The role of achiral pyrazolidinone templates in enantioselective Diels-Alder reactions: Scope, limitations, and conformational insights

Article abstract of DOI:10.1021/ja066425o

We have evaluated the role of achiral pyrazolidinone templates in conjunction with chiral Lewis acids in room temperature, enantioselective Diels-Alder cycloadditions. The role of the fluxional N(1) substituent was examined, with the bulky 1-naphthylmethyl group providing enantioselectivities up to 99% ee, while templates with smaller fluxional groups gave lower selectivities. High selectivities were also observed in reactions of 7d with chiral Lewis acids derived from relatively small chiral ligands, suggesting the pyrazolidinone templates are capable of relaying stereochemical information from the ligand to the reaction center. Lewis acids capable of adapting square planar geometries, such as Cu(OTf)₂, Cu(ClO₄)₂, and Pd(ClO₄)₂, were found to be particularly effective at providing high selectivities. Additionally, substitution at the C-5 position of the pyrazolidinone templates has been shown to be critical for optimal selectivity. Reactions of the optimal pyrazolidinone appended with a number of common dienophiles and various dienes demonstrate the utility of this achiral template. Furthermore, catalytic loadings could be lowered to 2.5 mol % with essentially no loss in selectivity, π-π interactions were evaluated as a means to explain the unusually high selectivity observed at room temperature. Finally, non-C₂-symmetric ligands were employed as a test to determine if chiral relay was operative.

Full text of DOI:10.1021/ja066425o

Published on Web 12/20/2006
The Role of Achiral Pyrazolidinone Templates in
Enantioselective Diels-Alder Reactions: Scope, Limitations,
and Conformational Insights
Mukund P. Sibi,* Levi M. Stanley, Xiaoping Nie, Lakshmanan Venkatraman,
Mei Liu, and Craig P. Jasperse
Contribution from the Department of Chemistry and Molecular Biology, North Dakota State
UniVersity, Fargo, North Dakota 58105
Received September 5, 2006; E-mail: mukund.sibi@ndsu.edu
Abstract: We have evaluated the role of achiral pyrazolidinone templates in conjunction with chiral Lewis
acids in room temperature, enantioselective Diels-Alder cycloadditions. The role of the fluxional N(1)
substituent was examined, with the bulky 1-naphthylmethyl group providing enantioselectivities up to 99%
ee, while templates with smaller fluxional groups gave lower selectivities. High selectivities were also
observed in reactions of 7d with chiral Lewis acids derived from relatively small chiral ligands, suggesting
the pyrazolidinone templates are capable of relaying stereochemical information from the ligand to the
reaction center. Lewis acids capable of adapting square planar geometries, such as Cu(OTf)2, Cu(ClO4)2,
and Pd(ClO4)2, were found to be particularly effective at providing high selectivities. Additionally, substitution
at the C-5 position of the pyrazolidinone templates has been shown to be critical for optimal selectivity.
Reactions of the optimal pyrazolidinone appended with a number of common dienophiles and various dienes
demonstrate the utility of this achiral template. Furthermore, catalytic loadings could be lowered to 2.5 mol
% with essentially no loss in selectivity. π-π interactions were evaluated as a means to explain the unusually
high selectivity observed at room temperature. Finally, non-C2-symmetric ligands were employed as a test
to determine if chiral relay was operative.
Introduction
tions. Can modified templates amplify the enantioselectivity
afforded by existing chiral Lewis acids? If so, amplification of
Chiral Lewis acid catalysis continues to be an effective way
to access enantiomerically pure compounds.1 A wealth of
research has been dedicated to the development of efficient
chiral Lewis acid catalysts.2 Much of this effort has focused on
the design of superior ligands capable of shielding distant
reactive centers when combined with an appropriate Lewis acid.3
Although this approach continues to provide unique solutions
in terms of asymmetric synthesis, the preparation of many of
the necessary ligands involves significant effort. An additional
limitation in many enantioselective reactions is the requirement
for low temperatures to achieve high levels of selectivity.
We have considered an approach involving improving the
design of achiral templates utilized for asymmetric transforma-
enantioselectivity might enable simpler, less elaborate, and more
accessible chiral ligands to induce high enantioselectivity for
reactions in which relatively inaccessible or expensive ligands
are currently required. The ability to amplify enantioselectivity
might also make room-temperature reactions practical.
Our approach focused on the use of pyrazolidinone templates
1 (Figure 1). Pyrazolidinones 1 are readily available and are
attractive because the R group can be easily varied and they
have no static (permanent) chirality. However, the fluxional N(1)
nitrogen does provide a chiral center. Following coordination
to a chiral Lewis acid, conformations 2 and 3 become diaste-
reotopic. We reasoned the reaction might proceed preferentially
through one of these two fluxional configurations, with the
chirality of the reactive configuration controlled by the chiral
Lewis acid. If so, the pyrazolidinone with induced chirality
might function much like a chiral auxiliary such as 4. Chiral
oxazolidinones 4 are effective in numerous diastereoselective
reactions4 and, when used in conjunction with matched chiral
Lewis acids, give strongly amplified stereoselectivity (double
(1) Thayer, A. M. Chem. Eng. News 2005, September 5, 40.
(2) (a) Jacobsen, E. N., Pfaltz, A., Yamamoto, H., Eds. ComprehensiVe
Asymmetric Catalysis I-III; Springer: Berlin, 1999. (b) Ojima, I., Ed.
Catalytic Asymmetric Catalysis in Organic Synthesis; Wiley-VCH: New
York, 2000. (c) Kobayashi, S., Jorgensen, K., A., Eds. Cycloaddition
Reaction in Organic Synthesis; Wiley-VCH: Weinheim, 2002. (d) For an
entire issue dedicated to enantioselective catalysis, see: Chem. ReV. 2003,
103, issue 8. (e) Johnson, J. S.; Evans, D. A. Acc. Chem. Res. 2000, 33,
325.
(3) For recent reviews, see: (a) Desimoni, G.; Faita, G.; Jorgensen, K. A. Chem.
ReV. 2006, 106, 3561. (b) Berthod, M.; Mignani, G.; Woodward, G.;
Lemaire, M. Chem. ReV. 2005, 105, 1801. (c) McManus, H. A.; Guiry,
P. J. Chem. ReV. 2004, 104, 4151. (d) Die´guez, M.; Pa`mies, O.; Claver, C.
Chem. ReV. 2004, 104, 3189. (e) Desimoni, G.; Faita, G.; Quadrelli, P.
Chem. ReV. 2003, 103, 3119. (f) Gennari, C.; Piarulli, U. Chem. ReV. 2003,
103, 3071. (g) Ghosh, A. K.; Mathivanan, P.; Cappiello, J. Tetrahedron:
Asymmetry 1998, 9, 1.
(4) For reviews, see: (a) Gnas, Y.; Glorius, F. Synthesis 2006, 1899. (b) Sibi,
M. P. Aldrichimica Acta 1999, 32, 93. (c) Ager, D. J.; Prakash, I.; Schaad,
D. R. Aldrichimica Acta 1997, 30, 3. (d) Ager, D. J.; Prakash, I.; Schaad,
D. R. Chem. ReV. 1996, 96, 835. (e) Gallou, I.; Senanayake, C. H. Chem.
ReV. 2006, 106, 2843. (f) Ghosh, A. K.; Fidanze, S.; Senanayake, C. H.
Synthesis 1998, 937. (g) Senanayake, C. H. Aldrichimica Acta 1998,
31, 3.
9
10.1021/ja066425o CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 395-405
395

A R T I C L E S
Sibi et al.
Scheme 1. Synthesis of Pyrazolidinone Dienophiles
Figure 1. Rationale for increased enantioselectivity with pyrazolidinone
substrates.
diastereoselection).5 We thought that if the pyrazolidinone R
group were to act in a consonant fashion with a chiral Lewis
acid, it could provide similar amplification of enantioselectivity.
