Synthesis of tricyclic phosphonopyrrolidines via IMDAF: Experimental and theoretical investigation of the observed stereoselectivity

Article abstract of DOI:10.1021/jo801138s

(Chemical Equation Presented) During the synthesis of tricyclic phosphonopyrrolidines via intramolecular Diels-Alder reactions of 1-acylamino(furan-2-yl)methyl phosphonates, two isomers are formed in most cases. The presence of a short three-atom tether together with spectroscopic data, including difference NOE, revealed that the cycloaddition occurred exo, but the phosphonate substituent on the tether had an exo or endo orientation. This was confirmed via X-ray analysis. A thermodynamic preference for the product with the phosphonate function in the endo position was observed experimentally and was confirmed theoretically. Density functional theory methods and several high-level post Hartree-Fock procedures were used to rationalize the observed isomer ratio of the IMDAF-reactions. This was done for two different types of reagents: with the activating carbonyl group in the tether or as a substituent on the tether. For the first type of molecules there is a large steric hindrance of the bulky tether substituents that disfavors the exo-isomer. In the latter case, there was a very small energy difference between the transition states causing a mixture of epimers being formed.

Full text of DOI:10.1021/jo801138s

Synthesis of Tricyclic Phosphonopyrrolidines via IMDAF:
Experimental and Theoretical Investigation of the Observed
Stereoselectivity
Diederica D. Claeys,†,§,| Kristof Moonen,† Bart I. Roman,†,| Victor N. Nemykin,‡
Viktor V. Zhdankin,‡ Michel Waroquier,§ Veronique Van Speybroeck,*,§ and
Christian V. Stevens*,†
Research Group SynBioC, Department of Organic Chemistry, Faculty of Bioscience Engineering, Ghent
UniVersity, Coupure links 653, B-9000 Ghent, Belgium, Department of Chemistry and Biochemistry,
UniVersity of Minnesota, Duluth, Minnesota 55812, and Center for Molecular Modeling, Ghent UniVersity,
Proeftuinstraat 86, B-9000 Ghent, Belgium
chris.steVens@ugent.be; Veronique.Vanspeybroeck@ugent.be
ReceiVed May 28, 2008
During the synthesis of tricyclic phosphonopyrrolidines via intramolecular Diels-Alder reactions of
1-acylamino(furan-2-yl)methyl phosphonates, two isomers are formed in most cases. The presence of a
short three-atom tether together with spectroscopic data, including difference NOE, revealed that the
cycloaddition occurred exo, but the phosphonate substituent on the tether had an exo or endo orientation.
This was confirmed via X-ray analysis. A thermodynamic preference for the product with the phosphonate
function in the endo position was observed experimentally and was confirmed theoretically. Density
functional theory methods and several high-level post Hartree-Fock procedures were used to rationalize
the observed isomer ratio of the IMDAF-reactions. This was done for two different types of reagents:
with the activating carbonyl group in the tether or as a substituent on the tether. For the first type of
molecules there is a large steric hindrance of the bulky tether substituents that disfavors the exo-isomer.
In the latter case, there was a very small energy difference between the transition states causing a mixture
of epimers being formed.
Introduction
lecular Diels-Alder reaction of carefully designed furanyl-R-
aminophosphonates. Furans can undergo Diels-Alder reactions
as the 4π diene components despite their aromaticity and hence
expected decrease in reactivity,3 both inter- and intramolecu-
larly.4 Furthermore, the IMDAF reaction (intramolecular
Diels-Alder reaction of furan) is particularly attractive as two
or more rings can be constructed in a single step with high regio-
and stereocontrol, providing a convenient entry into polycyclic
targets including natural products5 like prostaglandins5h,i and
terpenoids.5j The 7-oxabicyclo[2.2.1]heptane skeleton, formed
here, is present in biologically active natural products, but
moreover, the 7-oxanorbornanes and their unsaturated deriva-
Since the discovery of the biological activity of aminoalkyl-
phosphonates, e.g., as enzyme inhibitors with antibacterial, plant
growth regulatory, and neuromodulatory activities,1 many
researchers have also focused their attention on conformationally
constrained azaheterocyclic phosphonates.2 This paper reveals
the synthesis of tricyclic 2-phosphonopyrrolidines via intramo-
* To whom correspondence should be addressed. (C.V.S.) Fax: +32-9-
2646243. Tel: +32-9-2645957. (V.V.S.) Fax: +32-9-2646697. Tel: +32-9-
2646558.
† Department of Organic Chemistry, Ghent University.
‡ University of Minnesota.
§ Center for Molecular Modeling, Ghent University.
| D.D.C. and B.I.R. are Ph.D. fellow of the Research Foundation - Flanders
(FWO - Vlaanderen).
(3) (a) Lipshutz, B. H. Chem. ReV. 1986, 86, 795–819. (b) Wenkert, E.;
Moeller, P. D. R.; Piettre, S. R. J. Am. Chem. Soc. 1988, 110, 7188–7194. (c)
Domingo, L. R.; Picher, M. T.; Aurell, M. J. J. Phys. Chem. A 1999, 103, 11425–
11430. (d) Lautens, M.; Fillion, E. J. Org. Chem. 1996, 61, 7994–7995.
(4) Kappe, C. O.; Murphree, S. S.; Padwa, A. Tetrahedron 1997, 53, 14179–
14233.
