Memory of chirality in rebound cyclizations of α-amide radicals

Article abstract of DOI:10.1139/v2012-085

Reduction of (S)-N-(2-bromoallyl)-N-(tert-butyl)-2-methyl-3- phenylpropanamide with tributyltin hydride provides (3S,4S)-3-benzyl-1-(tert- butyl)-3,4-dimethylpyrrolidin-2-one with about 80% retention of chirality at the stereocenter adjacent to the amide carbonyl group. This memory of chirality is suggested to occur by transfer of chirality from a stereocenter to an axis, then from the axis back to a new stereocenter.

Full text of DOI:10.1139/v2012-085

1
Memory of chirality in rebound cyclizations of
␣-amide radicals
Aniruddha Sasmal, Tsuyoshi Taniguchi, Peter Wipf, and Dennis P. Curran
Abstract: Reduction of (S)-N-(2-bromoallyl)-N-(tert-butyl)-2-methyl-3-phenylpropanamide with tributyltin hydride provides
(3S,4S)-3-benzyl-1-(tert-butyl)-3,4-dimethylpyrrolidin-2-one with about 80% retention of chirality at the stereocenter
adjacent to the amide carbonyl group. This memory of chirality is suggested to occur by transfer of chirality from a
stereocenter to an axis, then from the axis back to a new stereocenter.
Key words: memory of chirality, radical reactions, amides, stereochemistry.
Résumé : La réduction du (S)-N-(2-bromoallyl)-N-(tert-butyl)-2-méthyl-3-phénylpropanamide par l’hydrure de tributylétain
conduit a la formation de la (3S,4S)-3-benzyl-1-(tert-butyl)-3,4-diméthylpyrrolidin-2-one avec 80 % de rétention de la
`
chiralité au stéréocentre adjacent du groupe carbonyle de l’amide. Il est suggéré que ce souvenir de la chiralité résulte d’un
transfert de chiralité du stéréocentre vers un axe, puis de cet axe vers un nouveau stéréocentre.
Mots-clés : souvenir de la chiralité, réactions radicalaires, amides, stéréochimie.
[Traduit par la Rédaction]
Radical reactions involve open-shell intermediates and
therefore tend to occur much more rapidly than ionic and
pericylic reactions involving closed-shell species.1 Indeed,
many radical reactions can occur on timescales comparable to
or faster than common conformational changes of organic
molecules, including simple bond rotations. This speed makes
radicals attractive for reactions that involve transient confor-
mational chirality.
Depending on the overall transformation, reactions with
transiently chiral intermediates can exhibit transfer of chirality
or memory of chirality (MOC).2 An example of each reaction
type is shown in Fig. 1. Cyclizations of o-haloanilides are
typified by the reductive cyclization of 1 to give 3 via the
intermediacy of axially chiral radical 2.3 Levels of transfer of
axial chirality from 1 to the new stereocenter in 3 are high
even when the ortho substituent R is hydrogen.3b Simple
rotation of the anilide N–Ar bond is a racemization reaction
that would erase the chirality transfer, so clearly cyclization of
2 is faster than this rotation.
Memory of chirality occurs at a stereocenter when a
formal substitution reaction proceeds stereoselectively even
though the reaction passes through a trigonal inter-
mediate.2a Rychnovsky and co-workers4 have shown that
radical reactions can be faster than ring interconversions,
while Giese and co-workers5 and Bertrand and co-workers6
have shown that onward reactions of diradicals (for example,
coupling) can compete effectively with simple ␴-bond rota-
tions. Bonjoch and co-workers7 discovered intriguing exam-
ples of MOC in monoradical reactions that involved hydrogen
transfer and cyclization of ␣-aminyl radicals.
In an example from Bertrand and co-workers,6b heating of 4
provides 6 with a high degree of retention of configuration at
the stereocenter bearing the carbomethoxy (CO2Me) group.
This reaction must involve an axially chiral intermediate like
5 that is generated by enediyne cyclization and 1,6-hydrogen
atom transfer. Onward reaction of ␣-aminyl radical 5 to form
the product must be faster than CH2–N bond rotation. In other
words, the MOC results from a back-to-back relay of chirality
transfer reactions: from a stereocenter to an axis, then from the
axis back to a new stereocenter.
