢁ0.06), NRB 1b (0.25ꢁ0.04), NRB 1c
(0.19ꢁ0.05) and NRB 1d (0.19ꢁ0.05), versus NRB #1489
(0.83ꢁ0.07), were observed to be significantly lower. Pleasingly,
NRB 1e (0.64ꢁ0.13) was found to be more readily accepted; such an
observation also provided reassurance that our adopted purifica-
tion methods were not inadvertently introducing undetectable
contaminants into our NRB mixtures, which as a consequence could
have been responsible for bait rejection (based on either taste or
smell) in the palatability trial. In terms of mortality, only NRB 1a (3/
Fig. 4. Influence of Lewis acid on the DielseAlder reaction of 2-fulvenylmethanol 4
and maleimide (6) (erythro/threo ratio above bar) (in conjunction with Table 9).
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

10
M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
Table 10
phenomena such as particle size and shape, consequently influ-
encing the rate of dissolution and absorption (and possibly even
metabolism), thus impacting on overall bioavailability and sub-
sequently toxicity; a study which sits outside the scope of this
manuscript.
Stereoisomeric composition of commercially sourced NRB batch#1489, and NRB
batches 1aee (see Fig. 1)
Norbormide batch Stereoisomeric composition (%)a
V
Y
U
W R
T
X
S
Endo/Exo Cis/Trans Threo/Erythro
#1489
1ab
22 45 10 23 0
29 29 10 14 6
20 28 10 18 8
0
3
5
0
5
6
0
4
5
Endo only 0.5
2.0
2.0
1.6
1.4
1.2
Erythro only
4.6
3.2
1.0
0.8
1.0
1.0
1bc
4. Experimental (chemistry)
1cd
14 18 8 13 16 10 13 8 1.1
1de
8
0
11 5 10 19 16 17 14 0.5
40 60 0
4.1. General experimental methods
1ef
0
0
0
0
Endo only 0.7
a
As determined by achiral RP-HPLC analysis (relative percentage based on area
All reagents were used as supplied unless otherwise stated.
Solvents were purified by standard methods. Analytical thin layer
chromatography (TLC) was carried out on pre-coated silica gel
plates (Merck/UV254) and products were visualized by UV fluo-
rescence. Flash chromatography was performed using silica gel
under the curve at
l
¼254 nm, mean of n¼2, SEꢁꢂ1%).
b
See Table 8, entry 17.
See Table 8, entry 14.
See Table 8, entry 25.
See Table 8, entry 33.
See Table 9, entry 16.
c
d
e
f
€
(Riedel-de Haen, particle size 0.032e0.063 mm). Distillation of
cyclopentadiene was carried out using a vigreux apparatus. Melting
points in degrees Celsius (ꢀC) were measured on an Electro-
thermalÒ and X-4 Melting Point Apparatus and are uncorrected.
Infrared spectra were recorded on a PerkineElmer Spectrum One
Fourier Transform IR spectrometer and the absorption maxima are
5) and NRB 1d (1/5) resulted in a lethal endpoint; compared to the
highly efficacious NRB #1489 (5/5)dbased on earlier literature
suggesting poor palatability,8e10 we were surprised by the efficacy
of the commercially sourced NRB control. Given that consumption
levels were largely comparable across NRB 1a-d, it is likely that the
higher levels of mortality observed for NRB 1a simply relates to the
fact that of the mixtures evaluated, it contained the highest pro-
portion of toxic isomers (namely NRB V and NRB Y).5 Similarly,
while consumption levels recorded for NRB 1e (mortality 0/5)
weren’t too dissimilar to those observed for NRB #1489, given that
NRB 1e was free of both isomers V and Y, it is not surprising that
such a mixture was subsequently demonstrated to be inherently
less toxic. In terms of plausible explanations for the low con-
sumption observed for mixtures 1ae1d, given that consumption
levels were largely comparable across the batches, despite their
stereoisomeric constitution being markedly different, it can only be
assumed that an as yet unaccounted phenomena is responsible for
such findings.
expressed in wavenumbers (n
, cmꢃ1). Nuclear Magnetic Resonance
(NMR) spectra were recorded on a Bruker AVANCE DRꢄ400 (1H,
400 MHz) spectrometer at 298 K. For 1H NMR data, chemical shifts
are described in parts per million (ppm) relative to tetramethylsi-
lane (d 0.00) and are reported consecutively as position (dH), rela-
tive integral, multiplicity (s¼singlet, d¼doublet, t¼triplet,
dd¼doublet of doublets, ddd¼doublet of doublet of doublets,
m¼multiplet), coupling constant (J/Hz) and assignment.