(1) Kafarski, P.; Lejczak, B. Phosphorus, Sulfur, Silicon Relat. Elem. 1991,
63, 193–215, and references therein.
(2) (a) Moonen, K.; Laureyn, I.; Stevens, C. V. Chem. ReV. 2004, 104, 6177–
6215. (b) Redmore, D. Chem. ReV. 1971, 71, 315–337. (c) Robiette, R.; Defacqz,
N.; Peeters, D.; Marchand-Brynaert, J. Curr. Org. Synth. 2005, 2, 453–477.
10.1021/jo801138s CCC: $40.75
Published on Web 09/12/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 7921–7927 7921

Claeys et al.
tives can undergo a variety of reactions making them quite useful
synthetic intermediates in the synthesis of natural products and
analogues.6
While absolute stereocontrol of substituents on the diene or
dienophile is often observed, this is not the case for tether
substituents, which regularly give mixtures of both isomers.7
A detailed spectroscopic elucidation of the products and their
stereochemistry was performed using DIFNOE-NMR and X-ray
analysis. In order to explain the origin of selectivity or duality
of both isomers, theoretical calculations were conducted. They
can reveal whether the process is thermodynamically or kineti-
cally preferred and elucidate the effect of the position of the
carbonyl group on the stereoselectivity. Previous studies have
shown that steric interactions and ring strain in the transition
states are the source of stereoselectivity in intramolecular
Diels-Alder reactions.8 Tether substitutions, ring constraints,
and (planar) functional groups are known to alter the confor-
mational distribution and to restrict the rotations.9,10 The possible
influence of these factors was first investigated on a compound
with the carbonyl group incorporated in the tether, as a 100%
selectivity was observed in that case, using DFT methods:
B3LYP,11 mPW1PW91,12 and BB1K.13 The same DFT meth-
ods, MP2,14 MP3,15 and CCSD16 procedures were used for these
compounds where both isomers were formed.
FIGURE 1. Hindered rotation after acylation of the amine.
previous research it was shown that, when aminoalkyl phos-
phonate 1 (Figure 1) was acylated at nitrogen, there was a
hindered rotation about the C1-N bond.17 It is, however, an
unresolved issue whether this rotational barrier would allow,
hinder, or proliferate the Diels-Alder reaction.
The aminoalkyl phosphonates were synthesized following our
previously reported protocol.18 Subsequent acylations yielded
the corresponding amides 5 in high purity (Scheme 1). With
pivaloyl chloride, however, complex mixtures were obtained
under different reaction conditions.
This acyl group induces both a steric and an electronic
advantage to obtain a reactive substrate for the IMDAF reaction.
When 1-(allylamino)-1-(furan-2-yl)methyl phosphonate 4 was
refluxed overnight in toluene, only minimal amounts, 16%, of
Diels-Alder adducts could be detected using the typical peaks
of C6HaHb at 1.4 and 1.7 ppm. Prolonged refluxing only resulted
in a breakdown of the starting material. However, when the
corresponding amides (5a-e) were refluxed in toluene, complete
conversion to the ring-closed products was observed. In all cases,
a mixture of isomers (6′ and 6′′) was formed, the identity of
which will be discussed further in this paper (Scheme 1).
Experiments with toluene and acetonitrile showed that the first
one was the solvent of choice in most cases: a faster formation
of product and smaller amounts of side products were observed,
probably due to the higher boiling point of toluene.
The order of reactivity was in accordance with the results
reported previously for similar substrates 7 without the phos-
phonate group19 (Scheme 1). With a chloroacetyl group, 20 h
of reflux was required, while the ring-closed product 6e was
already formed during the acylation at room temperature and
subsequent aqueous workup.
In our work, however, a remarkable influence of the phos-
phonate group could be observed. While amides 8a and 8c could
only be obtained in low yield19 because of the poor conversion
(Table 1), the corresponding phosphono amides 6 gave complete
conversion. However, column chromatography caused big losses
of the product due to the presence of a polar group such as the
phosphonate group. The positive influence of increased tether
substitution on IMDAF reactions is often attributed to the
geminal dialkyl substitution and explained by the Thorpe-Ingold
effect, the reactive rotamer, or the facilitated transition effect.4,20
The hindered rotation we already observed in the synthesis of
phosphono-ꢀ-lactams17 will be further investigated.
Experimental Results and Discussion
The suitability of 1-(allylamino)-1-(furan-2-yl)methyl phos-
phonates 5 for the IMDAF reaction was investigated. From
(5) (a) Zhang, H. J.; Padwa, A. Org. Lett. 2006, 8, 247–250. (b) Padwa, A.;
Ginn, J. D. J. Org. Chem. 2005, 70, 5197–5206. (c) Wang, Q.; Padwa, A. Org.
Lett. 2004, 6, 2189–2192. (d) Kouznetsov, V. V.; Zubkov, F. I.; Cruz, U. M.;
Voskressensky, L. G.; Mendez, L. Y. V.; Astudillo, L.; Stashenko, E. E. Lett.
Org. Chem. 2004, 1, 37–39. (e) Fokas, D.; Patterson, J. E.; Slobodkin, G.;
Baldino, C. M. Tetrahedron Lett. 2003, 44, 5137–5140. (f) Padwa, A.; Eidell,
C. K.; Lynch, S. M. Heterocycles 2002, 58, 227–242. (g) Padwa, A.; Dimitroff,
M.; Liu, B. Org. Lett. 2000, 2, 3233–3235. (h) Sprague, P. W.; Heikes, J. E.;
Gougoutas, J. Z.; Malley, M. F.; Harris, D. N.; Greenberg, R. J. Med. Chem.
