Novel chiral 1-phosphono-1,3-butadiene for asymmetric hetero Diels-Alder cycloadditions with nitroso and azodicarboxylate dienophiles

Article abstract of DOI:10.1016/j.tetlet.2009.12.063

Chiral 1-phosphonodienes bearing a bicyclic (R,R)-1,3,2-dioxaphospholane or a (R,R)-1,3,2-diazaphospholidine auxiliary are potent dienes for asymmetric hetero Diels-Alder reactions. Their reactivity towards model nitroso and azodicarboxylate dienophiles has been studied by means of theoretical chemistry at the B3LYP/6-31G** level. This model, taking solvent effects into account, allowed us to identify parameters governing the stereoselectivity of this reaction. Our study emphasizes a synergy effect when increasing the steric hindrance of substituents of both partners. This led us to predict high levels of diastereoselectivity for one diene. Accordingly, we have hereby illustrated a convenient and original synthesis of this diene and its cycloadditions with commercially available nitroso and azodicarboxylate dienophiles.

Full text of DOI:10.1016/j.tetlet.2009.12.063

Tetrahedron Letters 51 (2010) 1052–1055
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Novel chiral 1-phosphono-1,3-butadiene for asymmetric hetero Diels–Alder
cycloadditions with nitroso and azodicarboxylate dienophiles
Jean-Christophe Monbaliu a, Bernard Tinant b, Daniel Peeters b, Jacqueline Marchand-Brynaert a,
*
a Unité de Chimie organique et médicinale, Département de Chimie, Université catholique de Louvain, Bâtiment Lavoisier, Place Louis Pasteur 1, B-1348 Louvain-la-Neuve, Belgium
b Unité de Chimie structurale et des mécanismes réactionnels, Département de Chimie, Université catholique de Louvain, Bâtiment Lavoisier, Place Louis Pasteur 1, B-1348
Louvain-la-Neuve, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
Chiral 1-phosphonodienes bearing a bicyclic (R,R)-1,3,2-dioxaphospholane or a (R,R)-1,3,2-diazaphosp-
holidine auxiliary are potent dienes for asymmetric hetero Diels–Alder reactions. Their reactivity towards
model nitroso and azodicarboxylate dienophiles has been studied by means of theoretical chemistry at
Received 28 October 2009
Revised 11 December 2009
Accepted 14 December 2009
Available online 16 December 2009
**
the B3LYP/6-31G level. This model, taking solvent effects into account, allowed us to identify parame-
ters governing the stereoselectivity of this reaction. Our study emphasizes a synergy effect when increas-
ing the steric hindrance of substituents of both partners. This led us to predict high levels of
diastereoselectivity for one diene. Accordingly, we have hereby illustrated a convenient and original syn-
thesis of this diene and its cycloadditions with commercially available nitroso and azodicarboxylate
dienophiles.
Keywords:
Chiral phosphonodienes
Nitroso dienophiles
Azodicarboxylate dienophiles
Hetero Diels–Alder reaction
DFT calculation
Ó 2009 Elsevier Ltd. All rights reserved.
Aminophosphonic acids and related compounds are recognized
as an important class of pharmacologically active molecules.1 Most
Two different strategies could be envisaged to build a chiral
phosphonodiene: (i) the direct introduction of the chirality on
the phosphorus atom8 (e.g., diene 1g, Fig. 2); (ii) the introduction
of the chirality on an adjacent atom, with a nitrogen relay nearby
the phosphorus atom (e.g., diene 1f, Fig. 2). The first way was aban-
doned because the synthesis of diene 1g led to an inseparable mix-
ture of diastereoisomers (see Supplementary data (SD)). For the
of the described synthetic strategies concern
a-aminophosphonic
compounds, despite the fact that the interest for b-,
c
- and d-
homologs is still growing. As a part of a research program2 dedi-
cated to the synthesis of aminophosphonic compounds, we have
studied the hetero Diels–Alder (HDA) cycloadditions of achiral 1-
phophonodienes with activated nitrogen-containing dienophiles,
namely nitroso and azodicarboxylate compounds.3 Indeed, the
resulting [4+2] cycloadducts can be considered as synthons for
the preparation of various aminophosphonic compounds (Fig. 1).