General procedure used to study the effect of chiral ligand and/
or Lewis acid on the synthesis of 2-fulvenylmethanol 4: Method A:
To
a solution of 2-benzoylpyridine (2) (2 equiv) or 2-
benzoylpyridine N-oxide (12) (2 equiv), chiral ligand (0 or
2 equiv) and/or Lewis acid (0 or 2 equiv) in ethanol/tetrahydrofuran
(0.5 M, 1:1 v/v) at 10 ꢀC was added dropwise a solution of cyclo-
pentadienyl sodium in tetrahydrofuran (1 equiv, 2 M), and the
mixture stirred at 10 ꢀC for a further 4 h, then at rt if no reaction had
occurred (followed by achiral RP-HPLC and chiral NP-HPLC analysis
of the crude mixture).
3. Conclusions
In summary, we have comprehensively demonstrated that the
stereochemical outcome of the DielseAlder reaction between 2-
fulvenylmethanol 4 and maleimide (6), a key step in the synthesis
of the rat selective toxicant NRB, is highly sensitive to both reaction
medium and Lewis acid. To a lesser extent, we have also shown that
the stereochemistry of pre-cursor 2-fulvenylmethanol 4 can be
influenced using chiral ligands in combination with Lewis acids.
Disappointingly, given the level of stereocontrol now available as
a result of this research, no clear relationship between stereoiso-
meric composition and palatability, in rats, was observed. Plausible
explanations for such findings may include pharmaceutic
General procedures used to study the effect of either solvent
(method B) or Lewis acid (method C) on the reaction of 2-
fulvenylmethanol 4 and maleimide (6): Method B: A solution of
2-fulvenylmethanol 4 (1 equiv) and maleimide (6) (1 equiv) in
solvent (0.1 M) was heated at 80 ꢀC for 16 h (followed by achiral RP-
HPLC analysis of the crude mixture). Method C: To a solution of
maleimide (6) (1 equiv) in dichloromethane (0.1 M) was added
Lewis acid (1 equiv), and the mixture stirred at rt for 10 min. 2-
Fulvenylmethanol 4 (1 equiv) was then added and the mixture
stirred at rt for a further 24 h (followed by achiral RP-HPLC analysis
of the crude mixture).
Table 11
No-choice bait-palatability and efficacy trial observations for commercially sourced NRB batch#1489, and NRB batches 1aee in SpragueeDawley rats (see Table 10 and Fig. 1)
Norbormide batch
Consumption/rata
Mortality
Mean bait consumption (g)d
Mean bait consumption as a % of body mass
Mean NRB dose consumed (and range) (mg/Kg)
#1489b
1ab
2.50
0.63
0.73
0.57
0.54
1.65
0.83ꢁ0.07
0.21ꢁ0.06
0.25ꢁ0.04
0.19ꢁ0.05
0.19ꢁ0.05
0.64ꢁ0.13
83ꢁ7 (71e108)
21ꢁ6 (7e45)
5/5
3/5
0/5
0/5
1/5
0/5
1bb
25ꢁ4 (15e38)
19ꢁ5 (11e38)
19ꢁ5 (4e30)
1cc
1db
1ec
64ꢁ13 (32e107)
a
Rats presented with 1% w/w toxicant in a highly palatable bait formulation (5 g) following 1 day ‘pre-feeding’ with a highly palatable bait formulation free of toxicant.
b
c
n¼5 (female rats).
n¼5 (3 male and 2 female rats).
Over 2 h. Data represented as mean ꢁ SE.
d
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
11
4.1.1.