However, unlike permanently chiral substrates such as 4, the
pyrazolidinone substrates need not be synthesized from chiral
pool precursors. Thus, we hoped to establish a “chiral relay”
situation in which the chiral Lewis acid would bias the fluxional
nitrogen, establish the pyrazolidinone as an effective chiral
auxiliary where the nitrogen substituent (the “relay group”) could
participate in face shielding, and if possible have the chiral
ligand with static chirality and the relay group work synergisti-
cally to amplify enantioselectivity.6
Scheme 2. Enantioselective Reactions Utilizing Pyrazolidinone
Substrates
Davies introduced a related concept for relaying stereochem-
ical information using a chiral auxiliary to install asymmetry.7
Similar approaches have also been evaluated in other labora-
tories, most notably by Clayden8 and Hitchcock.9 Our approach
differs in that chirality is not inherent to our substrate but is
indirectly relayed from a chiral Lewis acid through the template
to the reaction centers.10-13
(5) (a) Evans, D. A.; Miller, S. J.; Lectka, T. J. Am. Chem. Soc. 1993, 115,
6460. (b) Evans, D. A.; Barnes, D. M.; Johnson, J. S.; Lectka, T.; von
Matt, P.; Miller, S. J.; Murry, J. A.; Norcross, R. D.; Shaughnessy, E. A.;
Campos, K. R. J. Am. Chem. Soc. 1999, 121, 7582. (c) Evans, D. A.; Miller,
S. J.; Lectka, T.; von Matt, P. J. Am. Chem. Soc. 1999, 121, 7559.
(6) For reviews on the concept of stereochemical relay and related science,
see: (a) Corminboeuf, O.; Quaranta, L.; Renaud, P.; Liu, M.; Jasperse,
C. P.; Sibi, M. P. Chem. Eur. J. 2003, 9, 29. (b) Walsh, P. J.; Lurain, A.
E.; Balsells, J. Chem. ReV. 2003, 103, 3297. (c) Clayden, J.; Vassiliou, N.
Org. Biomol. Chem. 2006, 4, 2667. (d) Mikami, K.; Yamanaka, M. Chem.
ReV. 2003, 103, 3369.
(7) (a) Bull, S. D.; Davies, S. G.; Fox, D. J.; Garner, A. C.; Sellers, T. G. R.
Pure Appl. Chem. 1998, 70, 1501. (b) Bull, S. D.; Davies, S. G.; Epstein,
S. W.; Ouzman, J. V. A. Chem. Commun. 1998, 659. (c) Bull, S. D.; Davies,
S. G.; Fox, D. J.; Sellers, T. G. R. Tetrahedron: Asymmetry 1998, 9, 1483.
(d) Bull, Steven D.; Davies, Stephen G.; Epstein, Simon W.; Leech, Michael
A.; Ouzman, Jacqueline V. A. J. Chem. Soc., Perkin Trans. 1 1998, 2321.
(8) (a) Clayden, J.; Pink, J. H.; Yasin, S. A. Tetrahedron Lett. 1998, 39, 105.
(b) Betson, M. S.; Clayden, J.; Helliwell, M.; Johnson, P.; Lai, L. W.; Pink,
J. H.; Stimson, C. C.; Vassiliou, N.; Westlund, N.; Yasin, S. A.; Youssef,
L. H. Org. Biomol. Chem. 2006, 4, 424. (c) Clayden, J.; Westlund, N.;
Frampton, C. S.; Helliwell, M. Org. Biomol. Chem. 2006, 4, 455.
(9) For use of the relay concept in chiral auxiliary development see: (a) Casper,
D. M.; Blackburn, J. R.; Maroules, C. D.; Brady, T.; Esken, J. M.; Ferrence,
G. M.; Standard, J. M.; Hitchcock, S. R. J. Org. Chem. 2002, 67, 8871.
(b) Casper, D. M.; Burgeson, J. R.; Esken, J. M.; Ferrence, G. M.;
Hitchcock, S. R. Org. Lett. 2002, 4, 3739. (c) Hoover, T. R.; Hitchcock,
S. R. Tetrahedron: Asymmetry 2003, 14, 3233. (d) Hitchcock, S. R.; Casper,
D. M.; Vaughn, J. F.; Finefield, J. M.; Ferrence, G. M.; Esken, J. M.
J. Org. Chem. 2004, 69, 714. (e) Burgeson, J. R.; Renner, M. K.; Hardt, I.;
Ferrence, G. M.; Standard, J. M.; Hitchcock, S. R. J. Org. Chem. 2004,
69, 727. (f) Burgeson, J. R.; Dore, D. D.; Standard, J. M.; Hitchcock, S. R.
Tetrahedron 2005, 61, 10965.
(10) For examples that apply to the relay concept, see: (a) Quaranta, L.; Renaud,
P. Chimia 1999, 53, 364. (b) Quaranta, L. Ph.D. thesis, University of
Fribourg, Switzerland, 2000. (c) Quaranta, L.; Corminboeuf, O.; Renaud,
P. Org. Lett. 2002, 4, 39. (d) Corminboeuf, O.; Renaud, P. Org. Lett. 2002,
4, 1731.
(11) Balsells, J.; Walsh, P. J. J. Am. Chem. Soc. 2000, 122, 1802.
(12) For selective coordination of enantiotopic sulfone oxygen and relay of
stereochemistry, see: (a) Hiroi, K.; Ishii, M. Tetrahedron Lett. 2000, 41,
7071. (b) Wada, E.; Pei, W.; Kanemasa, S. Chem. Lett. 1994, 2345. (c)
Wada, E.; Yasuoka, H.; Kanemasa, S. Chem. Lett. 1994, 1637. (d) Evans,
D. A.; Campos, K. R.; Tedrow, J. S.; Michael, F. E.; Gagne´, M. R. J. Am.
Chem. Soc. 2000, 122, 7905. (e) Sugimoto, H.; Nakamura, S.; Hattori, M.;
Ozeki, S.; Shibata, N.; Toru, T. Tetrahedron Lett. 2005, 46, 8941. (f)
Sugimoto, H.; Nakamura, S.; Watanabe, Y.; Toru, T. Tetrahedron:
Asymmetry 2003, 14, 3043.
The pyrazolidinone templates of interest to this study were
prepared in a straightforward manner from the corresponding
â,â-disubstituted enoates (Scheme 1). Condensation of the
enoates with hydrazine hydrate gave the desired N(1) unsub-
stituted pyrazolidinones 5 in high yields. N(1) substitution was
accomplished by either reductive amination or selective alky-
lation of the more nucleophilic N(1) nitrogen. Finally, acylation
of 6 at N-2 occurs upon deprotonation with n-BuLi or LiHMDS
and subsequent addition of the appropriate R,â-unsaturated acid
chloride.
Prior reports from our laboratory have demonstrated the utility
of R,â-unsaturated pyrazolidinone imides in several asymmetric
transformations (Scheme 2). Pyrazolidinone substrates have
proven to be very effective for bisoxazoline/Lewis acid-
catalyzed Diels-Alder cycloadditions (eq 1),14 1,3-dipolar
cycloadditions of nitrones (eq 2),15 and nitrile oxides (eq 3) at
ambient temperature.16 Pyrazolidinone templates have also
(13) For the use of relay concept in chiral ligand design, see: (a) Sibi, M. P.;
Zhang, R.; Manyem, S. J. Am. Chem. Soc. 2003, 125, 9306. (b) Sibi,
M. P.; Stanley, L. M. Tetrahedron: Asymmetry 2004, 16, 3353. (c) Malkov,
A. V.; Stoncius, S.; MacDougall, K. N.; Mariani, A.; McGeoch, G. D.;
Kocovsky, P. Tetrahedron 2006, 62, 264. (d) Malkov, A. V.; Hand, J. B.;
Kocovsky, P. Chem. Commun. 2003, 1948. (e) Maughan, M. A. T.; Davies,
I. G.; Claridge, T. D. W.; Courtney, S.; Hay, P.; Davis, B. G. Angew. Chem.,
Int. Ed. 2003, 42, 3788.
9
396 J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007

Role of Achiral Pyrazolidinone Templates
A R T I C L E S
Table 1. Diels-Alder Reactions with Pyrazolidinone Templates Using Chiral Bisoxazolinesa
1
a For reaction details see Supporting Information. Endo/exo ratios were determined by H NMR, and ee determination was carried out after conversion
to the known benzyl ester using chiral HPLC. Absolute configuration was determined to be (1S,2S,3R,4R) by comparison of retention times of the known
benzyl ester. Yields for isolated column-purified material were 85-90%. b This experiment was carried out using Cu(SbF6)2 as the Lewis acid.
amplified enantioselectivity in the conjugate additions of
hydroxylamines17 and radicals.18
ated in Cu(OTf)2/bisoxazoline-catalyzed Diels-Alder reac-
tions.20 Cu(II)/bisoxazoline catalysts were chosen because of
the abundance of literature data for their catalysis of Diels-
Alder reactions involving oxazolidinones and related substrates.2e
Ready availability of various bisoxazoline ligands with differing
steric volumes allowed for the possibility to assess the impor-
tance of the chiral ligand size relative to the pyrazolidinone R
group in controlling enantioselection.
Diels-Alder reactions with substrates 7a-d were carried out
at room temperature using 15 mol % of the appropriate catalyst
and excess cyclopentadiene (Table 1).21 In each case the reaction
was also conducted using oxazolidinone substrate 15 for
comparison.