In addition to its versatility, the HDA reaction is also compatible
with asymmetric development.4,5 This led us to consider asymmet-
ric HDA cycloadditions of chiral phosphonodienes and nitrogen-
containing dienophiles as an entry towards aminophosphonic chir-
ons. Scarce examples of chiral phosphonodienophiles are illus-
trated by Wyatt,6,7 Katagiri8 and King9 with chirality either
directly surrounding the phosphorus atom itself or being placed
elsewhere on the chiral auxiliary. To our knowledge, only one
previous report by Wyatt mentioned the use of a chiral 4-phe-
nyl-1-phosphonodiene versus a cyclic azodicarboxylate dienophile
(4-phenyl-1,2,4-triazolin-3,5-dione) leading to the corresponding
cycloadduct in good yield but with low stereoselectivity.6
HO
HO
PO(OEt)2
OH
PO(OEt)2
PO(OEt)2
steps
O
N
O
+
N
R2
R2
2b
R2HN
1a
Figure 1. General HDA strategy. See Ref. 3b.
(R)
*
*
N
P
O
P
(R)
*
*
N
O
O
N
1f
1g
* Corresponding author.
E-mail address: jacqueline.marchand@uclouvain.be (J. Marchand-Brynaert).
Figure 2. Two examples of chiral phosphonodienes: dienes 1f–g.
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.12.063

J.-C. Monbaliu et al. / Tetrahedron Letters 51 (2010) 1052–1055
1053
(R)
(R)
R1
R1
O
R1
R1
O
X
P
X
P
X
X
P
(R)
(R)
Y
N
O
X
O
X
P
X
R1
X
R1
R1
R2
rot
R1
1
Ea
(R)
(S)
2
3
Y
N
Y
N
2a (Y=O, R2=Me)
2b (Y=O, R2=2-Tolyl)
2c (Y=NR2, R2=CO2Me)
R2
R2
4
Psyn-1
Panti-1
(R)-3 (Y=O)
(R)-4 (Y=NR2)
(S)-3 (Y=O)
1b (XR1=O), 1c (XR1=NH), 1d (XR1=NMe)
(S)-4 (Y=NR2)
1e (XR1=NiPr) and 1f (XR1=NBenzyl)
Figure 3. Asymmetric HDA cycloadditions of chiral phosphonodienes (1b–f) and nitroso (2a–b) or azo (2c) dienophiles: partners selected for the theoretical study.
second way, we used C2-symmetry auxiliaries, namely bicyclic
1,3,2-dioxaphospholane10 and 1,3,2-diazaphospholidines11 derived
from (R,R)-1,2-dihydroxy- and (R,R)-1,2-diaminocyclohexane,
respectively.
tential (
electrophilicity (
heterodienophiles are located above the chiral phosphonodienes in
the electrophilicity scale ( = 2.27 eV (2a), = 2.82 eV (2b) and
= 2.91 eV (2c)). Thus, in such HDA reactions they will act as
formal electrophiles while the phosphonobutadienes will act as
formal nucleophiles. The differences between dienes and dieno-
philes are between 0.6 and 1.6 eV, suggesting a Polar Diels–Alder
reaction (P-DA).13 The global electrophilicity of the dienes depends
on the nature of the chiral auxiliary. Diene 1b (X = O) is signifi-
l
) and chemical hardness (
g) as well as the global
x
) are listed in the SD. We found that all selected
x
x
Five different chiral dienes (1b–f) have been selected for a the-
oretical study of their HDA reactions with nitroso (2a–b) and azo-
dicarboxylate (2c) dienophiles (Fig. 3).The conformational
behaviour of the dienes and their respective cycloadditions are dis-
cussed at the B3LYP/6-31G** level, including condensed phase cal-
culations.12 This led to a realistic picture of the incidence of both
the XR1 substituent and dienophile nature on the diastereoselectiv-
ity. These theoretical insights have been then applied for the de-
sign of a new chiral phosphonodiene 1f. We propose herein an
original and convenient synthesis of (3aR,7aR)-2-(1-buta-1,3-die-
nyl)-[3a,4,5,6,7,7a-octahydro-1,3-dibenzyl-1,3,2-benzodiazaphosp-
hole]-2-oxide (1f) and the validation of this reagent in asymmetric
synthesis using nitroso (2b) and azodicarboxylate (2d–f) dieno-
philes.