a
-Phenyl-
a
-[6-phenyl-6-(2-pyridyl)-2-fulvenyl]-2-
tetrahydrofuran (25 mL). Purification by column chromatography
(hexane/ethyl acetate, 1:2) afforded 1b (white solid; 1.0 g,
1.95 mmol, 70%) as a mixture of eight stereoisomers. Rf (hexane/
ethyl acetate, 1:2): 0.35, 0.26, 0.20, 0.15; mp 131e143 ꢀC; 1H NMR
pyridinemethanol (4).4 As used to determine the effect of temper-
ature, concentration, reaction solvent and Lewis acid on the Diel-
seAlder reaction of racemic 2-fulvenylmethanol 4 and maleimide
(6). To a freshly prepared solution of sodium ethoxide (26 mmol;
from 0.60 g sodium) in ethanol (30 mL) at 0 ꢀC was added 2-
benzoylpyridine (2) (9.15 g, 52 mmol), followed by freshly dis-
tilled cyclopentadiene (3) (1.72 g, 26 mmol), added dropwise over
60 min, and the mixture stirred at 10 ꢀC for a further 6 h. The so-
lution was then concentrated in vacuo with purification by column
chromatography (petroleum ether/ethyl acetate, 4:1) affording 4
(orange solid; 8.89 g, 21 mmol, 82%) as a racemic mixture of four
stereoisomers. Rf (petroleum ether/ethyl acetate, 1:1): 0.56; mp
157e162 ꢀC (Lit.4 160e165 ꢀC); nmax(neat) cmꢃ1 3376, 3079, 1589,
(400 MHz, CDCl3): d 2.89e3.07 (0.35H, m, S/H-2, T/H-2, X/H-2, R/H-
2, X/H-3, S/H-3), 3.26e3.28 (0.05H, m, T/H-3), 3.35e3.56 (0.96H, m,
V/H-2, U/H-2, Y/H-3, U/H-3, W/H-3, R/H-3, S/H-4), 3.62e3.69
(0.88H, m, Y/H-2, W/H-2, V/H-3, W/H-4), 3.73e3.78 (0.20H, m, R/H-
1, X/H-1, T/H-4), 3.83e3.88 (0.28H, m, V/H-1, U/H-1), 3.94e3.98
(0.35H, m, Y/H-4, X/H-4), 4.11e4.16 (0.16H, m, U/H-4, R/H-4),
4.31e4.33 (0.20H, m, V/H-4), 4.35e4.38 (0.09H, m, T/H-1, S/H-1),
4.44e4.47 (0.48H, m, Y/H-1, W/H-1), 5.62e5.66 (0.76H, m, V/H-6,
Y/H-6, U/OH, W/OH), 5.74e5.76 (0.13H, m, R/H-6, T/H-6), 5.80
(0.04H, s, S/OH), 5.83 (0.28H, s, Y/OH), 5.86 (0.07H, s, X/OH), 5.88
(0.20H, s, V/OH), 6.05e6.08 (0.28H, m, U/H-6, W/H-6), 6.15 (0.07H,
dd, J¼3.4, 1.1 Hz, X/H-6), 6.19 (0.05H, T/OH), 6.20 (0.04H, dd, J¼3.4,
1.1 Hz, S/H-6), 6.54 (0.08H, s, R/OH), 6.77e7.57 (16H, m, Ar),
1568, 1431, 1371, 1150, 1035, 952, 755; 1H NMR (CDCl3):
d 5.99e6.00
(0.5H, m, trans H-1), 6.02e6.04 (0.5H, m, cis H-1), 6.16 (1H, br s,
OH), 6.33e6.36 (0.5H, m, cis H-4), 6.38e6.41 (0.5H, m, trans H-4),
6.55e6.58 (1H, m, H-3), 7.14e7.67 (16H, m, Ar), 8.52 (1H, d,
7.67e7.95 (1H, m, NH), 8.39e8.63 (2H, m, Pyr).
a
J¼4.8 Hz,
a
Pyr), 8.59 (0.5 H, d, J¼4.9 Hz, cis
aPyr), 8.68 (0.5H, d,
J¼4.9 Hz, trans
aPyr).