1985, 28, 1580–1590. (i) Sneden, A. T. Synlett 1993, 5, 313–322. (j) Yu, S. Y.;
Beese, G.; Keay, B. A. J. Chem. Soc., Perkin Trans. 1 1992, 2729–2731.
(6) Vogel, P.; Cossy, J.; Plumet, J.; Arjona, O. Tetrahedron 1999, 55, 13521–
13624.
(7) (a) Namboothiri, I. N. N.; Ganesh, M.; Mobin, S. M.; Cojocaru, M. J.
Org. Chem. 2005, 70, 2235–2243. (b) Zubkov, F. I.; Boltukhina, E. V.; Turchin,
K. F.; Borisov, R. S.; Varlamov, A. V. Tetrahedron 2005, 61, 4099–4113. (c)
Varlamov, A. V.; Zubkov, F. I.; Boltukhina, E. V.; Sidorenko, N. V.; Borisov,
R. S. Tetrahedron Lett. 2003, 44, 3641–3643.
(8) (a) Weingarten, M. D.; Prein, M.; Price, A. T.; Snyder, J. P.; Padwa, A.
J. Org. Chem. 1997, 62, 2001–2010. (b) Turner, C. I.; Williamson, R. M.;
Paddon-Row, M. N.; Sherburn, M. S. J. Org. Chem. 2001, 66, 3963–3969. (c)
Tantillo, D. J.; Houk, K. N.; Jung, M. E. J. Org. Chem. 2001, 66, 1938–1940.
(9) Jung, M. E.; Gervay, J. J. Am. Chem. Soc. 1991, 113, 224–232.
(10) Bur, S. K.; Lynch, S. M.; Padwa, A. Org. Lett. 2002, 4, 473–476.
(11) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Stephens,
P. J.; Devlin, F. J.; Chabalowski, C. F.; Frish, M. J. J. Phys. Chem. 1994, 98,
11623–11627.
(12) Adamo, C.; Barone, V. J. Chem. Phys. 1998, 108, 664–675.
(13) (a) Zhao, Y.; Lynch, B. J.; Truhlar, D. G. J. Phys. Chem. A 2004, 108,
2715–2719. (b) BB95/6-31+G(d) IOp(3/76)0580004200).
(14) (a) Head-Gordon, M.; Pople, J. A.; Frisch, M. J. Chem. Phys. Lett. 1988,
153, 503–506. (b) Frisch, M. J.; Head-Gordon, M.; Pople, J. A. Chem. Phys.
Lett. 1990, 166, 275–280. (c) Frisch, M. J.; Head-Gordon, M.; Pople, J. A. Chem.
Phys. Lett. 1990, 166, 281–289. (d) Head-Gordon, M.; Head-Gordon, T. Chem.
Phys. Lett. 1994, 220, 122–128. (e) Saebo, S.; Almlof, J. Chem. Phys. Lett. 1989,
154, 83–89.
(15) (a) Pople, J. A.; Seeger, R.; Krishnan, R. Int. J. Quant. Chem. Symp.
1977, 11, 149–163. (b) Pople, J. A.; Binkley, J. S.; Seeger, R. Int. J. Quant.
Chem. Symp. 1976, 10, 1–19.
(16) (a) Pople, J. A.; Krishnan, R.; Schlegel, H. B.; Binkley, J. S. Int. J.
Quantum Chem. 1978, 14, 545–560. (b) Bartlett, R. J.; Purvis, G. D. J. Quant.
Chem. 1978, 14, 561–581. (c) Purvis, G. D.; Bartlett, R. J. J. Chem. Phys. 1982,
76, 1910–1918. (d) Scuseria, G. E.; Janssen, C. L.; Schaefer, H. F., III. J. Chem.
Phys. 1988, 89, 7382–7387. (e) Scuseria, G. E.; Schaefer, H. F., III. J. Chem.
Phys. 1989, 90, 3700–3703.
The structure of the tricyclic pyrrolidines was confirmed by
its 2D DQFCOSY, HSQC, and HMBC spectra (Supporting
Information) and compared with the data from compounds 8.
1
The H NMR spectrum is characterized by a large difference
(17) (a) Van Speybroeck, V.; Moonen, K.; Hemelsoet, K.; Stevens, C. V.;
Waroquier, M. J. Am. Chem. Soc. 2006, 128, 8468–8478. (b) Stevens, C. V.;
Vekemans, W.; Moonen, K.; Rammeloo, T. Tetrahedron Lett. 2003, 44, 1619–
1622. (c) Moonen, K.; Stevens, C. V. Synthesis 2005, 3603–3612.
(18) Van Meenen, E.; Moonen, K.; Acke, D.; Stevens, C. V. ArkiVoc 2006,
i, 31–45.
(19) Ghelfi, F.; Parsons, A. F.; Tommasini, D.; Mucci, A. Eur. J. Org. Chem.
2001, 1845–1852.
(20) (a) Bruice, T. C.; Pandit, U. K. J. Am. Chem. Soc. 1960, 82, 5858–
5865. (b) Parrill, A. L.; Dolata, D. P. THEOCHEM 1996, 370, 187–202.