We wondered whether the electron-donating nitrogen sub-
stituent is indispensible in such types of MOC reactions.
Indeed, there is some old circumstantial evidence that it is not.
In 1968, Heiba and Dessau8 reported that the atom transfer
addition of chloroform to enantioenriched 7 gave a lactone
product (8) with a small residual optical rotation. The enan-
tiomeric excesses (ees) were not known for either 7 or 8, so the
significance of the residual rotation is unclear.
This transformation presumably involves addition of
Cl3C• to 7 followed by ester C–O bond rotation to give 9
and radical translocation9 by 1,5-hydrogen atom transfer to
give 10. MOC will occur if the 5-exo cyclization of 10 is faster
than rotation of the ␴ bonds to the carbonyl carbon and net
retention is expected. This sequence of translocation and
rebound cyclization is reminiscent of both Bertrand and
co-workers’6 and Bonjoch and co-workers’7 MOC processes.
Received 14 June 2012. Accepted 13 September 2012. Published at www.nrcresearchpress.com/cjc on 29 November 2012.
A. Sasmal, T. Taniguchi, P. Wipf, and D.P. Curran. Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260, USA.
Corresponding author: Dennis P. Curran (e-mail: curran@pitt.edu).
This article is part of a Special Issue dedicated to Professor Derrick Clive in recognition of his contributions to organic chemistry, especially
radical reactions.
Can. J. Chem. 90: 1–5 (2012)
doi:10.1139/v2012-085
Published by NRC Research Press

2
Can. J. Chem. Vol. 90, 2012
Fig. 1. Examples of radical intermediates in (a) transfer of chirality and (b) memory of chirality.
(a) Chirality transfer (from axis to stereocenter)
O
O
O
R'
I
NR'
R'
•
Bu3SnH
25ºC
N
N
R
R
R
3
1, R = H, Me, TMS...
R' = Me, CH2Ph...
2, cyclization faster
than N–Ar rotation
(b) Memory of chirality transfer (at C bearing CO2Me)
Ts
Ts
heat
H
N
O
MeO2C
H
N
MeO2C
O
4
5, CH2–Ar rotation and ring
closure are faster than CH2–N rotation
Ts
CO2Me
O
6
The memory effect can be spoiled by rotation of either the C–C
bond to the carbonyl group of 10 (racemization results) or the
C–O bond (an unfavorable geometry for cyclization results).
At the core of this postulate is the notion that ester-
substituted radicals like 10 resemble ␣-substituted acrylates
and related molecules.10 Despite conjugation, their ground-
state geometries are twisted owing to unfavorable steric inter-
actions in the otherwise favored planar forms.11 This is
important because the planar form of 10 is achiral. In short, the
assumption that 8 must inevitably be racemic (because a
trigonal intermediate is involved) is not on solid footing.
To start, we reproduced the cyclization of enantiopure 7 and
confirmed that the small rotation of 8 is not an artifact.12 But
we could not easily determine the enantiomeric composition of
8. It decomposed on GC injection, and the use of shift reagents
was complicated because 8 is already a mixture of diastereom-
ers (Supplementary data).
Instead of trying to solve the analytical problems with 8, we
retooled the precursor in three ways. First, we replaced the
ester with an amide. One of the two amide substituents is
always well-positioned for both 1,5-hydrogen transfer and
subsequent cyclization. This eliminates the problem of the
unfavorable rotation trans geometry at the ester C–O bond of
7. Second, the addition reaction of Cl3C• is not an essential
component of the process, so we designed it out, leaving an
alkenyl halide as a direct precursor of the alkenyl radical
needed for 1,5-hydrogen abstraction. Third, we removed the
geminal methyl groups of 7. This gives precursors 11a–11c
shown in the lower part of Fig. 2.