4.1.5. 5-( -Hydroxy- -2-pyridylbenzyl)-7-(a-2-pyridylbenzylidene)-
a
a
5-norbornene-2,3-dicarboximide (1c).4 Compound 1c was prepared
according to method B using racemic 2-fulvenylmethanol 4 (1.04 g,
2.51 mmol) and maleimide (6) (244 mg, 2.51 mmol) in methanol
(20 mL). Purification by column chromatography (hexane/ethyl ace-
tate,1:2) afforded 1c (white solid; 1.11 g, 2.17 mmol, 87%) as a mixture
of eight stereoisomers. Rf (hexane/ethyl acetate,1:2): 0.35, 0.26, 0.20,
4.1.2. -Phenyl-a-[6-phenyl-6-(2-pyridyl)-2-fulvenyl]-2-
a
pyridinemethanol bis-N-oxide (13).42 Prepared as a racemic mixture
for use as an HPLC reference in the asymmetric synthesis of com-
pound 13. To a solution of 1226 (136 mg, 0.68 mmol) in ethanol/
tetrahydrofuran (0.7 mL, 1:1 v/v) at 10 ꢀC was added dropwise
a solution of cyclopentadienyl sodium in tetrahydrofuran (0.17 mL,
0.34 mmol, 2 M) over 30 min, and the mixture stirred at 10 ꢀC for
a further 2 h. The solution was then concentrated in vacuo, the
resulting residue dissolved in methanol, filtered through a short
plug of silica and the filtrate concentrated in vacuo. Purification by
column chromatography (dichloromethane/acetone, 3:1) afforded
13 (orange solid; 109 mg, 0.20 mmol, 59%) as a racemic mixture of
four stereoisomers. Rf (dichloromethane/acetone, 3:1): 0.13; mp
180 ꢀC (dec) (Lit.42 223e224 ꢀC (EtOH/H2O)); nmax(neat) cmꢃ1 3055,
0.15; mp 131e145 ꢀC; 1H NMR (400 MHz, CDCl3):
d 2.90e3.00 (0.47H,
m, S/H-2, T/H-2, X/H-2, R/H-2), 3.04e3.10 (0.21H, m, X/H-3, S/H-3),
3.27e3.29 (0.10H, m, T/H-3), 3.36e3.60 (0.85H, m, V/H-2, U/H-2, Y/H-
3, U/H-3, W/H-3, R/H-3, S/H-4), 3.63e3.71 (0.58H, m, Y/H-2, W/H-2,
V/H-3, W/H-4), 3.74e3.75 (0.16H, m, R/H-1), 3.77e3.80 (0.23H, m, X/
H-1, T/H-4), 3.84e3.88 (0.22H, m, V/H-1, U/H-1), 3.94e4.00 (0.31H,
m, Y/H-4, X/H-4), 4.12e4.17 (0.24H, m, U/H-4, R/H-4), 4.32e4.39
(0.32H, m, T/H-1, S/H-1, V/H-4), 4.44e4.48 (0.31H, m, Y/H-1, W/H-1),
5.63e5.67 (0.53H, m, V/H-6, Y/H-6, U/OH, W/OH), 5.75e5.77 (0.26H,
m, R/H-6, T/H-6), 5.80 (0.08H, s, S/OH), 5.83 (0.18H, s, Y/OH),
5.86e5.88(0.27H, m, X/OH, V/OH), 6.06e6.09(0.21H, m, U/H-6, W/H-
6), 6.16 (0.13H, dd, J¼3.4,1.1 Hz, X/H-6), 6.19e6.21 (0.18H, m, S/H-6, T/
OH), 6.53 (0.16H, s, R/OH), 6.78e7.57 (16H, m, Ar), 7.77e7.85 (1H, m,
2924,1600,1482,1424,1253,1154, 764, 732; 1H NMR (CDCl3):
d 5.63
(0.5H, s, H-1), 6.05e6.07 (1H, m, H-1, H-4), 6.49e6.53 (1H, m, H-3,
H-4), 6.59 (0.5H, dd, J¼5.4, 1.3 Hz, H-3), 7.17e7.45 (16H, m, Ar),
8.12e8.22 (2H, m,
aPyr), 8.45 (1H, 2ꢄs, OH).
NH), 8.41e8.64 (2H, m,
aPyr).
4.1.3. 5-( -Hydroxy-a-2-pyridylbenzyl)-7-(a-2-pyridylbenzylidene)-
a
5-norbornene-2,3-dicarboximide (1a).4 Compound 1a was prepared
according to method B using racemic 2-fulvenylmethanol 4 (1.07 g,
2.58 mmol) and maleimide (6) (251 mg, 2.58 mmol) in 1,2-
dichloroethane (20 mL). Purification by column chromatography
(hexane/ethyl acetate, 1:2) afforded 1a (white solid; 1.15 g,