7922 J. Org. Chem. Vol. 73, No. 20, 2008

Synthesis of Tricyclic Phosphonopyrrolidines
SCHEME 1
TABLE 1. Acylation and IMDAF Reaction of 1-(Allylamino)-1-(furan-2-yl)methyl Phosphonate 4
no.
R
yield of 5 (%)
conversion/yield of 6a (%)
isomer ratio 6′′/6′
time (h)
T (°C)
conversion/yield of 7 to 8b (%)
a
b
c
d
e
CH2Cl
(CH2)2CH2Cl
iPr
CHCl2
CCl3
99
88
88
96
99c
100/75
100/47
100/53
100/94
100/99
27/73
20/80
17/83
35/65
21/79
20
4.5
7
110
110
110
110
82
43/40
62/60
100/97
100/99
3
1c
a The reported yields are those of isolated products after column chromatography or crystallization. b 30-40 h of reflux in acetonitrile (results from
ref 19). c The cycloadduct formed at room temperature during acylation and subsequent workup. The obtained mixture of 5e and 6e was further refluxed
for 1 h to obtain complete conversion to 6e.
in chemical shift of both NCH2 protons, clearly indicating that
they are on opposite sides of a rigid ring system. Similar to
pyrrolidines 8, the phosphonopyrrolidines 6 were isolated as a
mixture of two isomers, indicated as 6′ and 6′′ for the major
and minor isomer, respectively. Ghelfi and co-workers showed,
based on a NOESY experiment, that the configuration of the
tricyclic skeleton in 8 was exo and that the two isomers were
amide rotamers.19 The description exo is used if the group
attached to a nonbridgehead atom is pointing toward the oxa-
bridge and endo if it is pointing away from it. In the case of
phosphonopyrrolidines 6, the isomers might originate from three
diverse structural properties: endo- or exo-annulation, amide
rotamers, or configuration of the C2-center. Most studies using
similar substrates show exclusive formation of the exo-adduct
FIGURE 2. Comparison of typical coupling constants from literature
when the reaction is performed under thermodynamic con-
trol.4,5,7,10,19,21 Endo-fused, kinetic products are normally formed
under high-pressure-mediated conditions.7a Nevertheless, it is
reasonable to suggest the appearance of an endo-fused isomer
next to the exo-fused isomer based on the results of Tromp and
co-workers.22 They found that more steric substituents on the
tether can favor endo-addition. Therefore, the bulky phosphonate
group may be able to alter the reaction selectivity, yielding a
mixture of both adducts. However, having a close look at the
1H NMR and DQFCOSY data of both isomers 6′ and 6′′, the
differences in the multiplicities and coupling constants are found
to be too little to be the result of an opposite configuration at
the C5-bridgehead. Comparison with typical coupling constants
from similar compounds revealed the presence of an exo-fused
skeleton in both isomers (Figure 2).
sources (refs 22 and 23) with those measured in both isomers of
pyrrolidine 6.
In a further attempt to reveal the identity of the two isomers,
they were separated by recrystallization in acetone or diethyl
ether, giving the pure major isomer. The filtrate contained a
mixture of both isomers from which the minor isomer could be
recovered using column chromatography. Both isomers were
stable at room temperature for at least 1 month, so no amide
rotation is expected to occur during NMR acquisition. Further-
more, no rotamers were observed in the NMR spectra for the
open precursors 5. Therefore, it was concluded that the observed
isomers were not rotamers but originated from an incomplete
stereocontrol of the IMDAF-reaction at C2.
In order to reveal the position of the phosphonate group,
DIFNOE experiments were performed on both isomers. Con-
siderable nuclear Overhauser effects were observed between
C5H and C6HR and C4HR, confirming the exo-fused skeleton of
both isomers (Figure 3). A clear difference between both isomers
was observed, however, when the nuclear Overhauser effect at
C2H was studied: the bulky phosphonate is exo-oriented in the
major isomer 6e′. These results were confirmed using X-ray
analysis (Figure 3).
(21) (a) Pedrosa, R.; Sayalero, S.; Vicente, M.; Casado, B. J. Org. Chem.
2005, 70, 7273–7278. (b) Padwa, A.; Ginn, J. D.; Bur, S. K.; Eidell, C. K.;
Lynch, S. M. J. Org. Chem. 2002, 67, 3412–3424. (c) Paulvannan, K.; Chen,
T.; Jacobs, J. W. Synlett 1999, 1609–1611.
(22) Tromp, R. A.; Brussee, J.; van der Gen, A. Org. Biomol. Chem. 2003,
1, 3592–3599.
(23) (a) Feringa, B. L.; Gelling, O. J.; Melters, L. Tetrahedron Lett. 1990,
31, 7201–7204. (b) Van Royen, L. A.; Mijngheer, R.; De Clercq, P. J.
Tetrahedron Lett. 1983, 24, 3145–3148.
J. Org. Chem. Vol. 73, No. 20, 2008 7923

Claeys et al.