The first substrate studied was amide 11a, which exists as a
roughly 1:1 mixture of E/Z amide rotamers13 (structure 11a
shows the favorable rotamer for MOC chemistry). The study
of this compound proved to be a wrong turn, though we prefer
to regard it as a short scenic detour. Reduction of racemic 11a
with tributyltin hydride in refluxing toluene provided a crude
product that did not apparently contain either the target MOC
product or the directly reduced product. Instead, the major
product was 12, which was isolated in a 50% yield by flash
chromatography.
Compound 12 is a rearranged product of 1,4-phenyl transfer
that presumably forms by ipso cyclization of 13 to give 14
followed by fragmentation, as shown in Scheme 1. The mech-
anism was supported by a reduction of 11a with Bu3SnD,
which gave 12 with significant deuterium labeling on the
N–CH3 group. Related aryl transfer reactions are well-
known,14 although none apparently with this structural motif.
With optimization, this reaction could be useful for preparing
␤-aryl allyl amides and related molecules.
To eliminate competing phenyl migration, we targeted pre-
cursors 11b and 11c bearing N-tert-butyl groups in place of the
N-benzyl group of 11c. Precursors 11b and 11c exist as one
major rotamer in solution,13b which has the tert-butyl group
Published by NRC Research Press

Sasmal et al.
3
Fig. 2. Top: Could memory of chirality have occurred in this
old reaction of an ␣-chiral ester? Bottom: A retooled amide
11 eliminates potential problems.
cis to the amide oxygen atom, as drawn in Scheme 2. These
precursors were synthesized in racemic and enantiopure
forms,12 and the results of key reduction experiments are
summarized in Scheme 2. Direct reductive debromination of
these precursors was a consistent problem, and higher yields of
the target cyclized products (15) were obtained at higher temper-
atures. We settled on standard conditions involving syringe
pump addition of Bu3SnH and 2,2=-azobisisobutyronitrile
(AIBN) in toluene to a refluxing solution of the precursor
in toluene. The target products were isolated by flash chro-
matography, then analyzed by spectroscopy and chiral
GC.
The reduction of (rac)-11b in this way provided about a 1:1
mixture of diastereomers 15b in a 39% isolated yield. We
could not separate these diastereomers, nor could we make a
clear cis/trans assignment. Nonetheless, the sample of (rac)-
15b exhibited four peaks on chiral GC, showing that both pairs
of enantiomers were resolved. Cyclization of (S)-11b (Ͼ99:1
enantiomeric ratio (er)) provided the same ratio of diastereom-
ers 15b in about the same isolated yield, but now both diaste-
reomers were significantly enantioenriched to the extent of
about 84:16.
MOC with 7 is conceivable
O
O
CCl4
Me
O
Cl
Cl
O
(PhCO2)2
heat
Et
7, +20.6
1) add Cl3C•
Me
8, +0.9
Et
1) 5-exo
2) eliminate
Cl•
MOC
2) rotate C–O
O
O
O
O
Me
Et
1,5-H
Me
Et
H
H
transfer
Cl3C
Cl3C
While we learned the enantiomer ratios of 15b, it was not
easy to preparatively separate the diastereomers or to assign
their configurations in the mixture. To simplify the assignment
problem, we selected 11c with the hypothesis that one diaste-
reomeric radical resulting from cyclization would cyclize
again, whereas the other would not.15 Tin hydride reduction of
(rac)-11c provided a mixture of products including the product
of direct reductive debromination (about a 30% yield), a small
amount of an isomeric cyclized product (Ͻ5%),16 and the
target products (rac)-15c and tricycle 16. The products were
partially resolved by chromatography; the reported yields
(16% and 9%) are from pure fractions.
In contrast with 15b, 15c is a single diastereomer, which we
assign as having the vicinal ␣-benzyl and ␤-methyl groups
trans. The cis isomer was not detected in the crude product
mixture. We believe that the cyclization of the translocated
radical gives both isomers, one of which is reduced and the
other of which cyclizes (see the following). This is why the
yields of 15c (trans only) are about half of the yields of 15b
(cis/trans mixture).