Initial experiments were conducted using Me-box ligand 9,
whose C-4 methyl substituents were expected to be too small
to afford high enantioselection without amplification from the
pyrazolidinone auxiliaries. Selectivity was low when oxazoli-
dinone 15 was used (38% ee, entry 1). However, when
pyrazolidinone auxiliaries were used, a systematic increase in
In this study, we report a detailed investigation on the use of
pyrazolidinone templates in enantioselective Diels-Alder reac-
tions.14 The importance of the R group on the pyrazolidinone
nitrogen in reactions catalyzed by Cu(II)/bisoxazoline complexes
of varying size was further examined. A variety of different
Lewis acids were screened, and the study suggests that interplay
of the chiral ligand and the pyrazolidinone R group is most
cooperative for metals with pseudo-square-planar geometries.
In addition, the effect of substitution at C-5 has also been
evaluated.19 Optimal pyrazolidinone templates and chiral Lewis
acids have been employed to evaluate the scope of substrates
and dienes amenable to effective cycloaddition. Some insights
are provided by X-ray structures of the pyrazolidinones. Finally,
mechanistic and conformational possibilities have been ad-
dressed by: (1) employing non-C2-symmetric ligands and (2)
screening for the possibility of π-π interactions between
aromatic relay groups and the olefin of the dienophile.
(19) For the effect of gem dialkyl substitution on conformational control, see:
(a) Onimura, K.; Kanemasa, S. Tetrahedron Lett. 1992, 48, 8631. (b) Bull,
S. D.; Davies, S. G.; Key, M.-S.; Nicholson, R. L.; Savory, E. D. Chem.
Commun. 2000, 1721.
Results and Discussion
Proof of Principle Experiments. The role of the relay group
(20) For comprehensive reviews on enantioselective Diels-Alder reactions,
see: (a) Corey, E. J. Angew. Chem., Int. Ed. 2002, 41, 1650. (b) Dias,
L. C. J. Braz. Chem. Soc. 1997, 8, 289. For selected recent examples of
enantioselective Diels-Alder reactions, see: (c) Palomo, C.; Oiarbide, M.;
Garcia, J. M.; Gonzalez, A.; Arceo, E. J. Am. Chem. Soc. 2003, 125, 13942.
(d) Ryu, D. H.; Lee, T. W.; Corey, E. J. J. Am. Chem. Soc. 2002, 124,
9992. (e) Futatsugi, K.; Yamamoto, H. Angew. Chem., Int. Ed. 2005, 44,
1484. (f) Thadani, A. N.; Stankovic, A. R.; Rawal, V. H. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 5846. (g) Hawkins, J. M.; Nambu, M.; Loren, S.
Org. Lett. 2003, 5, 4293.
substituents on the fluxional N(1) nitrogen was initially evalu-
(14) For a preliminary account of this work, see: (a) Sibi, M. P.; Venkatraman,
L.; Liu, M.; Jasperse, C. P. J. Am. Chem. Soc. 2001, 123, 8444. For the
use of pyrazolidinone templates in Diels-Alder reactions of dihydropy-
ridines, see: (b) Nakano, H.; Tsugawa, N.; Fujita, R. Tetrahedron Lett.
2005, 46, 5677.
(15) Sibi, M. P.; Ma, Z.; Jasperse, C. P. J. Am. Chem. Soc. 2004, 126, 718.
(16) Sibi, M. P.; Itoh, K.; Jasperse, C. P. J. Am. Chem. Soc. 2004, 126, 5366.
(17) Sibi, M. P.; Liu, M. Org. Lett. 2001, 3, 4181.
(21) For detailed reaction conditions and substrate preparation procedures, see
Supporting Information.
(18) Sibi, M. P.; Prabagaran, N. Synlett 2004, 2421.
9
J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007 397

A R T I C L E S
Sibi et al.
Table 2. Screening of Metal Geometries with i-Pr-bisoxazolinea
enantioselectivity correlated increasing size of the R group (Me,
64% ee < Ph, 71% ee < 2-naphthyl, 79% ee < 1-naphthyl,
86% ee, entries 2-5). These results suggested that even when
a weak ligand was used at room temperature, good enantiose-
lectivity was possible. The results also suggested that in this
case the pyrazolidinone R group played a major, perhaps
dominant, role in face shielding. The results with 7d represent
a 3-fold enhancement of enantioselectivity relative to those with
smaller pyrazolidinone 7a, (13.1:1 vs 4.6:1 er) and a 6-fold
amplification relative to those with oxazolidinone 15 (13.1:1
vs 2.2:1 er).
entry
Lewis acid
yield (%)b
endo/exoc
ee (%)d
1
2
3
4
5
6
7
8
Cu(OTf)2
Cu(ClO4)2
Pd(ClO4)2
Ni(ClO4)2
NiBr2
Mg(ClO4)2
Mg(NTf2)2
Mg(OTf)2
MgI2
Zn(OTf)2
Zn(ClO4)2
Zn(NTf2)2
ZnI2
Fe(ClO4)3
Fe(ClO4)2
Sm(OTf)3
Sc(OTf)3
Yb(OTf)3
Yb(NTf2)3
Au(ClO4)3
90
92
94
85
23
81
92
95
68
73
87
87
24
39
29
73
81
89
62
41
92:08
92:08
93:07
85:15
81:19
91:09
91:09
84:16
90:10
90:10
91:09
93:07
90:10
81:19
84:16
82:18
83:17
77:23
83:17
89:11
84
85
96
60
05
23
07
07
07
45
66
02
03
14
08
03
03
03
05
02
Diels-Alder reactions using ligands 10 and 11, which have
medium-sized C-4 substituents (i-Pr 10, Bn 11), followed a
similar trend (entries 2-5) with the exception that 7c gave lower
enantioselectivities than 7b (entries 3, 4).22 Comparison of
results with pyrazolidinone substrate 7d and oxazolidinone
substrate 15 using ligand 10 was particularly interesting. In
analogous room-temperature experiments the pyrazolidinone
substrate gave 95% ee (entry 5) while the oxazolidinone
substrate gave only 23% ee (entry 1), which represents a 24-
fold increase in enantioselection (39:1 vs 1.6:1 er). These results
again supported the principle that pyrazolidinone templates can
amplify enantioselectivity to useful levels at room temperature,
using ordinary chiral ligands of modest steric bulk.
9
10
11
12
13
14
15
16
17
18
19
20
Ligands 12 and 13 are relatively strong ligands in terms of
steric demand. When ligand 12 was used, the pyrazolidinone
templates again amplified enantioselectivity relative to oxazo-
lidinone 15 (77-99% ee versus 54% ee), and again selectivity
increased with the size of the R group. Compounds 7a and 7b
gave 77% ee and 97% ee respectively, while 7c and 7d both
gave 99% ee (entries 2-5). The increase in enantioselectivity
between experiments with 7a and 7b where R ) Et and R )
Ph illustrated the pyrazolidinone R group still played a critical
role in enhancing enantioselectivity, but consistently high
enantioselectivity when R ) Ph, 2-naphthyl, and 1-naphthyl
indicated that the face-shielding ability of the chiral ligand in
conjunction with the pyrazolidinone R group was nearly
complete even with medium sized pyrazolidinone R groups.
a
b
For reaction details see Supporting Information. Yields are for
column-purified products. Endo/exo ratios were determined by 1H NMR.
Enantioselectivities were determined after conversion to the known benzyl
ester using chiral HPLC. Absolute configuration was determined to be
(1S,2S,3R,4R) by comparison of retention times of the known benzyl ester.
c
d
Lewis acids might induce chirality such that the pyrazolidinone
chirality works in concert with the chiral Lewis acid (matched
case) and amplify enantioselectivity. In other cases, the induced
chirality might display a mismatched, dissonant effect. It is also
possible that both configurations of the fluxional pyrazolidinone
might be present and reactive in other instances (see 2 and 3 in
Figure 1).
Our standard screening reaction used substrate 7b and ligand
10 with cycloaddition conducted at room temperature. These
were conditions not designed for optimal enantioselectivity; we
intentionally avoided low temperature and chose a ligand with
only a medium-sized substituent so that high enantioselectivity
would be possible only if strong amplification from the
pyrazolidinone was a contributing factor.
As shown in Table 2, maximal enantioselectivity resulted
when Cu(OTf)2, Cu(ClO4)2, and especially Pd(ClO4)2 were used
as Lewis acids (84%, 85%, and 96% ee, entries 1-3). The
improvement from Cu(ClO4)2 to Pd(ClO4)2 reflects a 4-fold
increase in enantioselectivity (12:1 vs 50:1 er). The 96% ee
observed with Pd(ClO4)2 was quite remarkable given the
reaction temperature and the unexceptional size of the ligand.