SCHEME 2
SCHEME 3
Furthermore, the observed stereoisomer ratio may be the result
of thermodynamic control and may not be the isomer ratio
formed in the initial reaction mixture. Equilibration can occur
under thermal conditions via a consecutive retro-Diels-Alder,
Diels-Alder reaction. To investigate this kind of behavior, pure
samples of major 6e′ and minor 6e′′ were heated in toluene
(110 °C). No change at all occurred to the minor isomer 6e′′
over a 20 h period. The major isomer 6e’ on the other hand,
was slowly converted to the minor isomer 6e′′. After 1 h at
110 °C, only 2% conversion was observed. This did not reflect
at all the 21/79 ratio observed after 1 h at 110 °C starting from
the open precursor 5e. When heating was continued for 20 h,
95% conversion to 6e′′ was observed. The slow conversion of
the major isomer to the minor isomer suggests retrocycloaddition
of the less stable cycloadduct. This is in agreement with the
stereochemistry generally observed during IMDAF reactions:
when a single bulky substituent is present in the tether, the most
stable cycloadduct will be formed in such a way as to minimize
nonbonded interactions.7a
Solvent effects may have an influence on the endo/exo
selectivity.24 Therefore, an experiment was set up to test this
latter effect in the laboratory for the IMDAF reaction leading
to pyrrolidine 6e. Methanol, acetonitrile, and toluene were used
as solvents at 65 °C leading to an endo/exo ratio of 18, 26, and
33%, respectively, after 2.25 h. So, the solvent does not cause
an inversion of stereoselectivity, but it causes some change on
the endo/exo ratio; the endo ratio increases with a decrease of
the solvent polarity.
An additional experiment was performed using a substrate
without an amide substituent. N-Allyl aminoalkyl phosphonate
4 was allylated using an excess of allyl bromide in the presence
of NaI. After refluxing for 4 h in acetone, a mixture of
diallylamine 9 and the ring-closed product 10 was obtained
(Scheme 2). This mixture was then refluxed in toluene during
4 h, yielding the ring-closed product 10. Notwithstanding the
presence of the chiral CHP-center, only one isomer could be
detected by NMR. Therefore, the relative stereochemistry at the
CHP-center was apparently fixed during the ring-closure reaction
or equilibration via retrocycloaddition was fast in this case.
This was also observed when the amide group was included
in the tricyclic skeleton. In this case, the dienophiles were
inserted during alkylation using cinnamoyl and acryloyl chloride.
The corresponding pyrrolidinones 13 were obtained as single
isomers upon refluxing the amides 12 in toluene or THF
(Scheme 3). However, the stereochemistry of the two stereo-
centers could not be determined via DIFNOE experiments.
Refluxing of 13a in toluene during 60 h did not result in
formation of the other epimer. In this case the acylation of the
sterically hindered tert-butylamine derivative succeeded, but
only with a very low degree of conversion and was followed
by ring closure in the same reaction step.
To conclude the experimental part learns that for compounds
6 with the acyl group as a substituent on the tether, two isomers
were formed, which are epimers at the C2-center. The major
isomer, with the phosphonate in exo-position, is the thermody-
(24) (a) Kong, S.; Evanseck, J. D. J. Am. Chem. Soc. 2000, 122, 10418–
10427. (b) Moreau, Y.; Loos, P.-F.; Assfeld, X. Theor. Chem. Acc. 2004, 112,
228–239. (c) Acevedo, O.; Jorgensen, W. L. J. Chem. Theory Comput. 2007, 3,
1412–1419.
FIGURE 3. Stereochemical analysis of isomers 6e′ and 6e′′. The
percentage of nuclear Overhauser effect with irradiation of C5H is
indicated. The X-ray structures of both isomers are depicted as well.
7924 J. Org. Chem. Vol. 73, No. 20, 2008

Synthesis of Tricyclic Phosphonopyrrolidines
SCHEME 4. Computationally Studied IMDAF Reactions
TABLE 2. Relativea Electronic (SCF without ZPE Corrections)
and Free Energies at 383.8 K (kJ/mol) for the Stationary Points of
the IMDAF Reaction of 12a to 13a with 6-31+G(d) Basis Set
namically less stable one. For the products 13 with the carbonyl
group being part of the tether, only one isomer was formed.
This isomer could not be converted into the other C2-epimer
by refluxing it in toluene.
B3LYP
∆E ∆G
0.0 0.0
BB1K
mPW1PW91
∆E ∆G
0.0 0.0
∆E
0.0
∆G
0.0
12a
Computational Details
TS endo
TS exo
13a endo
13a exo
111.3
125.6
-24.6
-8.9
118.5
133.8
-6.1
11.6
98.0
113.8
-72.9
-57.8
105.3
122.0
-54.4
-37.2
89.1
103.8
-67.3
-51.5
96.4
112.0
-48.8
-31.0
All ab initio calculations were carried out with the GAUSS-
IAN 03 software package.25 All geometries have been optimized
at the B3LYP11 level with the 6-31+G* basis set and were
followed by frequency calculations that confirmed the nature of
the stationary points and were used in the thermochemical
analysis. Diffuse functions were added as phosphorus is a third-
row element.26 Previous studies have shown B3LYP is very
accurate for many Diels-Alder reactions and performs well in
reproducing the effects on activation energies but shows
systematic errors for hetero systems and incorrectly predicts the
strain of the norbornene framework, leading to an underestima-
tion of the reaction exothermicity.27,28
As a correct energy calculation of the TSs was relevant
because of the small preference for one of the epimers, single-
point energies were calculated with new generation functionals,
i.e., mPW1PW9112 and BB1K13 as well as with post-HF MP214
and MP315 methods, and these numbers were validated against
CCSD.16
a Relative to the energy of the reactant.
1, 5, and 7 are linked), since they appear as enantiomeric
couples. The R-isomer was used for the calculations. For the
product only the exo-oriented tether was modeled, which was
based on the known stereochemical preference of the reaction.