9
10, rotation
spoils MOC
retooled amides 11 based on the intermediacy of 9
O
Me
R
Br
R1
11a, R = R1 = CH2Ph
N
11b, R = Et, R1 = tBu
11c, R = CH2Ph, R1 = tBu
1) direct radical generation from a bromide
2) amide favors H-transfer and cyclization
3) remove allylic Me groups
Scheme 1. An interesting phenyl transfer reaction provides a short
Reductions of (S)-11c and (R)-11c provided the products
15c in similar yields to the racemate, but now significantly
enantioenriched. The ratio of enantiomers (S)-15c/(R)-15c was
78:22 starting from (S)-11c and 21:79 starting from (R)-11c.
These values of chirality transfer are comparable to those seen
in the diastereomers of 15b.
With a diastereopure sample now available, the assignment
of absolute configuration was accomplished by experimental
and computational rotation methods. The measured specific
rotations of 15c are ϩ44.6 and –44.3 (c 0.25, CHCl3). The
specific rotation of (S,S)-15c was calculated by established
techniques to be ϩ125 in CHCl3.17 From the measured rota-
tion of enantioenriched samples, the rotations of enantiopure
samples of 15c are about ϩ77 and –77, respectively. While the
values of measured and calculated rotations differ somewhat,
it seems clear because of the large magnitudes involved that
the dextrorotatory enantiomer has the (S,S) configuration. This
means that the conversion of 11c to 15c (and by analogy 11b
scenic detour.
O
O
Bu3SnH
AIBN
CH2Ph
Br
CH3
Me
Me
N
N
Ph
PhCH2
PhCH2
toluene
reflux
11a
Bu3Sn•
12, 50%
fragmentation
O
O
CH2Ph
Me
Me
ipso
N
N
PhCH2
PhCH2
cyclization
13
14
Published by NRC Research Press

4
Can. J. Chem. Vol. 90, 2012
Scheme 2. Memory of chirality in reductions of 11b and 11c.
O
Bu3SnH
AIBN
tBu
O
H3C
H3C
CH3CH2
H3C
N
t
Br
N Bu
CH3CH2
toluene
reflux
(rac)-11b
(rac)-15b, 39%, 1:1 cis:trans
(S)-11b, >99:1 er
15b, 38%, both diastereomers
84:16 er
O
O
H3C
Bu3SnH
AIBN
tBu
N
O
H3C
PhCH2
H3C
t
+
N Bu
t
PhCH2
Br
N Bu
H3C
toluene
reflux
(rac)-11c
(S)-11c, 99:1 er (shown)
(rac)-15c, 16%, trans only
(S,S)-15c, 18%, 78:22 er
(rac)-16, 9%
(R)-11c, >99:1 er (not shown) (R,R)-15c, 18%, 21:79 er
Fig. 3. Proposed pathways in the reduction of 11c.
O
O
O
tBu
tBu
tBu
PhCH2
H3C
Bu3Sn•
1,5-H
H3C
PhCH2
H3C
N
N
N
H
transfer
Br
H
PhCH2
17
18
(S)-11c
O
O
H3C
PhCH2
•H2C
H3C
rebound,
retention
+
t
t
N Bu
N Bu
•
19-trans,
reduced to (S,S)-15c
19-cis,
removed by cyclization
to 15b) occurs predominately (about 80%) with the retention
of configuration.
group undergo a sequence of radical reactions whose key steps
are radical translocation (by 1,5-hydrogen transfer) and rebound
5-exo cyclization of the resulting ␣-amidyl radical to the resulting
alkene. Even though a trigonal radical intermediate is formed, the
␥-lactam product is produced with about 80% retention of con-
figuration. This MOC is suggested to occur by transfer of chiral-
ity from a stereocenter to an axis, then from the axis back to a new
stereocenter. The work suggests that generation of ␣-amide and
related ␣-carbonyl (␣-ester, ␣-keto, and so on) radicals by radical
translocation might be a general route to reactions involving
MOC of preformed tert-␣-stereocenters in such precursors.
Figure 3 shows suggested pathways for the reduction of
(S)-11c. Bromine abstraction provides alkenyl radical 17,
which in turn abstracts a hydrogen atom adjacent to the car-
bonyl group. Because of geometric requirements, the ab-
stracted hydrogen is not aligned with the carbonyl ␲ orbitals
but is instead positioned roughly anti to the carbonyl group.