A local potential energy surface (PES) scan allowed localizing
two stationary points for each diene studied, susceptible to under-
go a HDA cycloaddition. Both conformers are characterized by a
s-cis conformation of the butadienyl moiety, but show differences
in the O@P–C(1)–C(2) dihedral angles: Panti and Psyn conformers
feature the P@O bond, respectively, in anti and syn conformations
regarding the butadienyl moiety (rotation around the P–C(1) bond,
Fig. 3). The Psyn conformer is the most stable, whatever the diene
studied. The nature of XR1 has a slight incidence on this energetic
discrimination. Nevertheless, the diene 1f showed the highest level
x
cantly higher in the electrophilicity scale (
x
= 1.65 eV) than dienes
= 1.31 eV (1e)
1c–f (X = NR1):
x
= 1.41 eV (1c), = 1.38 eV (1d), x
x
and
x = 1.41 eV (1f). But in this series of dienes, the incidence of
the N-substitution (R1) appeared to be quite low.
The transition states (TSs) of the cycloaddition reactions of the
selected dienes with some model dienophiles were examined. As
the transition states related to the Panti (TSan) and Psyn (TSsy)
conformers led, respectively, to the (R)-P–C(1) and (S)-P–C(1) ster-
eoisomers (see Fig. 3 and SD), the subjacent conformational equi-
librium may be considered as a key element for stereochemical
drifts (i.e., HDA reactions leading to low diastereoselectivities).
The energetic differences between TSan and TSsy will thus give
information in terms of stereoselectivities (D
Esael, Table 2). The
TSs associated to the proximal (nitroso) and endo approaches will
be the only considered ones for discussion, since the distal (nitroso)
and exo pathways are disfavoured by 4–10 kcal molÀ1 and 3–
6 kcal molÀ1, respectively. These TSs are characterized by a highly
of conformational discrimination (DEconf, Table 1) both in gas and
condensed phase [tetrahydrofuran (THF), dichloroethane (DCE)
and acetonitrile (ACN)].
Table 2
Selectivities computed for the HDA reactions of dienes 1b–f and dienophiles 2a–c
Concerning the activation barrier for this conformational equi-
librium (Eraot, Fig. 3 and Table 1), the nature of XR1 has a stronger
incidence. An activation barrier of 11.7 kcal molÀ1 was found for
the diene 1f.
For both conformers of dienes 1c–f (X = N), the substituents
attached to the nitrogen atoms were found in a pseudo-equatorial
position. In agreement with the previous observations of Bennani
and Hanessian,11 there is a clear differentiation of the preferred
orientation of R1 groups on either size of the chiral auxiliary.