From now on, exo and endo always point at the orientation of
the phosphonate group on the C2-position of the tether,
respectively, pointing toward the oxabridge or away from it.
i. Formation of Pyrrolidinone 13a. The relative energies
and free energies of the endo- and exo-cycloadducts 13a and
their respective TSs, with respect to the reactant, are given in
Table 2. The product with the phosphonate in the endo-position
is energetically more stable than the exo-analogue due to reduced
steric hindrance between the phosphonate group and the
oxanorbornene and a higher distance between the electronegative
oxygens of the phosphonatesin which the double-bonded O is
most electronegativesand the oxabridge of the oxanorbornene
skeleton. These distances are given in Table 3.
B3LYP is known to have problems with calculating the
energy of the oxanorbornene skeleton.11,27 In this case, the
relative free energies of the cycloadducts are too high and ∆G
of the exo-isomer is even positive. Nevertheless, the energy
difference between them can be used as a good indicator for
their relative energetic stability, as the strain is about the same
in both the endo- and the exo-isomer. This was confirmed by
the good correspondence with the single-point energy calcula-
tions using the mPW1PW9112 and BB1K13 functional.
The absence of a nuclear Overhauser effect between C5H and
C2H cannot completely exclude the exo-position of the phos-
phonate, mainly because this NOE effect was small for the exo-
isomer of pyrrolidine 6e. An experimental indication that the
endo-isomer is formed indeed can be deduced from the coupling
constant JHP of the proton at C2 of product 6e: for the exo-
isomer this is 14.0 Hz, for the endo-isomer 4.7 Hz. Thus, the
value of 5.5 Hz found for pyrrolidinone 13a also points in the
direction of the endo-isomer.
Theoretical Results and Discussion
As the tricyclic phosphonopyrrolidine 6e was most easily
formed, an X-ray was taken of both these isomers, and to
have a means for comparison, 6e was used for the calculation
as well. In addition, the acryloyl derivative 13a was used
because only one isomer was formed in excellent yield. The
reactions considered and the numbering of atoms are
presented in Scheme 4.
Until now, we have always mentioned the formation of one
or two isomers, but actually two or four different isomers are
present (keeping in mind that the stereochemistry at positions
(25) Frish, M. J. et al. Gaussian 03, Revision C.02; Gaussian, Inc.:
Wallingford; CT, 2004; see the Supporting Information for the full reference.
(26) (a) Leyssens, T.; Peeters, D. THEOCHEM 2004, 673, 79–86. (b)
Leyssens, T.; Peeters, D. THEOCHEM 2004, 686, 71–82.
(27) Guner, V.; Khuong, K. S.; Leach, A. G.; Lee, P. S.; Bartberger, M. D.;
Houk, K. N. J. Phys. Chem. A 2003, 107, 11445–11459.
(28) (a) Domingo, L. R.; Aurell, M. J.; Arno, M.; Saez, J. A. J. Org. Chem.
2007, 72, 4220–4227. (b) Pieniazek, S. N.; Houk, K. N. Angew. Chem. 2006,
118, 1470–1473. (c) Pieniazek, S. N.; Houk, K. N. Angew. Chem., Int. Ed. 2006,
45, 1442–1445. (d) Padwa, A.; Crawford, K. R.; Straub, C. S.; Pieniazek, S. N.;
Houk, K. N. J. Org. Chem. 2006, 71, 5432–5439. (e) Dinadayalane, T. C.; Vijaya,
R.; Smitha, A.; Sastry, G. N. J. Phys. Chem. A 2002, 106, 1627–1633. (f)
Domingo, L. R.; Picher, M. T.; Andre´s, J. J. Org. Chem. 2000, 65, 3473–3477.
J. Org. Chem. Vol. 73, No. 20, 2008 7925

Claeys et al.
TABLE 3. Geometric Parameters for the Stationary Points of the IMDAF Reaction of 12a to 13a of the B3LYP/6-31+G(d)-Optimized
Structures (Distances in Å, Angles in deg)
a
b
c
aO
aH
ꢁC1C2NC4
ꢁC2NC4C5
ꢁO10C1C2N
10P11
aO aO
10O12
O
BnO12
arO12
aH
10
Cl
12a
4.06
4.08
3.21
4.09
3.17
4.56
5.22
4.61
5.23
4.57
4.56
4.30
2.93
4.27
2.88
4.51
2.46
2.37
2.52
2.45
4.68
3.37
2.51
3.21
2.52
120.68
-12.49
18.04
-10.81
16.70
-168.43
-12.99
6.05
-6.09
1.22
-27.15
-71.79
70.33
-84.27
80.94
TS endo
TS exo
13a endo
13a exo
b
c
a Ocl is the oxygen of the methoxy groups closest to the oxabridge. HBn is the H-atom at the benzylic position closest to O12. Har is the aromatic
hydrogen atom closest to O12.
TABLE 4. Relativea Electronic (SCF without ZPE Corrections) and Free Energies at 354.8 K (kJ/mol) for the Stationary Points of the
IMDAF Reaction of 5e to 6e with 6-31+G(d) Basis Set
B3LYP
BB1K
mPW1PW91
∆E ∆G
MP2
MP3
CCSD
∆E
0.0
101.5
101.9
-26.0
-12.3
∆G
0.0
113.7
113.6
-1.8
9.5
∆E
0.0
87.7
85.4
-76.2
-63.8
∆G
0.0
100.0
97.2
-52.1
-42.0
∆E
∆G
∆E
∆G
∆E
∆G
5e
0.0
80.6
0.0
92.8
TS endo
TS exo
6e endo
6e exo
1.4
0.0
1.9
0.0
1.0
0.0
1.5
0.0
1.0
0.0
1.6
0.0
78.7
90.4
-67.7
-54.6
-43.6
-32.9
a Relative to the energy of the reactant for DFT methods and to the TS exo for MPx and CCSD.