This was even more apparent when the bisoxazoline ligand
13 derived from amino indanol was employed. Enantioselec-
tivities were uniformly high (95-96% ee) regardless of the size
of the pyrazolidinone R group (entries 2-5). From this data it
appeared that ligand 13 was the primary controller of enanti-
oselection; but there was still significant amplification in
enantioselectivity when pyrazolidinone substrates were em-
ployed (95-96% ee) versus the non-relay substrate 15 (83%
ee, entry 1). These results suggest the relay group on nitrogen
still played a critical role in enhancing selectivity, even though
the correlation between size of the pyrazolidinone R group and
selectivity was not observed in this series of experiments. The
principle that combination of an excellent ligand with a
pyrazolidinone auxiliary/template can lead to very high enan-
tioselectivities was supported.
(22) Analogous experiments corresponding to entries 2, 3, and 5 (substrates
7a, 7b, and 7d, respectively) in Table 1 employing Cu(OTf)2/(S,S)-2,2′-
isopropylidene-bis(4-phenyl-2-oxazoline) as the chiral Lewis acid were also
conducted. Although a similar trend of increasing enantioselectivity with
increasing size of the fluxional group was observed, the enantioselectivities
were significantly lower in each case. The following enantioselectivities
and endo:exo ratios were obtained with this ligand: (1) for substrate 7a,
28% ee, 88:12 endo:exo; (2) for substrate 7b, 38% ee, 91:09 endo:exo;
and (3) for substrate 7d, 46% ee, 86:14 endo:exo. The sense of absolute
stereochemistry of the cycloadducts was identical to that obtained with
ligands 9-13.
Screening of Lewis Acids and Metal Geometries. Having
established that pyrazolidinones amplify the enantioselectivity
induced by Cu(OTf)2/bisoxazoline complexes, we screened a
series of additional Lewis acids (Table 2). We expected the
impact of the pyrazolidinone group might be dependent on
whether a chiral Lewis acid has a square planar, tetrahedral,
octahedral, or higher coordination environment. Some chiral
9
398 J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007

Role of Achiral Pyrazolidinone Templates
A R T I C L E S
Table 3. Size Effects of Chiral Relay Groups with Various Lewis
The results with the copper and palladium Lewis acids suggested
that amplification of enantioselectivity was excellent for Lewis
acids with square planar geometry. It is well established that
Cu(II)/bisoxazoline catalysts react via distorted square planar
geometries upon complexation to bidentate substrates.5c There
are a number of instances of high selectivity in Diels-Alder
reactions catalyzed by square planar palladium(II) complexes.23
However, the degree to which Pd(II) was superior to Cu(II) in
this case is noteworthy.
Acidsa
entry
substrate
Lewis acid
yield (%)b
endo/exoc
ee (%)d
On the basis of literature reports of Ni(II) complexes with
both square planar and octahedral geometries, Ni(ClO4)2 and
NiBr2 were subsequently screened.24 Ni(ClO4)2 gave 60% ee
(entry 4), while reactivity and selectivity were poor for NiBr2-
catalyzed reactions (entry 5). The reactivity and selectivity
differences may be explained by the strongly dissociating nature
of the perchlorate counterion. The ClO4- counterions are likely
dissociated, leading to a cationic square planar complex, while
the Br- counterions should not be dissociated, leading to a
neutral, less reactive octahedral complex.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
7a
7b
7d
7a
7b
7d
7a
7b
7d
7a
7b
7d
7a
7b
7d
7a
7b
7d
7a
7b
7d
7a
7b
7d
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(ClO4)2
Cu(ClO4)2
Cu(ClO4)2
Pd(ClO4)2
Pd(ClO4)2
Pd(ClO4)2
Ni(ClO4)2
Ni(ClO4)2
Ni(ClO4)2
Mg(ClO4)2
Mg(ClO4)2
Mg(ClO4)2
Zn(ClO4)2
Zn(ClO4)2
Zn(ClO4)2
Fe(ClO4)3
Fe(ClO4)3
Fe(ClO4)3
Fe(ClO4)2
Fe(ClO4)2
Fe(ClO4)2
90
90
85
86
92
93
54
94
97
96
85
93
79
81
86
70
87
84
19
39
12
33
29
44
96:04
92:08
93:07
92:08
92:08
93:07
91:09
93:07
95:05
87:13
85:15
83:17
87:13
91:09
91:09
88:12
91:09
92:08
87:13
81:19
80:20
81:19
84:16
86:14
56
84
95
78
85
92
45
96
99
56
62
60
20
23
20
54
66
74
23
14
00
21
08
04
We then chose to examine Lewis acids that have been
reported to give tetrahedral or octahedral metal geometries. Of
the Mg(II) Lewis acids25 screened, only Mg(ClO4)2 provided
any significant level of enantioselection (23% ee, entry 6), while
Mg(NTf2)2, Mg(OTf)2, and MgI2 all gave nearly racemic product
(entries 7-9). Zn(ClO4)2 gave the best results of the Zn(II) salts
screened (66% ee, entry 11). Zn(OTf)2 gave cycloadduct in 45%
ee (entry 10). Zn(NTf2)2 provided nearly racemic product in
good yield (entry 12), while reactivity and selectivity for ZnI2
were poor (24% yield, 3% ee, entry 13). Fe(III) and Fe(II) Lewis
acids were then screened on the basis of previous reports of
success when combined with bisoxazoline ligands.26 However,
reactivities and selectivities were poor when Fe(ClO4)3 and Fe-
(ClO4)2 (entries 14 and 15) were used. On the basis of these
results it was determined that tetrahedral or octahedral com-
plexes were not optimal for enantioselective Diels-Alder
reactions with pyrazolidinone substrates.
a
b
For reaction details see Supporting Information. Yields are for
column-purified products. Endo/exo ratios were determined by 1H NMR.
Enantioselectivities were determined after conversion to the known benzyl
c
d
ester using chiral HPLC. Absolute configuration was determined to be
(1S,2S,3R,4R) by comparison of retention times of the known benzyl ester.
Dependence of Enantioselectivity on the Size of the
Pyrazolidinone R Group with Various Lewis Acids. Our
initial experiments with Cu(OTf)2/i-Pr bisoxazoline complex
demonstrated the dramatic effects of increasing the size of the
pyrazolidinone R group (Table 3, entries 1-3). It was thus
desirable to see whether this size dependence would be observed
with other Lewis acid/metal geometries. Not surprisingly, a
similarly strong size effect was observed with Cu(ClO4)2 as the
Lewis acid (entries 4-6). For Pd(ClO4)2 the effect was even
more dramatic. Reactions with 7a, where R ) Et, gave 45% ee
(entry 7). However, substrates 7b and 7d with R ) Ph and
1-naphthyl, respectively, gave 96% ee and 99% ee (entries 8
and 9).
Two possibilities exist for the increased enhancements
observed with Pd(ClO4)2 versus Cu(ClO4)2 (compare entries 5
and 6 with entries 8 and 9). One possible explanation is that
the reactive Pd(II) complex may be a less distorted square planar
complex than the analogous Cu(II) complex, leading to more
efficient shielding by the pyrazolidinone R group and/or the
C-4 isopropyl group.5c,28 Alternatively, the N-Pd-O bond
lengths in the Pd(II) complex are likely significantly longer than
the N-Cu-O bond lengths in the Cu(II) complex.29 This
Entries 16-19 show results from a number of Lewis acids
capable of providing higher-coordinate complexes. Although
reactivities were moderate to good (62-89% yield) with Sm-
(OTf)3, Sc(OTf)3, Yb(OTf)3, and Yb(NTf2)3, enantioselectivities
were negligible (entries 16-19). Finally, a Au(III) salt was used
on the basis of reports of square planar Au(III) complexes, but
Au(ClO4)3 gave essentially racemic product in 41% yield (entry
20).27 This result is likely due to poor solubility of the Au(III)/
ligand complex in dichloromethane.
(23) (a) Ghosh, A. K.; Matsuda, H. Org. Lett. 1999, 1, 2157. (b) Nakano, H.;
Takahashi, K.; Okuyama, Y.; Senoo, C.; Tsugawa, N.; Suzuki, Y.; Fujita,
R.; Sasaki, K.; Kabuto, C. J. Org. Chem. 2004, 69, 7092. (c) Nakano, H.;
Okuyama, Y.; Suzuki, Y.; Fujita, R.; Kabuto, C. Chem. Commun. 2002,
1146. (d) Nakano, H.; Suzuki, Y.; Kabuto, C.; Fujita, R.; Hongo, H.
J. Org. Chem. 2002, 67, 5011. (e) Manchen˜o, O. G.; Arraya´s, R. G.;
Carretero, J. C. Organometallics, 2005, 24, 557. (f) Nakano, H.; Tsugawa,
N.; Fujita, R. Tetrahedron Lett. 2005, 46, 5677.