Entry
Dienes
Media
Dienophiles
A
2a
B
2b
C
2c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1b
Gas
1.7
0.7
0.6
0.3
1.8
0.6
0.4
0.4
2.3
1.3
1.2
0.9
2.9
2.3
2.2
2.0
3.1
3.7
3.6
3.4
1.7
0.8
0.7
0.6
0.9
0.6
0.5
0.6
1.8
0.6
0.4
0.4
2.0
1.8
1.7
1.6
2.8
3.1
3.0
2.7
0.7
0.8
0.7
0.7
0.5
1.3
1.1
1.2
2.6
1.4
1.3
1.4
3.2
1.4
1.4
1.4
5.0
7.5
6.0
4.6
XR1 = O
THF
DCE
ACN
Gas
THF
DCE
ACN
Gas
THF
DCE
ACN
Gas
THF
DCE
ACN
Gas
1c
X = N, R1 = H
We further considered the global electrophilicity (x) of the se-
lected reagents.13 The static global properties, namely chemical po-
1d
X = N, R1 = Me
Table 1
Energetic discrimination between the Panti and Psyn conformers
1e
X = N, R1 = iPr
D
Econf ( kcal molÀ1
)
Eraot
Gas
Gas
THF
DCE
ACN
1f
1b
1c
1d
1e
1f
2.1
3.8
4.0
4.9
11.7
0.8
0.9
0.9
1.4
1.6
0.5
0.6
0.9
1.2
1.2
0.5
0.6
0.9
1.2
1.2
0.4
0.5
0.8
0.8
1.4
X = N, R1 = Bn
THF
DCE
ACN
D
Esael values are given in kcal molÀ1
.

1054
J.-C. Monbaliu et al. / Tetrahedron Letters 51 (2010) 1052–1055
Table 3
asynchronous bond formation, the C(4)–N bond being almost com-
pletely formed (1.87–2.06 Å) and the C(1)–Y bond remaining close
to the sum of the corresponding van der Waals radii (2.66–2.88 Å).
No significant differences are observed between TSsy and TSan in
terms of bond formation (pictures of TSs for the reactions of 1f
with 2a–c are provided in SD).
Experimental conditions of HDA cycloadditions (Scheme 1)
Entry React.
T
Solv. Time (h) Conv. (%) dr (31P) Product
1
2
3
4
5
6
7
1f+2b 90 °C DMF 12
99
99
99
/
35
27
30
1:1
1:1
1:1
/
1:0.6
1:0
1:0
3b
3b
3b
/
1f+2b
D
DCE
12
1
5 d
1
3
1f+2b MWa
1f+2d
D
DCE
For each case in Table 2, the TSsy was lower in energy than the
corresponding TSan. Generally, the introduction of a solvent re-
duces the difference between the two diastereogenic TSs, except
for the cycloaddition 1c+2c14 and for diene 1f. These results
showed a slight incidence of the solvent polarity. Comparing the
different dienes, it appeared that the nature of the heteroatom
(O, N) has a weak incidence on the selectivities (entries 1A–C
and 5A–C). For diene 1b, cycloadditions with both nitrosomethane
(2a) and 2-nitrosotoluene (2b) proceed with a selectivity of
1.7 kcal molÀ1 (entries 1A–B). The diastereoselectivity drops to
0.7 kcal molÀ1 with dienophile 2c (entry 1C). Considering the
diazaphospholidino dienes (1c–f), the selectivity obtained with
2a is higher than that for 2b (entries 5A–B, 9A–B, 13A–B and
17A–B). By increasing the steric hindrance of the diene R1 groups,
a slight increase in selectivity is observed: for the dienophile 2a the
selectivity increases twice when going from diene 1c to diene 1f
(entries 5A, 9A, 13A and 17A), while for dienophile 2b the increase
in selectivity is more pronounced passing from diene 1c to diene 1f
(entries 5B, 9B, 13B and 17B). The cycloadditions of dienophile 2c
with dienes 1b–f clearly showed a synergy effect due to the steric
hindrance of both the R1 and R2 substituents (entries 5C, 9C, 13C
and 17C). The highest level of stereoselectivity was obtained with
diene 1f.
The same trends were observed in solvent (entries 18C, 19C and
20C). It may be also expected that working with sterically more
hindered azodicarboxylates 2d–f (R2 = Et, iPr, tBu) will increase
the diastereoselectivity of the reaction. This prediction has been
experimentally confirmed.