TABLE 5. Geometric Parameters for the Stationary Points of the
IMDAF Reaction of 5e to 6e of the B3LYP/6-31+G(d)-Optimized
Structures (Distances in Å, Angles in deg)
a
ꢁC1C2NC4 ꢁC2NC4C5 ꢁO10C1C2N
10
10P11
10O12
O
aO
aO
aO
Cl
5e
TS endo 4.08
TS exo 3.06
6e endo 4.09
6e exo 3.02
3.14
3.50
3.07
4.54
2.96
4.58
2.90
-43.52
-32.26
-23.02
-25.18
-15.51
80.24
8.59
49.33
8.19
85.95
-61.61
92.05
-75.71
98.58
4.59
3.42
4.57
3.37
35.78
a Ocl is the oxygen of the methoxy groups closest to the oxabridge.
FIGURE 4. Structures of the TSs of the endo- and exo-isomer of 13a
at B3LYP/6-31+G(d). Distances in angstroms.
seen from the distances given in Table 3. The rotational potential
in terms of the dihedral angle ꢁC1C2NC4 was calculated at a
B3LYP/6-31+G(d) level of theory and is given as Figure S1
in the Supporting Information.
Both pathways were then explored to find an explanation for
the formation of a single isomer of pyrrolidinone 13a (Table
2). This IMDAF reaction has a large energy barrier toward the
TS, which is in agreement with the high temperature required
for the reaction. At 110 °C, the precursor rotamers will be in
equilibrium and the kinetic isomer ratio of the cycloadducts will
depend only on the difference in the free energy of the TS of
both isomers (Curtin-Hammett principle29). The TSs are
depicted in Figure 4, and relevant geometrical parameters for
TSs and precursors are given in Table 3. The free energy of
the TS for the exo-isomer is higher than for the endo-isomer
by 15-17 kJ/mol depending on the functional; causing the
exclusive formation of the endo-isomer.
To trace the origin of this energetic difference, we took a
closer look at the TSs and reactant. The major differences
between the TSs of the different isomers are the dihedral angles
ꢁO10C1C2N and ꢁC1C2NC4. The first one corresponds to the
rotation of the furan group and has an opposite value for both
isomers. The latter is associated with the hindered C2-N rotation
we described previously.17 This rotation has a barrier of about
70 kJ/mol at the HF/3-21+g* level and shows two minima at
-45° and at 115° that have about the same energy. The exo-
TS needs a larger distortion of the dihedral angle, and this might
be one of the origins for the higher energy of the TS. This
increase can be attributed to steric interactions, mainly of the
benzyl and furan group with the phosphonate moiety as can be
For formation of pyrrolidinone 13a we find exclusive
formation of the endo-isomer corresponding to the experimental
results of full selectivity, irrespective of the functional used.
ii. Formation of Pyrrolidines 6e. The energies and important
geometrical parameters of the stationary points in the cycload-
dition to the exo- and endo-isomer of 6e are given in Tables 4
and 5. Again, B3LYP underestimates the free energy of the
reaction, but the energy difference between both epimers is
independent of the functional. The endo-cycloadduct is energeti-
cally preferred in order to minimize the steric hindrance and
electrostatic repulsion between the phosphonate group and the
oxabridge of the oxanorbornene skeleton. These distances
between the oxygen atoms are listed in Table 5. The energy
difference between the endo- and exo-isomer of pyrrolidinone
13a is of the same order of magnitude as for the pyrrolidine
6e. If the retro-Diels-Alder reaction is possible, this difference
in stability between the two isomers can explain the conversion
of the major isomer into the minor isomer under thermodynamic
control.
Next, we wanted to unravel the experimentally observed
isomer ratio during the synthesis of pyrrolidine 6e. To test
whether the Curtin-Hammet29 conditions are fulfilled, the
reaction was followed in methanol. Methanol was chosen
because of its lower boiling point which does not allow the
retro-Diels-Alder reaction to occur, as observed experimentally.
Moreover, the 31P NMR peaks of both isomers are better
(29) Seeman, J. I. Chem. ReV. 1983, 2, 83–134.
7926 J. Org. Chem. Vol. 73, No. 20, 2008

Synthesis of Tricyclic Phosphonopyrrolidines
the allyl group as dienophile the endo- and exo-position require
about the same amide rotation and there is a subtle energy
difference that requests a high level of theory calculation. From
the calculations, it is clear that B3LYP fails to estimate the
reaction free energy of these strained cycloadducts.
Experimental Section
Typical Procedure for the Synthesis of r-Aminophosphonates
4 and 11a,b. See ref 18.
Typical Procedure for the Acylation of r-Aminophosphonates
(5 and 12a,b). To a solution of 5 mmol of 1-aminoalkyl phosphonate
in 10 mL of dry THF were added 10 mmol of pyridine and a
solution of the acid chloride of choice in 2 mL of dry THF under
a nitrogen atmosphere. The reaction mixture was stirred for 0.5-2
h at room temperature. It was washed with saturated NaHCO3(aq)
and then with a 1 M HCl(aq) solution. After drying with MgSO4,
filtration of the solids, and evaporation of the solvent under reduced
pressure, the product was obtained in high purity.