(24) (a) Venkataraman, D.; Du, Y.; Wilson, S. R.; Hirsch, K. A.; Zhang, P.;
Moore, J. S. J. Chem. Educ. 1997, 74, 915. (b) Kanemasa, S.; Adachi, K.;
Yamamoto, H.; Wada, E. Bull. Chem. Soc. Jpn. 2000, 73, 681. (c) Evans,
D. A.; Downey, C. W.; Hubbs, J. L. J. Am. Chem. Soc. 2003, 125, 8706.
(25) (a) Corey, E. J.; Ishihara, K. Tetrahedron Lett. 1992, 33, 6807. (b) Carbone,
P.; Desimoni, G.; Faita, G.; Filippone, S.; Righetti, P. Tetrahedron 1998,
54, 6099.
(26) (a) Sibi, M. P.; Petrovic, G. Tetrahedron: Asymmetry 2003, 14, 2879. (B)
Usuda, H.; Kuramochi, A.; Kanai, M.; Shibasaki, M. Org. Lett. 2004, 6,
4387. (c) Bromidge, S.; Wilson, P. C.; Whiting, A. Tetrahedron Lett. 1998,
39, 8905. (d) Bart, S. C.; Hawrelak, E. J.; Schmisseur, A. K.; Lobkovsky,
E.; Chirick, P. J. Organometallics 2004, 23, 237.
(27) (a) Yanagisawa, A. In Lewis Acids in Organic Synthesis; Yamamoto, H.;
Ed., Wiley-VCH: Weinheim, 2000; Vol. 2, p 575. (b) Mansour, M. A.;
Lachicotte, R. J.; Gysling, H. J.; Eisenberg, R. Inorg. Chem. 1998, 37,
4625. (c) Ortner, K.; Abram, U. Polyhedron 1999, 749.
9
J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007 399

A R T I C L E S
Sibi et al.
Table 4. Diels-Alder Reactions on Templates with C-5 Variationa
increase in bond length may in turn result in more efficient
shielding by the pyrazolidinone R group. While it is possible
that the C-4 isopropyl group provides improved face shielding
with Pd(II), this seems unlikely since an increase in N-Pd-O
bond lengths effectively moves the C-4 substituent further from
the reaction center.
While reactions involving Cu(II) and Pd(II) showed a strong
correlation between enantioselectivity and the size of the
pyrazolidinone R group, this pattern was not shared by Ni(II).
Substrates 7a, 7b, and 7d gave 56%, 62%, and 60% ee,
respectively (entries 10-12). These results suggested that the
reactions proceed through a metal geometry significantly
different from that for Cu(II) and Pd(II), such that the pyrazo-
lidinone R group provides very little face shielding.
The dependence of enantioselectivity on the size of the R
group varied widely for Lewis acids that probably involve
tetrahedral or octahedral complexes. There was no apparent
dependence on size of the R group when Mg(ClO4)2 was
employed as the Lewis acid (entries 13-15). With Zn(ClO4)2,
a systematic increase in enantioselectivity correlated with
increasing size of R, with 7a, 7b, and 7d giving 54%, 66%,
and 74% ee, respectively (entries 16-18). On the other hand,
both Fe(ClO4)3 and Fe(ClO4)2 gave diminishing selectivity on
the already negligible enantioselectivity as the R group increased
from ethyl to 1-methylnaphthyl (entries 19-21 for Fe(ClO4)3
and 22-24 for Fe(ClO4)2).
Several generalizations can be drawn from these results. First,
it appears that for bisoxazoline/Lewis acid complexes with
square-planar-type geometries, pyrazolidinone auxiliaries tend
to amplify enantioselectivity to a large degree, and the degree
of amplification is maximized as the size of the pyrazolidinone
R group increases. Second, even with non-square planar
geometries, there appears to be a significant dependence on the
size of the pyrazolidinone group, although not always in a
predictable way. Many of the less selective Lewis acids included
in our survey are not well matched with the isopropyl bisox-
azoline ligand, so both the overall enantioselectivity and the
dependence on the pyrazolidinone size are muted. It does not
necessarily follow that pyrazolidinones might not provide
meaningful amplification of enantioselectivity when more
a
For reaction details see Supporting Information. Endo/exo ratios were
determined by 1H NMR, and ee determination was carried out after
conversion to the known benzyl ester using chiral HPLC. Absolute
configuration was determined to be (1S,2S,3R,4R) by comparison of
retention times of the known benzyl ester. Yields for isolated column-
purified material averaged 85-95%, except for R1 ) Bn (yields around
65%).
effective chiral Lewis acids involving Mg(II), Zn(II), Ni(II), and
other metals are used.15-18
C-5 Variation Effects. With a greater understanding of the
importance of metal geometry and of the size of the pyrazoli-
dinone nitrogen substituent, the effect of C-5 substitution was
investigated (Table 4). We had previously demonstrated that
the nitrogen substituent plays an important face-shielding role
when C-5 was dimethyl substituted (Table 1), but questions
remained as to whether this substitution was necessary in order
to place the chiral relay substituent on nitrogen into close
proximity to the reaction center. Thus, substrates 20-22a where
R1 ) H were prepared. Diastereoselectivities and enantioselec-
tivities were markedly lower for substrates where R1 ) H (50%
ee for 20a, 65% ee for 21a, and 63% ee for 22a) when compared
with substrates where R1 ) Me (56% ee for 20b, 84% ee for
21b, and 95% ee for 22b).
The lack of relay group effect was evident upon examination
of entry 1, where increasing the size of the nitrogen substituent
R2 from benzyl (65% ee) to 1-naphthylmethyl (63% ee) did
not amplify the enantioselectivity. This was in stark contrast to
entry 2 where a significant positive effect on enantioselectivity
was observed when the size of the pyrazolidinone R2 group was
increased. This difference can be explained by the lack of steric
interaction between R1 substituents and the R2 relay group when
R1 ) H, which allows a favored conformation in which the
nitrogen substituent is rotated away from the reaction center
(Scheme 3). In the case where R1 ) Me, the conformation in
which the N(1) substituent is rotated away from the reaction
center is likely disfavored by an adverse interaction between
the benzyl group and the C5-methyl hydrogen atoms. Thus, the
conformation in which the relay group is positioned near the
reaction center is favored, leading to enhanced face shielding.
On the basis of these results an investigation of additional
C-5 substituents was undertaken to evaluate whether increased
steric volume at this position would lead to increased enanti-
oselectivity. Additional substrates with R1 ) Et, -(CH2)4-,
(28) For examples of distorted square planar Cu(II)/bisoxazoline complexes,
see: (a) Evans, D. A.; Peterson, G. S.; Johnson, J. S.; Barnes, D. M.;
Campos, K. R.; Woerpel, D. A. J. Org. Chem. 1998, 63, 4541. (b) Evans,
D. A.; Rovis, T.; Johnson, J. S. Pure Appl. Chem. 1999, 71, 1407. (c)
Evans, D. A.; Johnson, J. S.; Burgey, C. S.; Campos, K. R. Tetrahedron
Lett. 1999, 40, 2879. (d) Evans, D. A.; Johnson, J. S.; Olhava, E. J. J. Am.
Chem. Soc. 2000, 122, 1635. (e) Thorhauge, J.; Roberson, M.; Hazell,
R. G.; Jørgensen, K. A. Chem. Eur. J. 2002, 8, 1888. (f) Evans, D. A.;
Miller, S. J.; Lectka, T.; Von, Matt, P. J. Am. Chem. Soc. 1999, 121, 7559.
(g) Evans, D. A.; Scheidt, K. A.; Johnson, J. N.; Willis, M. C. J. Am. Chem.
Soc. 2001, 123, 4480. (h) Evans, D. A.; Rovis, T.; Kozlowski, M. C.;
Downey, C. W.; Tedrow, J. S. J. Am. Chem. Soc. 2000, 122, 9134. (i)
Audrain, H.; Thorhauge, J.; Hazell, R. G.; Jørgensen, K. A. J. Org. Chem.
2000, 65, 4487. (j) Evans, D. A.; Seidel, D.; Rueping, M.; Lam, H. W.;
Shaw, J. T.; Downey, C. W. J. Am. Chem. Soc. 2003, 125, 12692. For
examples of pseudo-square-planar Pd(II)/bisoxazoline complexes, see: (k)
Von, Matt, P.; Lloyd-Jones, G. C.; Minidis, A. B. E.; Pfaltz, A.; Macko,
L.; Neuburger, M.; Zehnder, M.; Ru¨egger, H.; Pregosin, P. S. HelV. Chim.
Acta 1995, 78, 265. (l) Pfaltz, A. Acc. Chem. Res. 1993, 26, 339. (m)
Hoarau, O.; A¨ıt-Haddou, H.; Daran, J. C.; Cramaile`re, D.; Balavoine,
G. G. A. Organometallics 1999, 18, 4718.
(29) For Cu-N and Cu-O bond lengths in Cu(II)/bisoxazoline/bidentate oxygen
donor complexes, see: (a) Evans, D. A.; Rovis, T.; Kozlowski, M. C.;
Tedrow, J. S. J. Am. Chem. Soc. 1999, 121, 1994. For a representative
Pd-N bond length in a Pd(II)/bisoxazoline complex, see reference 27(l).