This reaction erases the stereocenter but provides axially chiral
radical 18. Rebound cyclization of 18 to the alkene completes
the MOC process, which must occur with retention. Now the
diasteromeric products 19-trans and 19-cis take different path-
ways with 19-trans being reduced as usual. However, 19-cis is
derailed by cyclization to the aryl ring, thereby simplifying the
product analysis.
Supplementary data
Supplementary data are available with the article through
the journal Web site http://nrcresearchpress.com/doi/suppl/
10.1139/v2012-085.
In summary, N-(2-bromo-2-propenyl)-N=-tert-butyl amides
bearing a tertiary stereocenter adjacent to the amide carbonyl
Published by NRC Research Press

Sasmal et al.
5
(10) (a) Porter, N. A. In Radicals in Organic Synthesis, 1st ed.;
Renaud, P., Sibi, M. P., Eds.; Wiley-VCH: Weinheim, 2001;
Vol. 1, pp 416–440; (b) Porter, N. A.; Giese, B.; Curran, D. P.
Acc. Chem. Res. 1991, 24 (10), 296. doi:10.1021/ar00010a003.
(11) (a) Simakov, P. A.; Martinez, F. N.; Horner, J. H.; Newcomb,
M. J. Org. Chem. 1998, 63 (4), 1226. doi:10.1021/jo971774ϩ;
(b) Musa, O. M.; Choi, S. Y.; Horner, J. H.; Newcomb, M.
J. Org. Chem. 1998, 63 (3), 786. doi:10.1021/jo9717907.
(12) Experimental, computational, and compound characterization
data (except for compound 16) can found in the thesis of A.
Sasmal, University of Pittsburgh, 2011: open access at http://
d-scholarship.pitt.edu/10754/ (see the Supplementary data).
Data for 16: IR (thin firm, cm–1) ␯max: 1670;.1H NMR
(300 MHz, CDCl3) ␦: 1.15 (9H, s), 1.24 (3H, s), 2.29Ϫ2.37 (1H,
m), 2.43 (1H, d, J ϭ 14.4 Hz), 2.54 (1H, dd, J ϭ 14.4, 4.5 Hz),
2.74 (1H, dd, J ϭ 9.9, 4.5 Hz), 2.83 (1H, dd, J ϭ 14.4, 5.1 Hz),
2.93 (1H, d, J ϭ 14.4 Hz), 3.54 (1H, t, J ϭ 9.6 Hz), 7.11Ϫ7.15
(4H, m). 13C NMR (100 MHz, CDCl3) ␦: 25.6, 27.2, 33.7, 36.8,
38.3, 46.8, 49.0, 53.6, 126.2, 126.5, 127.5, 127.9, 136.4, 137.6,
178.4. HR-MS (ESI) calcd for C17H23NONa ([M ϩ Na]ϩ):
280.1677; found: 280.1681.
(13) (a) Oki, M. The Chemistry of Rotational Isomers; Springer-
Verlag: New York, 1993; (b) Stewart, W. E.; Siddall, T. H.
Chem. Rev. 1970, 70 (5), 517. doi:10.1021/cr60267a001.
(14) (a) Bowman, W. R.; Storey, J. M. D. Chem. Soc. Rev. 2007, 36
(11), 1803. doi:10.1039/b605183a; (b) Studer, A.; Bossart, M.
Tetrahedron 2001, 57 (48), 9649. doi:10.1016/S0040-
4020(01)00990-5; (c) Studer, A. In Radicals in Organic Syn-
thesis, 1st ed.; Renaud, P., Sibi, M. P., Eds.; Wiley-VCH:
Weinheim, 2001; Vol. 2, pp 44–60.
(15) Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of
Radical Reactions: Concepts, Guidelines, and Synthetic
Applications; VCH: Weinheim, 1996.
(16) This tentatively assigned product results from 1,6-transfer of a
benzylic hydrogen atom to the initial alkenyl radical, followed
by 6-exo cyclization of the resulting benzyl radical.
(17) (a) Mukhopadhyay, P.; Wipf, P.; Beratan, D. N. Acc. Chem.