Synthesis of Tricyclic Phosphonopyrrolidines. Procedure for
the Preparation of Dimethyl (3-Allyl-10-oxa-3-azatricyclo-
[5.2.1.01,5]dec-8-en-2-yl)phosphonate (10). A mixture of R-amino-
phosphonate 4 (5 mmol), allyl bromide (15 mmol), sodium iodide
(15 mmol), and potassium carbonate (35 mmol) was refluxed in 5
mL of acetone. After 4 h, the mixture was cooled to room
temperature; the insoluble salts were filtered off, and the filtrate
was evaporated under reduced pressure. This resulted in a mixture
of diallylamine 9 (90%) and the ring-closed product. Refluxing in
toluene completed the IMDAF reaction. Product 10 was purified
using column chromatography (EtOAc, Rf ) 0.25) which caused
the yield to drop to 28%.
Procedure for the Preparation of Dimethyl (3-tert-Butyl-4-oxo-
6-phenyl-10-oxa-3-azatricyclo[5.2.1.01,5]dec-8-en-2-yl)phospho-
nate (13b). A solution of 5 mmol of 1-(tert-butylamino)-1-(furan-
2-yl)methyl phosphonate (11b) and 10 mmol of pyridine in 12 mL
of dry THF was stirred at room temperature under a nitrogen
atmosphere. A solution of 7.5 mmol of cinnamoyl chloride in 3
mL of dry THF was added. The mixture was then refluxed for 7 h.
It was washed with saturated NaHCO3(aq) solution and then with 1
M HCl(aq) and dried with MgSO4. After filtration of the solids and
evaporation of the solvent under reduced pressure, a brown oil was
obtained from which the ring-closed product could be obtained in
pure form using column chromatography.
Typical Procedure for the IMDAF Reaction (6 and 13a). A
solution of 3.5 mmol of a suitable 1-(alkylamino)-1-(furan-2-
yl)methyl phosphonate in 14 mL of toluene was refluxed until
complete disappearance of the starting material was obtained
(monitoring by 31P NMR). Then the solvent was removed under
reduced pressure, and the corresponding adducts were obtained in
good purity. Further purification could be performed using column
chromatography or crystallization from ether or acetone.
FIGURE 5. Structures of the TSs of the endo- and exo-isomer of 6e
at B3LYP/6-31+G(d). Distances in angstroms.
resolved. As the ratio of major and minor isomer remained
constant throughout the course of the reaction, the Curtin-
Hammet principle holds. So again the relative energies of the
TSs were studied. The structures and parameters are given in
Figure 5 and Tables 4 and 5. The energetic difference of both
TSs is very small and depends on the functional. Therefore, a
higher level of theory was used to calculate these subtle
differences. MP214 and MP315 calculations predict indeed a
lower free energy of the exo-TS by 1.9 to 1.5 kJ/mol, and these
figures were validated using a CCSD16 calculation. This
confirms the formation of a mixture of the endo- and exo-
cycloadduct.
The retro-Diels-Alder reaction of the exo-isomer has an
activation free energy that is, depending on the method, 33-42
kJ/mol more than the forward reaction and indicates that the
slow thermodynamic conversion of the exo- isomer to the endo-
epimer is possible.
The theoretical study for formation of pyrrolidine 6e indeed
shows that the endo-cycloadduct is thermodynamically preferred
and that there is a competition between the formation of both
isomers. CCSD calculations confirm the experimentally ob-
served preference for the epimer in which the phosphonate group
is oriented toward the oxa-bridge.
Conclusions
The IMDAF-reaction of 1-acyl-1-alkylamino(furan-2-yl)-
methyl phosphonates results in complex azaheterocyclic phos-
phonates in a small number of synthetic steps. These products
might be used as novel conformationally constrained amino acid
analogues. The presence of a phosphonate substituent on the
tether raises the conversion of the IMDAF reaction significantly.
Furthermore, a high degree of stereocontrol is observed during
the cycloaddition reaction. Only the exo-fused products were
obtained. For derivatives containing the carbonyl group in the
tether two isomers were formed, originating from incomplete
stereocontrol of the phosphite addition. If the allyl group serves
as dienophile, four isomers are formed resulting from an
incomplete kinetic control of the position of the phosphonate
as tether substituent during the IMDAF reaction. However, the
most stable stereoisomers, having an endo-oriented phosphonate
group, are formed under thermodynamic control.
Acknowledgment. We thank the Research Foundation -
Flanders (Fonds voor Wetenschappelijk Onderzoek - Vlaanderen)
and the research board of Ghent University for financial support
of this research.
Supporting Information Available: General experimental
methods, experimental procedures, spectroscopic data of all new
compounds with detailed peak assignments, crystallographic
information files of 6e′ and 6e′′, computational data, and full
reference. This material is available free of charge via the
Internet at http://pubs.acs.org.
This kinetic and thermodynamic control could be computa-
tionally reproduced. With the acyl group functioning as dieno-
phile the isomers with the phosphonate group in the exo-position
require a larger distortion from their minimum energy precursor
than the endo-isomers and are sterically more hindered, as can
be seen in their respective TSs. For aminophosphonates 5 with
JO801138S
J. Org. Chem. Vol. 73, No. 20, 2008 7927