For
a representative PdlO bond length, see: (b) Bugarcic, Z. D.;
Nandibewoor, S. T.; Hamza, M. S. A.; Heinemann, F.; van Eldik, R. Dalton
Trans. 2006, 2984.
9
400 J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007

Role of Achiral Pyrazolidinone Templates
A R T I C L E S
Scheme 3. Role of C-5 Substitution in Determining Enantioselectivity
Table 5. Evaluation of Substrate Scopea
-(CH2)5-, and Bn were prepared with R2 ) Et (20c-f), Bn
(21c-f), and 1-CH2naphthyl (22c-f). Although enantioselec-
tivities increased as the size of the R1 substituents was increased
irrespective of the identity of the R2 relay group, the trend was
quite apparent with substrates where R2 ) Bn (see the 21a-f
column, Table 4). In this case the baseline selectivity for R1 )
Me was 84% ee (entry 2). When substrate 21c with R1 ) Et
was used, the enantioselectivity increased to 87% (entry 3).
Enantioselectivity was further increased when the R1 groups
were cyclopentyl (88% ee, entry 4) or cyclohexyl (92% ee, entry
5). When substrate 21f with R1 ) Bn was employed, the
enantioselectivity rose to 95% even with the medium sized
benzyl R2 relay group (entry 6). These results clearly indicated
that the nature of C-5 substitution was critical for optimal
selectivity in reactions involving pyrazolidinone templates.
Study of the Scope of Pyrazolidinone Dienophiles in
Diels-Alder Reactions. Knowledge of the elements necessary
for high selectivity with pyrazolidinone templates was then
applied to templates with a number of common dienophiles
appended. Although we had determined that relay substrates
containing bulkier C-5 substituents provided optimal enanti-
oselectivity, substrates with C-5 methyl groups were chosen as
most practical for further evaluation due to the commercial
availability of ethyl 3,3-dimethyl acrylate, which is the starting
material necessary for preparation of the pyrazolidinone core.
The N(1)-CH2naphthyl substituent was determined to be the
pyrazolidinone substituent of choice based on previous results
(Tables 1 and 3). Cu(OTf)2-catalyzed reactions with acrylate
26 gave 85% ee (Table 5, entry 1). However, the enantiose-
lectivity of this reaction was improved to 90% ee upon lowering
the reaction temperature to 0 °C (entry 3). Pd(ClO4)2 was also
an effective Lewis acid for 26, providing product in 87% ee at
room temperature (entry 2). As reported previously, Cu(OTf)2
and Pd(ClO4)2 both proved excellent for crotonate 7d, giving
95% ee and 99% ee, respectively (entries 4 and 5). When
cinnamate substrate 28 bearing the 1-naphthlylmethyl group was
employed under standard Cu(OTf)2-catalyzed conditions, reac-
tivity was very poor. Reactivity was increased marginally by
using cinnamate 27, which contains the benzyl R group (entry
6). Yields were still poor (35%), and enantioselectivity was
moderate (70% ee). However, moderate yields were obtained
with 28 with Pd(ClO4)2 (entry 7). In this case the Diels-Alder
adduct was isolated in 57% yield with 88% ee. Fumarate
substrate 29 gave only 76% ee in the Cu(OTf)2-catalyzed
reactions (entry 8). This result was improved with Pd(ClO4)2,
which resulted in product isolated in 91% ee (entry 9). Cu-
(OTf)2-catalyzed reaction of 29 was further improved to give
96% ee when the reaction was performed at -20 °C (entry 10).
yield
(%)
ee
(%)
entry
R1
R2
temp
Lewis acid
endo/exo
1
2
3
4
5
6
7
8
9
10
H (26)
H (26)
H (26)
1-Naph rt
1-Naph rt
1-Naph 0 °C
1-Naph rt
1-Naph rt
Cu(OTf)2
Pd(ClO4)2
Cu(OTf)2
Cu(OTf)2
Pd(ClO4)2
Cu(OTf)2
Pd(ClO4)2
Cu(OTf)2
Pd(ClO4)2
93
89
93
85
97
35
57
94
80
99
91:09
95:05
93:07
93:07
95:05
85:15
83:17
86:14
93:07
93:07
85
87
90
95
99
70
88
76
91
96
Me (7d)
Me (7d)
Ph (27)
Ph (28)
CO2Et (29) 1-Naph rt
CO2Et (29) 1-Naph rt
Ph
rt
1-Naph rt
CO2Et (29) 1-Naph -20 °C Cu(OTf)2
a
For reaction details see Supporting Information. Endo/exo ratios were
determined by 1H NMR, and ee determination was carried out using chiral
HPLC.
It again bears mention that these selectivities were induced using
only the modest isopropyl bisoxazoline ligand 10; most likely
a stronger ligand would provide higher enantioselectivities.
Evaluation of Dienes with Pyrazolidinone Acrylates. A
sample of common dienes was then evaluated using acrylate
substrate 26 with the 1-naphthylmethyl relay group (Table 6).
Room-temperature Cu(OTf)2-catalyzed reaction of isoprene (35)
with 26 gave cycloadduct in 87% ee (entry 1). Enantioselectivity
increased moderately when the reaction temperature was
lowered to 0 °C with product obtained in 90% ee (entry 3). A
more dramatic increase in enantioselectivity was observed using
Pd(ClO4)2 (entry 2), which boosted the selectivity for isoprene
addition to 97% ee and also improved the yield to 92%. A
similar trend was observed for reaction of 2,3-dimethylbutadiene
(36) and cyclohexadiene (37) with 26. Cu(OTf)2-catalyzed
reactions of 2,3-dimethylbutadiene at room temperature and 0
°C gave 88% ee and 92% ee (entries 4 and 6), respectively,
while the corresponding experiments with cyclohexadiene gave
85% ee and 90% ee (entries 7 and 9). But when Pd(ClO4)2-10
was used at room temperature, both 2,3-dimethylbutadiene and
cyclohexadiene provided cycloadducts in 98% ee (entries 5 and
8). The superiority of Pd(II) over Cu(II) in terms of both
reactivity and enantioselectivity was even more pronounced for
cycloaddition of 1-(phenoxycarbonyl)-1,2-dihydropyridine (38)
to pyrazolidinone acrylates 34 and 26. Pd(II)-catalyzed cycload-
dition to N-benzyl substrate 34 gave product in 77% yield and
92% ee (entry 11), while product was isolated in only 38% yield
9
J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007 401

A R T I C L E S
Sibi et al.
Table 6. Diene Study with Acrylate 26 or 34a
Table 7. Lewis Acid Loading Studya
entry
mol % CLA
Lewis acid
yield (%)b
endo/exoc
ee (%)d
1
2
3
4
5
6
7
8
2.5
5
10
15
2.5
5
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Cu(OTf)2
Pd(ClO4)2
Pd(ClO4)2
Pd(ClO4)2
Pd(ClO4)2
75
86
90
85
91
97
94
97
93:07
93:07
93:07
93:07
96:04
95:05
95:05
95:05
95
95
95
95
99
99
99
99
10
15
a
b
For reaction details see Supporting Information. Yields are for
column-purified products. Endo/exo ratios were determined by 1H NMR.
Enantioselectivities were determined after conversion to the known benzyl
c
d
ester using chiral HPLC. Absolute configuration was determined to be
(1S,2S,3R,4R) by comparison of retention times of the know benzyl ester.
a
For reaction details see Supporting Information. Endo/exo ratios were
ligand loadings are possible in systems in which the rate of
uncatalyzed background cycloaddition is low.
determined by 1H NMR, and ee determination was carried out using chiral
HPLC.
Conformational Insights Based on the Crystal Structure
for 7d. Although indirect evidence for the role of the pyrazo-
lidinone R group was strong at this point, additional insights
were provided by the crystal structure of substrate 7d (Figure
2). The carbonyls in this crystal structure have an interesting
syn orientation. Other pyrazolidinone substrates such as 44
crystallize with the carbonyls anti; thus, the structure for 7d
may reflect crystal packing effects. Nevertheless, the structure
for 7d provides an interesting image of how the naphthylmethyl
substituent might position itself when the carbonyls are locked
into syn alignment, as would be expected following Lewis acid
coordination. It is easy to envision from structure 7d how the
aryl portion of the N(1) substituent might block the bottom face
of the enoyl group.
and 72% ee in the corresponding Cu(II)-catalyzed experiment
(entry 10). Yield and enantioselectivities were likewise improved
when N(1)-naphthylmethyl substrate 26 was employed. Pd(II)-
catalyzed cycloaddition gave product in 83% yield and 98% ee
(entry 13), while product was isolated in only 49% yield and
77% yield in the corresponding Cu(II)-catalyzed experiment
(entries 11 and 12). Entry 13 represents the best enantioselec-
tivity reported to date for Diels-Alder cycloaddition of 1-(aryl-
loxycarbonyl)- or 1-(alkoxycarbonyl)-1,2-dihydropyridines.14b
Related isoquinuclidines have been proposed as key intermedi-
ates in the synthesis of a number of natural products, which
continue to be of interest.30
Lewis Acid Loading Study. Upon determining that the
reactivity and selectivity of the template/catalyst systems were
sufficient to be of interest, we recognized the need for catalysts
loadings below the 15 mol % loadings reported to this point.