Res. 2009, 42 (6), 809. doi:10.1021/ar8002859; (b) Kondru,
R. K.; Chen, C. H.-T.; Curran, D. P.; Beratan, D. N.; Wipf, P.
Tetrahedron Asymmetry 1999, 10 (21), 4143. doi:10.1016/
S0957-4166(99)00443-7; (c) Kondru, R. K.; Wipf, P.; Beratan,
D. N. Science 1998, 282 (5397), 2247. doi:10.1126/
science.282.5397.2247.
Acknowledgements
We thank the National Institutes of Health for funding of
this work.
References
(1) Newcomb, M. In Radicals in Organic Synthesis, 1st ed.;
Renaud, P., Sibi, M., Eds.; Wiley-VCH: Weinheim, 2001; Vol.
1, pp 317–336.
(2) (a) Zhao, H. W.; Hsu, D. C.; Carlier, P. R. Synthesis 2005, 1
doi:10.1055/s-2004-834931; (b) Kawabata, T.; Fuji, K. Top.
Stereochem. 2003, 23, 175. doi:10.1002/0471224499.ch3.
(3) (a) Shirakawa, S.; Liu, K.; Maruoka, K. J. Am. Chem. Soc. 2012,
134 (2), 916. doi:10.1021/ja211069f; (b) Petit, M.; Lapierre,
A. J. B.; Curran, D. P. J. Am. Chem. Soc. 2005, 127 (43), 14994.
doi:10.1021/ja055666d; (c) Petit, M.; Geib, S. J.; Curran, D. P.
Tetrahedron 2004, 60 (35), 7543. doi:10.1016/j.tet.2004.05.116;
(d) Curran, D. P.; Liu, W. D.; Chen, C. H.-T. J. Am. Chem. Soc.
1999, 121 (47), 11012. doi:10.1021/ja993329x.
(4) (a) Dalgard, J. E.; Rychnovsky, S. D. Org. Lett. 2004, 6 (16),
2713. doi:10.1021/ol049038x; (b) Buckmelter, A. J.; Kim, A. I.;
Rychnovsky, S. D. J. Am. Chem. Soc. 2000, 122 (39), 9386.
doi:10.1021/ja002068k.
(5) (a) Sinicropi, A.; Barbosa, F.; Basosi, R.; Giese, B.; Olivucci,
M. Angew. Chem. Int. Ed. 2005, 44 (16), 2390. doi:10.1002/
anie.200461898; (b) Giese, B.; Wettstein, P.; Stahelin, C.; Bar-
bosa, F.; Neuburger, M.; Zehnder, M.; Wessig, P. Angew.
Chem. Int. Ed. 1999, 38 (17), 2586. doi:10.1002/(SICI)1521-
3773(19990903)38:17Ͻ2586::AID-ANIE2586Ͼ3.0.CO;2-K.
(6) (a) Mondal, S.; Nechab, M.; Vanthuyne, N.; Bertrand, M. P. Chem.
Commun. (Camb.) 2012, 48. doi:10.1039C2CC17830C; (b) Nechab,
M.; Campolo, D.; Maury, J.; Perfetti, P.; Vanthuyne, N.; Siri, D.;
Bertrand, M. P. J. Am. Chem. Soc. 2010, 132 (42), 14742. doi:10.1021/
ja106668d.
(7) Quirante, J.; Diaba, F.; Vila, X.; Bonjoch, J.; Lago, E.; Molins,
E. C. R. Acad. Sci. Paris. Chim. 2001, 4, 513. doi:10.1016/
S1387-1609(01)01261-0.
(8) Heiba, E.-A. I.; Dessau, R. M. J. Am. Chem. Soc. 1967, 89 (9),
2238. doi:10.1021/ja00985a050.
(9) (a) Robertson, J.; Pillai, J.; Lush, R. K. Chem. Soc. Rev. 2001,
30 (2), 94. doi:10.1039/b000705f; (b) Curran, D. P.; Kim, D.;
Liu, H. T.; Shen, W. J. Am. Chem. Soc. 1988, 110 (17), 5900.
doi:10.1021/ja00225a052.
Published by NRC Research Press