2.24 mmol, 87%) as a mixture of eight stereoisomers. Rf (hexane/
ethyl acetate, 1:2): 0.35, 0.26, 0.20, 0.15; mp 137e146 ꢀC; 1H NMR
4.1.6. 5-( -Hydroxy- -2-pyridylbenzyl)-7-(a-2-pyridylbenzylidene)-
a
a
5-norbornene-2,3-dicarboximide (1d).4 Compound 1d was pre-
pared according to method B using racemic 2-fulvenylmethanol 4
(1.04 g, 2.51 mmol) and maleimide (6) (244 mg, 2.51 mmol) in
dimethylsulfoxide (20 mL). Purification by column chromatography
(hexane/ethyl acetate, 1:2) afforded 1d (white solid; 880 mg,
1.72 mmol, 69%) as a mixture of eight stereoisomers. Rf (hexane/
ethyl acetate, 1:2): 0.35, 0.26, 0.20, 0.15; mp 134e145 ꢀC; 1H NMR
(400 MHz, CDCl3): d 2.89e2.99 (0.18H, m, S/H-2, T/H-2, X/H-2, R/H-
2), 3.04e3.09 (0.09H, m, X/H-3, S/H-3), 3.26e3.28 (0.03H, m, T/H-
3), 3.35e3.58 (1.02H, m, V/H-2, U/H-2, Y/H-3, U/H-3, W/H-3, R/H-
3, S/H-4), 3.62e3.69 (0.86H, m, Y/H-2, W/H-2, V/H-3, W/H-4),
3.74e3.78 (0.15H, m, R/H-1, X/H-1, T/H-4), 3.84e3.88 (0.39H, m, V/
H-1, U/H-1), 3.95e3.98 (0.34H, m, Y/H-4, X/H-4), 4.11e4.16 (0.17H,
m, U/H-4, R/H-4), 4.32e4.36 (0.34H, m, T/H-1, S/H-1, V/H-4),
4.44e4.46 (0.43H, m, Y/H-1, W/H-1), 5.63e5.66 (0.82H, m, V/H-6,
Y/H-6, U/OH, W/OH), 5.75e5.77 (0.09H, m, R/H-6, T/H-6),
5.81e5.87 (0.65H, m, Y/OH, V/OH, X/OH, S/OH), 6.06e6.08 (0.26H,
m, U/H-6, W/H-6), 6.15e6.21 (0.12H, m, X/H-6, S/H-6, T/OH), 6.53
(0.06H, s, R/OH), 6.78e7.58 (16H, m, Ar), 7.89e8.13 (1H, m, NH),
(400 MHz, CDCl3): d 2.90e3.00 (0.66H, m, S/H-2, T/H-2, X/H-2, R/H-
2), 3.03e3.10 (0.30H, m, X/H-3, S/H-3), 3.27e3.29 (0.17H, m, T/H-3),
3.35e3.59 (0.72H, m, V/H-2, U/H-2, Y/H-3, U/H-3, W/H-3, R/H-3, S/
H-4), 3.62e3.69 (0.39H, m, Y/H-2, W/H-2, V/H-3, W/H-4),
3.72e3.78 (0.52H, m, R/H-1, X/H-1, T/H-4), 3.84e3.88 (0.13H, m, V/
H-1, U/H-1), 3.95e3.98 (0.27H, m, Y/H-4, X/H-4), 4.11e4.16 (0.24H,
m, U/H-4, R/H-4), 4.33e4.34 (0.08H, m, V/H-4), 4.37e4.39 (0.31H,
m, T/H-1, S/H-1), 4.46e4.47 (0.21H, m, Y/H-1, W/H-1), 5.62e5.67
(0.34H, m, V/H-6, Y/H-6, U/OH, W/OH), 5.74e5.77 (0.36H, m, R/H-6,
T/H-6), 5.81 (0.14H, s, S/OH), 5.84 (0.11H, s, Y/OH), 5.86 (0.16H, s, X/
OH), 5.89 (0.08H, s, V/OH), 6.06e6.08 (0.15H, m, U/H-6, W/H-6),
6.16 (0.16H, dd, J¼3.4, 1.1 Hz, X/H-6), 6.20e6.21 (0.31H, m, S/H-6, T/
OH), 6.54 (0.19H, s, R/OH), 6.78e7.63 (16H, m, Ar), 7.75e7.88 (1H, m,
8.42e8.61 (2H, m,
a
Pyr).
4.1.4. 5-( -Hydroxy-
a
a
-2-pyridylbenzyl)-7-(
a
-2-pyridylbenzylidene)-
NH), 8.43e8.63 (2H, m, aPyr).
5-norbornene-2,3-dicarboximide (1b).4 Compound 1b was prepared
according to method B using racemic 2-fulvenylmethanol 4 (1.14 g,
2.75 mmol) and maleimide (6) (267 mg, 2.75 mmol) in
4.1.7. 5-(
a-Hydroxy-a-2-pyridylbenzyl)-7-(a-2-pyridylbenzylidene)-
5-norbornene-2,3-dicarboximide (1e).4 Compound 1e was prepared
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014

12
M. Jay-Smith et al. / Tetrahedron xxx (2016) 1e12
19. Rai, V.; Mobin, S. M.; Namboothiri, I. N. N. Tetrahedron: Asymmetry 2007, 18,
2719.