As such, cycloaddition of cyclopentadiene to substrate 7d was
chosen to evaluate if catalyst loadings could be decreased in
Cu(OTf)2- and Pd(ClO4)2-catalyzed reactions (Table 7). The
baseline 15 mol % catalyst loading experiment for Cu(OTf)2
provided product in 85% yield and 95% ee (entry 4). We were
pleased to observe no loss in enantioselectivity upon lowering
the catalytic loading from 15 to 2.5 mol % (entries 1-4). This
trend was also seen in Pd(ClO4)2-catalyzed experiments. Catalyst
loadings of 2.5, 5, 10, and 15 mol % all gave high yields of
cycloadducts (91-97% yield) in 99% ee (entries 5-8). Although
results from room-temperature catalyst loading experiments with
crotonate 7d may not correlate to all relay substrate/diene
combinations, they do suggest that high selectivities at low
Figure 2. Crystal structures of 44 and 7d.
π-Stacking Studies. The crystal structure for 7d suggested
the possibility that π-stacking might play a role in the amplifica-
tion of enantioselectivity afforded by pyrazolidinones.31 We have
conducted several studies to address this possibility (Table 8).
(30) (a) Kuehne, M. E.; Marko, I. In Syntheses of Vinblastine-type Alkaloids.
The Alkaloids. Antitumor Bisindole Alkaloids from Catharanthus roseus
(L.); Brossi, A., Suffness, M., Eds.; Academic Press: San Diego, 1990;
Vol. 37, pp 77. (b) Popik, P.; Skolnick, P. In Pharmacology of Ibogain
and Ibogaine-related Alkaloids. The Alkaloids. Chemistry and Biology;
Cordell, G. A., Ed.; Academic Press: Sand Diego, 1999; Vol. 52, p 197.
(c) Redding, M. T.; Fukuyama, T. Org. Lett. 1999, 1, 973. (d) Martin,
S. F.; Rueger, H.; Williamson, S. A.; Grzejszczak, S. J. Am. Chem. Soc.
1987, 109, 6124.
(31) Evans, D. A.; Chapman, K. T.; Bisaha, J. J. Am. Chem. Soc. 1988, 110,
1238.
9
402 J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007

Role of Achiral Pyrazolidinone Templates
A R T I C L E S
Table 8. π-π Interaction Studya
Table 9. Reactions with Relay Templates Using C1 Chiral
Ligandsa
yield
(%)b
ee
(%)d
entry
R
Lewis acid
endo/exoc
1
2
3
4
5
6
7
8
Ph (7b)
Ph (7b)
1-Naph (7d)
Cu(OTf)2
Pd(ClO4)2
Cu(OTf)2
Pd(ClO4)2
Cu(OTf)2
Pd(ClO4)2
Cu(OTf)2
Pd(ClO4)2
Cu(OTf)2
Pd(ClO4)2
Cu(OTf)2
Pd(ClO4)2
90
94
85
97
92
91
91
89
92
88
91
85
92:08
93:07
93:07
95:05
91:09
96:04
89:11
94:06
90:10
94:06
92:08
93:07
84
96
95
99
66
97
88
98
86
97
86
99
1-Naph (7d)
cyclohexyl (45a)
cyclohexyl (45a)
4-MeO-C6H4 (45b)
4-MeO-C6H4 (45b)
4-F-C6H4 (45c)
4-F-C6H4 (45c)
4-MeO2C-C6H4 (45d)
4-MeO2C-C6H4 (45d)
9
10
11
12
a
b
For reaction details see Supporting Information. Yields are for
column-purified products. Endo/exo ratios were determined by 1H NMR.
Enantioselectivities were determined after conversion to the known benzyl
c
d
ester using chiral HPLC. Absolute configuration was determined to be
(1S,2S,3R,4R) by comparison of retention times of the know benzyl ester.
a
Endo/exo ratios were determined by 1H NMR. Enantioselectivities
were determined after conversion to the known benzyl ester using chiral
HPLC. Absolute configuration was determined to be (1S,2S,3R,4R) by
comparison of retention times of the known benzyl ester. Yields for isolated
column-purified material averaged 85-90%. Reaction at -23 °C using
50 mol % chiral Lewis acid.
b
We prepared substrate 45a with a cyclohexylmethyl substituent
that lacks a π-system for comparison to the N(1)-benzyl
substrate 7b. Initial results with Cu(OTf)2-catalyzed reaction
of 45a gave 66% ee (entry 5), which is significantly lower than
the values of 84% ee achieved with 7b where R ) Ph (entry 1)
and 95% with 7d where R ) 1-naphthyl (entry 3). This result
initially suggested that π-π interaction may be important.
However, analogous experiments catalyzed by Pd(ClO4)2 gave
97% ee with cyclohexyl substrate 45a (entry 6), a selectivity
which is comparable to those observed with phenyl and
1-naphthyl substrates 7b and 7d (96% and 99% ee, entries 2
and 4). These results suggest that π-π interaction is not essential
for high enantioselectivity.
We also prepared additional substrates 45b-d, which incor-
porate electron-donating or -withdrawing groups onto the
N-benzyl group. We reasoned the 4-MeO-C6H4 π-system would
be more electron rich relative to the Ph π-system, which could
enhance a π-stacking interaction with the electron-deficient
crotonate group. On the other hand, the 4-F-C6H4 and 4-MeO2C-
C6H4 π-systems would be less electron rich relative to the Ph
π-system, leading to less interaction with the crotonate. On the
basis of this proposal, increased enantioselectivity was expected
with 45b, whereas decreased enantioselectivity was expected
with 45c and 45d. This reasoning proved to be false as electron-
rich 45b gave 88% ee (entry 7), while electron-poor 45c and
45d both gave 86% ee (entries 9 and 11) in Cu(OTf)2
experiments. The parent benzyl 7b gave 84% ee (entry 1). These
results point to the insignificance of π-π interactions in
determining enantioselectivity. Steric volume appears to be the
foremost factor responsible for enhanced selectivity. This
conclusion is further supported by the similar trend observed
in Pd(II)-catalyzed experiments (compare entries 2, 8, 10, and
12).
demonstrating chiral relay was devised using non-C2-symmetric
ligands 50-52 (Table 9), which afford almost no enantiose-
lectivity when applied to oxazolidinone substrate 15 (entry 1).
Use of a non-C2-symmetric ligand such as 50 most likely
provides a square planar complex, but two distinctly different
complexes are possible (Figure 3). The most probable complex
47 has the ligand’s isopropyl group occupying an opposite
quadrant that is remote relative to both the reactive crotonate
center and the pyrazolidinone benzyl group. That the isopropyl
group is too distant to provide any meaningful face shielding
is reflected by the negligible enantioselectivities observed using
achiral oxazolidinone 15 and substrate 49 (entries 1 and 2).
Figure 3. Possible square planar complexes with non-C2-symmetric ligands.
For reactions occurring through complex 47 (Figure 3),
appreciable enantioselectivity seems likely only if the pyrazo-
lidinone benzyl group provides the actual face shielding at the
reactive center. This would then be evidence of the “chiral relay”
concept, in which a permanently chiral center (in this case a
rather remote one) is able to bias the configuration of the
fluxional nitrogen, and it is the substituent on the fluxional
nitrogen that actually shields the reactive center. The chirality
Evaluation of Pyrazolidinone Templates When Using Non-
C2 Symmetric Ligands. An even more stringent test for
9
J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007 403

A R T I C L E S
Sibi et al.
at the fluxional nitrogen controls enantioselectivity, but the
chirality at the fluxional nitrogen is induced or “relayed” from
the fixed chiral center.
Reactions using ligand 50 and a series of pyrazolidinones
showed a clear correlation between the size of the pyrazolidinone
R group and enantioselectivity. Substrates 49 and 7a with small
pyrazolidinone R groups (H and Et) gave 4 and 29% ee (Table
9, entries 2 and 3). Enantioselectivities increased to 47 and 56%
ee using benzyl and 1-naphthylmethyl pyrazolidinones 7b and
7d (entries 4 and 5). Optimal enantioselectivity was once again
observed with 1-naphthylmethyl pyrazolidinones 7d (69% ee,
entry 6), and could be increased to 88% at -23 °C (entry 7).