We have synthesized the chiral diene 1f in two steps from 1a in
62% overall yield (Scheme 1 and SD for experimental details). Its
structure was confirmed by X-ray diffraction analysis (see SD).
Under classical thermal conditions, the reaction of diene 1f with
2-nitrosotoluene (2b) led quantitatively to the corresponding cyl-
coadduct 3b, after 12 h in refluxing DCE (Scheme 1, see SD for
experimental details).
1f+2d MWb
4d (R2 = Et)
4e (R2 = iPr)
4f (R2 = tBu)
1f+2e MWb
1f+2f
MWb
5
a
In DCE, 100 °C, 500 W.
In toluene/DMF 30:1, 120 °C, 750 W.
b
We next considered the cycloadditions of diene 1f and azodi-
carboxylate dienophiles (R2 = Et, 2d; R2 = iPr, 2e and R2 = tBu, 2f).
In DCE under thermal conditions, no reaction was observed, even
under prolonged heating (degradation of the chiral diene). Never-
theless, under MW heating, the desired cycloadditions occurred
(Table 3, entries 4–7) and the heterocycles 4d–f were isolated in
modest yields (Scheme 1 and SD for experimental details). This
experimental result is in agreement with the larger charge trans-
fer for the HDA of (1f+2b) than for (1f+2d–f), favoring the (1f+2b)
P-DA reaction.13,15 13C NMR spectra (benzene-d6) of cycloadducts
4d–f showed typical features of the 1-phosphono-3,6-dihydro-
1
1,2-hydrazine pattern (C(6)–P at 55.91 ppm with JCÀP = 128.4 Hz)
and 31P NMR data provided unambiguously the stereochemical
information. Raising the size of the azodicarboxylate substituents
led to an increase in the selectivity. Indeed, the 31P NMR spectrum
of the crude mixture 4d revealed two signals (36.3 and 35.2 ppm)
while the 31P NMR spectra of 4e–f contained only one signal (38.8
and 38.1 ppm, respectively). By comparison with the previous
observations in the 1,2-oxazine and 1,2-pyridazine series, both
13C and 31P NMR values are consistent with the formation of sin-
gle diastereoisomers, namely cycloadducts (S)-4e and (S)-4f (see
SD).
In conclusion, this work describes the first computational study
of chiral bicyclic 1,3,2-dioxaphospholane-(1b) 1,3,2-diazaphosp-
holidine (1c–f) dienes and their asymmetric HDA cycloadditions
with nitroso and azodicarboxylate dienophiles. Among the series
of dienes, diene 1f showed the highest computed stereoselectivi-
ties. However, experimentally, the reaction 1f+2b led to a 1:1 mix-
ture of separable diastereomers 3b. This can be attributed to (i) a
low energetic discrimination between the Panti and the Psyn con-
formers of chiral diene 1f; (ii) the high asynchronicity of bond for-
mation with the nitroso-partner. Our theoretical model showed a
synergy effect when increasing the steric hindrance of both the
XR1 and R2 substituents. The computed high level of stereoselectiv-
ity for the reaction 1f+2c prompted us to consider the reactions of
diene 1f with commercially available azodicarboxylates 2d–f. The
cycloadditions with i-propyl and t-butyl azodicarboxylates 2e
and 2f led to the formation of single diastereoisomers, cycload-
ducts (S)-4e and (S)-4f, respectively.
The selectivity was not influenced by the solvent polarity.
Thermal instability of diene 1f prevents GC determination of the
selectivity, but 31P NMR is helpful. Indeed, the two diastereoiso-
mers (R)-3b and (S)-3b clearly show different 31P NMR shifts
(33.2 and 34.6 ppm). In all cases (Table 3, entries 1–3), we recov-
ered a 1:1 diastereomeric mixture of 3b cycloadducts, as expected
from the low selectivity level computed. Microwave (MW) heating
was also investigated in the view of increasing the diastereoselec-
tivity, but without success. HPLC analytical separation of the dia-
stereoisomers could be realized using a CHIRACEL OD-H column
(see SD).