20. Yang, H.-M.; Li, L.; Li, F.; Jiang, K.-Z.; Shang, J.-Y.; Lai, G.-Q.; Xu, L.-W. Org. Lett.
according to method C using racemic 2-fulvenylmethanol 4 (1.14 g,
2.76 mmol), maleimide (6) (268 mg, 2.76 mmol) and indium(III)
chloride (610 mg, 2.76 mmol) in dichloromethane (20 mL). Purifi-
cation by column chromatography (hexane/ethyl acetate, 1:2)
afforded 1e (white solid; 0.93 g, 1.82 mmol, 66%) as a mixture of
two stereoisomers. Rf (hexane/ethyl acetate, 1:1): 0.07, 0.11; mp
2011, 13, 6508.
21. Porter, N. A.; Feng, H.; Kavrakova, I. K. Tetrahedron Lett. 1999, 40, 6713.
22. Johar, P. S.; Araki, S.; Butsugen, Y. J. Chem. Soc., Perkin Trans. 1, 1992, 711.
23. Scettri, A.; Massa, A.; Palombi, L.; Villano, R.; Rosaria, M.; Acocella, M. R. Tet-
rahedron: Asymmetry 2008, 19, 2149.
146e156 ꢀC; 1H NMR (400 MHz, CDCl3):
d
3.37 (0.6H, dd, J¼7.9,
24. Perdicchia, D.; Jørgensen, K. A. J. Org. Chem. 2007, 72, 3565.
25. Sibi, M.; Shay, J. J.; Ji, J. Tetrahedron Lett. 1997, 38, 5955.
4.5 Hz, W/H-3), 3.51 (0.4H, dd, J¼7.9, 4.9 Hz, U/H-2), 3.57 (0.4H, dd,
J¼7.9, 4.5 Hz, U/H-3), 3.64 (0.6H, ddd, J¼4.5, 1.7, 1.4 Hz, W/H-4), 3.67
(0.6H, dd, J¼7.9, 5.0 Hz, W/H-2), 3.85 (0.4H, ddd, J¼4.9, 3.4, 1.6 Hz,
U/H-1), 4.16 (0.4H, ddd, J¼4.5, 1.6, 1.3 Hz, U/H-4), 4.47 (0.6H, ddd,
J¼5.0, 3.4, 1.7 Hz, W/H-1), 5.64 (1H, s, U/OH, W/OH), 6.06e6.08 (1H,
m, U/H-6, W/H-6), 6.83e7.59 (16H, m, Ar), 7.85 (0.4H, s, NH), 7.88
~
26. Holmquist, M.; Blay, G.; Munoz, M. C.; Pedro, J. R. Org. Lett. 2014, 16, 1204.
~
27. Barroso, S.; Blay, G.; Munoz, M. C.; Pedro, J. R. Org. Lett. 2011, 13, 402.
€
28. Landa, A.; Minkkila, A.; Blay, G.; Jørgensen, K. A. Chem.dEur. J. 2006, 12, 3472.
€
29. Landa, A.; Richter, B.; Johansen, R. L.; Minkkila, A.; Jørgensen, K. A. J. Org. Chem.
2007, 72, 240.
30. Livieri, A.; Boiocchi, M.; Desimoni, G.; Faita, G. Chem.dEur. J. 2012, 18, 11662.
31. Sauer, J.; Sustmann, R. Angew. Chem., Int. Ed. Engl. 1980, 19, 779.
32. Coda, C. A.; Desimoni, G.; Ferrari, E.; Righetti, P. P.; Tacconi, G. Tetrahedron Lett.
1984, 40, 1611.
33. Cativiela, C.; Garcia, J. I.; Mayoral, J. A.; Salvatella, L. Chem. Soc. Rev. 1996, 209.
34. Cativiela, C.; Garcia, J. I.; Mayoral, J. A.; Avenoza, A.; Peregrina, J. M.; Roy, M. A. J.
Phys. Org. Chem. 1991, 4, 48.
(0.6H, s, NH), 8.41e8.42 (0.4H, m, U/
a
Pyr), 8.44e8.46 (0.4H, m, U/
Pyr).
a
Pyr), 8.49e8.51 (0.6H, m, W/ Pyr), 8.61e8.63 (0.6H, m, W/a
a
5. Experimental (bait palatability and efficacy trial in Spra-
gueeDawley rats)
35. Cativiela, C.; Garcia, J. I.; Mayoral, J. A.; Royo, A. J.; Salvatella, L. Tetrahedron:
Asymmetry 1993, 4, 1613.