This increase in enantioselectivity with increasing size of the
pyrazolidinone relay group R followed the same trend as was
observed with C2-symmetric bisoxazoline ligands (Table 1), and
gave the same absolute configuration as when isopropyl
bisoxazoline ligand 10 was used. While the magnitude of the
enantioselectivity was diminished using ligand 50 versus ligand
10, it is actually rather surprising how modest the diminishment
of selectivity was.
Analogous trends were observed using nonsymmetric and
readily accessible ligands 51 and 52. In each case, there was
no selectivity using oxazolidinone 15 or N-H pyrazolidinone
49 (entries 1 and 2), but as the pyrazolidinone R groups got
larger, increasing levels of enantioselectivity resulted. All of
these results seem to demonstrate that chiral relay was operative
when the pyrazolidinone R group is sufficiently large and that
the pyrazolidinone R groups provide the actual face shielding
when ligands such as 50-52 are used. It is interesting to note
that similar results were obtained with ligands 50 and 52, capable
of forming six-membered chelates versus ligand 51, which forms
a five-membered chelate.
NMR Data, Model, and Discussion. As a means to further
evaluate the location and/or configuration of the fluxional group
in solution, an extensive 1D nOe NMR study was conducted
on a complex of Pd(OTf)2/ligand 12/substrate 7d.32 Although
this study was not conclusive due to the small enhancements
observed, it suggested C-5 substitution was critical for orienta-
tion of the fluxional group near the olefin. In fact a small
enhancement was observed at the C-8 proton of the naphthyl
ring upon irradiation of the proton R to the carbonyl at 5.37
ppm. This observation is consistent with the fluxional group in
near proximity to the olefin.
The results in Tables 1 and 9 can be rationalized using a
distorted square-planar-geometry model relative to Cu(II) or Pd-
(II), with the ligand occupying two sites while the template is
coordinated in a bidentate fashion in an s-cis conformation (the
complex with non-C2 50 is shown in Figure 4). The ligand
isopropyl group in the upper left quadrant orients the pyrazo-
lidinone R group to the rear of the lower right quadrant, where
it blocks the back face of the crotonate. Thus, the 1-naphthyl-
methyl group “relays” chirality from the ligand’s isopropyl
group to the reactive centers in the substrate.
Figure 4. Distorted square planar models with ligands 50 and 10.
substituent act in concert (matched double stereoselection) to
mutually amplify the selectivity. In both non-C2 and C2-
symmetric cases the proposed models suggest approach of the
diene to the re face, resulting in the observed (1S,2S,3R,4R)
configuration.
For C2-symmetric chiral bisoxazolines, a question is why the
pyrazolidinone substituent would occupy the lower right rear
quadrant rather than the front quadrant. The fluxional substituent
can either be cis to the upper right bisoxazoline substituent and
trans to the upper left substituent, or trans to the upper right
bisoxazoline substituent and cis to the upper left substituent.
The first situation would seemingly be a matched situation
resulting in amplified enantioselectivity (as is observed), the
second situation would result in a mismatch leading to reduced
enantioselectivity. While the latter situation seems intuitively
more likely, it obviously does not conform to the actual
experimental observations.
We do not have a complete explanation for the underlying
organizational factors. Unfortunately we have been unable to
obtain crystal structures for any Cu(II)/bisoxazoline/pyrazoli-
dinone complexes or for the Pd(II) analogues. One possibility
is that, when the fluxional substituent is in the front quadrant,
then neither face of the dienophile is accessible, with the chiral
ligand blocking the back and the fluxional substituent blocking
the front. If so, then reaction might be preferable via the
conformation in which the fluxional group is on the same side
as the ligand substituent, so that the back face is very well
blocked but the front face is completely exposed. Thus, a Curtin-
Hammett situation might be involved in which both configura-
tions are in rapid equilibrium, but reaction via the matched
configuration is much faster than reaction via the mismatched
configuration.
Other factors may also impact the location of the fluxional
group: (1) the copper complexes are probably significantly
distorted from planarity; (2) the energy differential between
having the fluxional group in the front versus the back is
relatively small, and actually favors the matched configuration
in which the fluxional substituent is in the back; and (3) the
In reactions with the C2-symmetric chiral bisoxazoline 10
(Figure 4), the additional ligand isopropyl group would occupy
the upper right rear quadrant and actually reinforce shielding
of the back face of the crotonate. Thus, in reactions with the
C2 ligands both the pyrazolidinone R group and the C-4
(33) Variable-temperature NMR experiments on the complex derived from Pd-
(ClO4)2, bisoxazoline 10, and pyrazolidinone substrate 7d showed no line
broadening with lower temperature (from room temperature to -35 °C),
suggesting nitrogen inversion is sufficiently rapid that it is not observed
on the NMR time scale.
(32) Experimental details, analysis, and spectra for nOe experiments are provided
in the Supporting Information.
9
404 J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007

Role of Achiral Pyrazolidinone Templates
A R T I C L E S
barrier to nitrogen inversion appears to be very small, such that
rapid interconversion of the fluxional substituent is likely.33
The above discussion has focused on how pyrazolidinone
groups can amplify enantioselectivity via matched double
enantioselection, in which the fluxional substituent provides
active face shielding at the reactive center. This certainly seems
to be operative with some of the weak ligands investigated, such
as methyl bisoxazoline 6 and nonsymmetric ligands 50-52.
However, it is also likely that pyrazolidinone templates may
both amplify enantioselection and provide other benefits in
situations where the fluxional substituent is not the primary face-
shielding element. The pyrazolidinone templates in structures
7a-d are significantly larger than traditional auxiliaries like
oxazolidinone, and with adjustable size. By simply occupying
more space in a metal complex, other elements in an array may
be pushed closer together. If a chiral ligand and a reactive center
are squeezed closer together, amplification of enantioselectivity
may result, even if the pyrazolidinone substituent is not itself
directly shielding the face of a reactive center. The bulky
pyrazolidinone may simply be pushing the ligand nearer to the
reactive center so that the chiral ligand itself provides more
effective face shielding. This effect may be operative when
strong ligands such as 13 are used, and this basis for amplified
enantioselectivity need not be limited to square planar Lewis
acids. While the effect was not entirely clear in this Diels-
Alder study using Mg(II)/7 (Table 3), we have elsewhere
documented that the use of pyrazolidinone auxiliaries amplify
enantioselectivity for Mg(II)/13-catalyzed additions of hydroxy-
lamines and radicals, and do so to an increasing degree as the
fluxional group gets larger.17,18 The large size of the pyrazoli-
dinone template may also benefit regioselectivity as we have
observed for enantioselective nitrile oxide cycloadditions. One
other attractive feature of pyrazolidinone substrates that we have
observed in other studies is that they have relatively strong
coordinating abilities toward many Lewis acids (relative to
oxazolidinone analogues). As a result Lewis-acid-catalyzed
processes involving pyrazolidinones may be more tolerant of a
wider range of functionalilty in both reactants and solvents, since
the Lewis acid is less likely to be sequestered from the
pyrazolidinone substrate by the other functional groups.
Conclusion
We have further demonstrated utility of pyrazolidinone
templates in enantioselective Diels-Alder reactions. Enanti-
oselectivities in room-temperature reactions were found to be
highly dependent on the size of the dienophile’s pyrazolidinone
R group. High selectivities were observed in reactions with
chiral Lewis acids capable of providing square planar metal
geometries. Cu(II) and Pd(II) complexes provided excellent
enantioselectivities when substrates with the 1-naphthylmethyl
group were employed, with Pd(II) normally being superior.
Substitution at the C-5 position of the pyrazolidinone core was
found to play a significant role in enhancing selectivity.
Reactions with common dienophiles and dienes were performed
to demonstrate the utility of pyrazolidinone substrates. Studies
indicated that π-π interactions were not an important compo-
nent of the high selectivities obtained. Some weak ligands and
non-C2-symmetric ligands were used to further demonstrate that
chiral relay is indeed operative, in which actual face shielding
is probably provided by the pyrazolidinone fluxional group
rather than the ligand itself.
Acknowledgment. Financial support for this program was
provided by the National Science Foundation (Grant NSF-CHE-
0316203). We thank Dr. Peter Daniels for assistance in solving
X-ray crystal structures, Dr. John Bagu for assistance with NMR
techniques, and Mrs. Rina Miyabe for experimental assis-
tance. We also thank Dr. Shankar Manyem for helpful discus-
sions.
Supporting Information Available: Characterization data for
compounds and experimental procedures. This material is
available free of charge via the Internet at http://pubs.acs.org.
JA066425O
9
J. AM. CHEM. SOC. VOL. 129, NO. 2, 2007 405