(R)
Bn
(R)
(R)
N
Bn
Bn
N
N
H
(R)
(R)
(R)
NH
b
P O
P O
N
N
PO(OEt)2
a
POCl2
Bn
R2
R2
Bn
Bn
2b or 2d-f
O
N
(S) N
N
1f
or
Table 3
(Bn=CH2Ph)
Tol
1a
major diastereoisomer (4d)
single diastereoisomer (4e-f)
3b
Scheme 1. Synthesis and cycloadditions of diene 1f. Reagents and conditions: (a) TMSBr, DCM, rt; (COCl)2, DMFcat, THF, 0 °C; (b) (R,R)-N,N-dibenzyl-1,2-diaminocyclohexane,
pyridine, THF, 0 °C to rt, 15 h (62%).

J.-C. Monbaliu et al. / Tetrahedron Letters 51 (2010) 1052–1055
1055
4. Fringuelli, F.; Taticchi, A. The Diels–Alder Reaction; John Wiley & Sons, Ltd, 2002.
5. Yamamoto, H.; Kawasaki, M. Bull. Chem. Soc. Jpn. 2007, 80, 595–607.
6. Wyatt, P. B.; Villalonga-Barber, C.; Motevalli, M. Tetrahedron Lett. 1999, 40,
149–152.
7. Cheng, E. W. C.; Mandalia, R. T.; Motevalli, M.; Mothia, B.; Patanwadia, Y.;
Wyatt, P. B. Tetrahedron 2006, 62, 12398–12407.
Acknowledgements
This work was supported by FRIA through a fellowship to J.C.M.,
and by F.R.S.-FNRS through its support in accessing computational
facilities. J.M.B. is a senior research associate of F.R.S.-FNRS.
8. Katagiri, N.; Yamamoto, M.; Iwaoka, T.; Kaneko, C. J. Chem. Soc. Commun. 1991,
1429–1430.
9. Ware, R. W.; King, S. B. J. Org. Chem. 2000, 65, 8725–8729.
10. Maffei, M.; Buono, G. Tetrahedron 2003, 59, 8821–8825.
11. Bennani, Y.; Hanessian, S. Tetrahedron 1996, 52, 13837–13866.
12. Frisch, M. J. et al Gas phase calculations were done using the GAUSSIAN 98 suite of
programs, Revision A.11; Gaussian: Pittsburgh, PA, 1998. The single point PCM
SCRF method is used to evaluate solvent effects.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tetlet.2009.12.063.
13. (a) Domingo, L. R.; Sàez, J. A. Org. Biomol. Chem. 2009, 7, 3576–3583; (b)
Domingo, L. R.; Picher, M. T.; Arroyo, P. Eur. J. Org. Chem. 2006, 2570–2680. and
references cited therein.
References and notes
14. For the cycloaddition 1c+2c, a strong hydrogen bond-like interaction appeared
between the N–H moiety of the chiral auxiliary and the C@O bond of the
dienophile.
15. Domingo, L. R.; Saéz, J. A.; Zaragozà, R. J.; Arnò, M. J. Org. Chem. 2008, 73, 8791–
8799.
1. Kukhar, V.; Hudson, H. Aminophosphonic and Aminophosphinic Acids: Chemistry
and Biological Activity; John Wiley & Sons, Ltd, 2000.
2. Monbaliu, J.-C.; Robiette, R.; Peeters, D.; Marchand-Brynaert, J. Tetrahedron Lett.
2009, 50, 1314–1317. and references cited therein.
3. (a) Monbaliu, J.-C.; Marchand-Brynaert, J. Synthesis 2009, 11, 1876–1880; (b)
Monbaliu, J.-C.; Marchand-Brynaert, J. Tetrahedron Lett. 2008, 49, 1839–1842.