36. Cativiela, C.; Garcia, J. I.; Mayoral, J. A.; Salvatella, L. Can. J. Chem. 1994, 72, 308.
37. Berson, J. A.; Hamlet, Z.; Mueller, W. A. J. Am. Chem. Soc. 1962, 84, 297.
38. Cativiela, C.; Garcia, J. I.; Gil, J.; Martinez, R. M.; Mayoral, J. A.; Salvatella, L.;
Urieta, J. S.; Mainar, A. M.; Abraham, M. H. J. Soc. Perkin Trans. 2 1997, 653.
39. Cativiela, C.; Garcia, J. I.; Mayoral, J. A.; Royo, L.; Salvatella, L.; Assfeld, X.; Ruiz-
Lopez, M. F. J. Phys. Org. Chem. 1992, 5, 230.
Please see Supplementary data.
Acknowledgements
40. Yates, P.; Eaton, P. J. Am. Chem. Soc. 1960, 82, 4436.
41. Ali, M. B. J. Am. Chem. Soc. 1984, 106, 3806.
42. Mohrbacher, R. J.; Paragamian, V.; Carson, E. L.; Puma, B. M.; Rasmussen, C. R.;
Meschino, J. A.; Poos, G. I. J. Org. Chem. 1966, 31, 2149.
43. Brown, J. E.; De Witt, E. G.; Shapiro, H. US Patent 2,864,843, 1958.
44. Pearson, R. G. J. Am. Chem. Soc. 1963, 85, 3533.
The authors would like to acknowledge the Ministry of Business,
Innovation and Employment, New Zealand and Connovation Ltd,
New Zealand for their financial support and assistance (M.J.-S.).
Thanks also go to Jenn Bothwell (Lincoln University, New Zealand)
for her invaluable help during the bait trials.
45. Zhang, Y. Inorg. Chem. 1982, 21, 3889.
46. Fischer, T.; Sethi, A.; Welton, T.; Woolf, J. Tetrahedron Lett. 1999, 40, 793.
47. Earle, M. J.; McCormac, P. B.; Seddon, K. R. Green Chem. 1999, 23.
48. De la Hoz, A.; Diaz-Ortis, A.; Moreno, A.; Langa, F. Eur. J. Org. Chem. 2000, 3659.
49. Majetich, G.; Hicks, R. J. Microw. Power EE 1995, 30, 27.
50. Gedye, R. N.; Rank, W.; Westaway, K. C. Can. J. Chem. 1991, 69, 706.
51. Kohler, E. P.; Kable, J. J. Am. Chem. Soc. 1935, 57, 917.
52. Woodward, R. B.; Baer, H. J. Am. Chem. Soc. 1944, 66, 645.
53. Hong, B.-C.; Shr, Y.-J.; Liao, J.-H. Org. Lett. 2002, 4, 663.
54. Barluenga, J.; Tomas, M.; Suarez-Sobrino, A.; Lopez, L. A. J. Chem. Soc., Chem.
Commun. 1995, 1785.
Supplementary data
Supplementary data (Raw HPLC data (corresponding to Tables
7e9) and bait-palatability and efficacy trial experimental pro-
tocols) associated with this article can be found in the online ver-
sion, at doi http://dx.doi.org/10.1016/j.tet.2016.07.014.
55. Javed, T.; Mason, T. J.; Phull, S. S.; Baker, N. R.; Robertson, A. Ultrason. Sonno-
chem. 1995, 2, S3.
References and notes
56. Berlan, J.; Mason, T. J. Ultrasonics 1992, 30, 203.
57. Kaur, N.; Kishore, D. J. Pharm. Chem. Res. 2012, 4, 991.
1. Roszkowski, A. P.; Poos, G. I.; Mohrbacher, R. J. Science 1964, 144, 412.
2. Roszkowski, A. P. J. Pharmacol. Exp. Ther. 1965, 149, 288.
3. Bova, S.; Cima, L.; Golovina, V.; Luciani, S.; Cargnelli, G. Cardiovasc. Drug Rev.
2001, 19, 226.
4. Brimble, M. A.; Muir, V. J.; Hopkins, B.; Bova, S. Arkivoc 2004, 1.
5. Poos, G. I.; Mohrbacher, R. J.; Carson, E. L.; Paragamian, V.; Puma, B. M.; Ras-
mussen, C. R.; Roszkowski, A. P. J. Med. Chem. 1966, 9, 537.
6. Kusano, T. J. Fac. Agri. Tottori Univ. 1975, 5, 15.
58. Grieco, P. A.; Nunes, J. J.; Gaul, M. D. J. Am. Chem. Soc. 1990, 112, 4595.
59. Forman, M. A.; Dailey, W. P. J. Am. Chem. Soc. 1991, 113, 2761.
60. Kobayashi, S. Lanthanide triflate catalyzed carbon-carbon bond-forming re-
actions in organic synthesis In Lanthanides: Chemistry and Use in Organic Syn-
thesis; Kobayashi, S., Ed.; Springer: Heidelberg, 1999.
61. Kobayashi, S.; Ogawa, C. Chem.dEur. J. 2006, 12, 5954.
62. Song, C. E.; Shim, W. H.; Roh, E. J.; Lee, S.; Choi, J. H. Chem. Commun. 2001, 1122.
63. Unless stated otherwise dielectric constants were obtained from: Vogel, A. I. In
Vogel’s Textbook of Practical Organic Chemistry, 5th ed.; Furniss, B. S., Hannaford,
A. J., Smith, P. W. G., Tatchell, A. R., Eds.; Longman: Harlow, 1989.
64. Unless stated otherwise normalized polarity values were obtained from:
Reichardt, C.; Welton, T. Empirical Parameters of Solvent Polarity In Solvents
and Solvent Effects in Organic Chemistry; Wiley-VCH GmbH & Co. KGaA: 2010.
65. Handbook of Organic Solvent Properties; Smallwood, I. M., Ed.; John Wiley &
Sons: New York, 1996.
7. Shimizu, T. Appl. Ent. Zool. 1983, 18, 243.
8. Greaves, J. H. J. Hyg. 1966, 64, 275.
9. Ogushi, K.; Iwao, T. Eisei Dobutsu (in Japanese) 1970, 21, 181.
10. Rennison, B. D.; Hammond, L. E.; Jones, G. L. J. Hyg. 1968, 66, 147.
11. Greaves, J. H.; Rowe, F. P.; Redfern, R.; Ayres, P. Nature 1968, 219, 402.
12. Nadian, A.; Lindblom, L. Int. J. Pharm. 2002, 242, 63.
13. Rennison, D.; Laita, O.; Bova, S.; Cavalli, M.; Hopkins, B.; Linthicum, D. S.;
Brimble, M. A. Bioorg. Med. Chem. 2012, 20, 3997.
14. Rennison, D.; Laita, O.; Conole, D.; Jay-Smith, M.; Knauf, J.; Bova, S.; Cavalli, M.;
Hopkins, B.; Linthicum, D.; Brimble, M. A. Bioorg. Med. Chem. 2013, 21, 5886.
15. Choi, H.; Conole, D.; Atkinson, D. J.; Laita, O.; Jay-Smith, M.; Pagano, M. A.;
Ribaudo, G.; Cavalli, M.; Bova, S.; Hopkins, B.; Brimble, M. A.; Rennison, D. Chem.
Biodiv 2016, 13, 762.
66. Sastry, N. V.; Dave, P. N. Proc. Indian Acad. Sci. Chem. Sci. 1997, 109, 211.
67. Gu, X.; Song, X.; Shao, C.; Zeng, P.; Lu, X.; Shen, X.; Yang, Q. Int. J. Electrochem.
Sci. 2014, 9, 8045.
68. Mukherjee, L. M.; Grunwald, E. J. Phys. Chem. 1958, 62, 1311.
€
69. Weingartner, H. Z. Phys. Chem. 2006, 220, 1395.
16. Mohrbacher, R. J.; Almond, H. R., Jr.; Carson, E. L.; Rosenau, J. D.; Poos, G. I. J. Org.
70. Earle, M.; Wasserscheid, P.; Schulz, P.; Olivier-Bourbigou, H.; Favre, F.; Vaultier,
M.; Kirschning, A.; Singh, V.; Riisager, A.; Fehrmann, R.; Kuhlmann, S. Organic
Synthesis In Ionic Liquids in Synthesis; Wiley-VCH GmbH & Co. KGaA: 2008.
Chem. 1966, 31, 2141.
17. Ojida, A.; Yamano, T.; Taya, N.; Tasaka, A. Org. Lett. 2002, 4, 3051.
18. Smaardijk, A. A.; Wynberg, H. J. Org. Chem. 1987, 52, 135.
Please cite this article in press as: Jay-Smith, M.; et al., Tetrahedron (2016), http://dx.doi.org/10.1016/j.tet.2016.07.014