1,5-Asymmetric Induction in Squarate Cascades. Conformational Control of Helicity by Chiral Amino Substituants during Conrotatory Octatetraene Cyclization Prior to β-Elimination

Article abstract of DOI:10.1021/jo9722921

Both the sigmatropic and electrocyclic rearrangement pathways that can arise when a pair of alkenyl anions are added to a squarate ester have high stereochemical demands. The distinction is nontrivial. When cis addition occurs initially, the stereoinduction that materializes at this point is fully transmitted into the product (s). The more commonly observed trans addition exhibits fleeting stereochemical consequences because of rapid equilibration of the octatetraenyl intermediates. In this instance, product distribution is governed by the relative rates of conrotatory cyclization at this advanced stage. Herein reported is a complete dissection of a squarate cascade when a stereogenic center attached to an amino substituent effects 1,5-asymmetric induction prior to β-elimination of the entire fragment. Deuterium labeling permits a direct measure of the contrasting kinetic imbalances associated with the two possible modes of alkenyl anion addition. Furthermore, quantitative analysis of the partitioning experienced by the two helical octatetraenes is readily accomplished. This work constitutes the first example where a complete dynamic profile for these complex processes has been possible. The fact that long-range asymmetric induction has been instrumental in solving the mechanistic puzzle is noteworthy.

Full text of DOI:10.1021/jo9722921

2022
J . Org. Chem. 1998, 63, 2022-2030
1,5-Asym m etr ic In d u ction in Squ a r a te Ca sca d es. Con for m a tion a l
Con tr ol of Helicity by Ch ir a l Am in o Su bstitu en ts d u r in g
Con r ota tor y Octa tetr a en e Cycliza tion P r ior to â-Elim in a tion
Leo A. Paquette* and J insung Tae
Evans Chemical Laboratories, The Ohio State University, Columbus, Ohio 43210
Received December 18, 1997
Both the sigmatropic and electrocyclic rearrangement pathways that can arise when a pair of alkenyl
anions are added to a squarate ester have high stereochemical demands. The distinction is
nontrivial. When cis addition occurs initially, the stereoinduction that materializes at this point
is fully transmitted into the product(s). The more commonly observed trans addition exhibits fleeting
stereochemical consequences because of rapid equilibration of the octatetraenyl intermediates. In
this instance, product distribution is governed by the relative rates of conrotatory cyclization at
this advanced stage. Herein reported is a complete dissection of a squarate cascade when a
stereogenic center attached to an amino substituent effects 1,5-asymmetric induction prior to
â-elimination of the entire fragment. Deuterium labeling permits a direct measure of the contrasting
kinetic imbalances associated with the two possible modes of alkenyl anion addition. Furthermore,
quantitative analysis of the partitioning experienced by the two helical octatetraenes is readily
accomplished. This work constitutes the first example where a complete dynamic profile for these
complex processes has been possible. The fact that long-range asymmetric induction has been
instrumental in solving the mechanistic puzzle is noteworthy.
Squarate ester cascades are now well-recognized to be
polyolefinic intermediates to experience mutual equili-
bration. The stereochemically determining step is linked
to the rates of 8π conrotatory closure of these helices, with
that coil experiencing the minimal steric impedance to
cyclization giving rise to the major (or exclusive) product.
There are at least two interesting ways of potentially
controlling the outcome of the latter mechanistic pathway
in systematic fashion. The first is to position a stereo-
differentiating R′ group proximal to the octatetraene
bonding sites as in 1. For 1a , the R′ group finds itself
intercalated inside the spiral instead of being positioned
on the exterior as in 1b. The steric congestion present
powerful synthetic transforms for the rapid construction
of complex polycyclic molecules.1 Two variants are
known, both having the formidable potential for generat-
ing new C-C bonds and up to five new stereogenic
centers in a single step. As an added benefit, the highly
functionalized products are amenable to further chemical
change,2 thereby providing very convenient access to
additional structurally intricate substances.
When the two alkenyllithium reagents add in cis
fashion to the squarate, a pathway not generally observed
unless chelation control operates,3-6 product stereochem-
istry is set at this very early step. This is because the
ensuing dianionic oxy-Cope rearrangement very reliably
transmits the original stereoselectivity into the structural
scaffolding of the final product(s).
Trans introduction of the alkenyl anions, the custom-
arily dominant reaction channel because of its associated
steric and electrostatic advantages, is quickly followed
by conrotatory 4π ring opening of the cyclobutene di-
alkoxide to a doubly charged octatetraenyl intermediate.1
The stereochemical information resident within the first-
formed dialkoxides is subsequently lost because of the
remarkable ability of these helical (and therefore chiral)
(1) (a) Negri, J . T.; Morwick, T.; Doyon, J .; Wilson, P. D.; Hickey,
E. R.; Paquette, L. A. J . Am. Chem. Soc. 1993, 115, 12189. (b) Paquette,
L. A.; Morwick, T. J . Am. Chem. Soc. 1995, 117, 1451. (c) Paquette, L.
A.; Doyon, J . J . Am. Chem. Soc. 1995, 117, 6799. (d) Paquette, L. A.;
Morwick, T. J . Am. Chem. Soc. 1997, 119, 1230. (e) Paquette, L. A.;
Doyon, J . J . Org. Chem. 1997, 62, 1723.
(2) Morwick, T. M.; Paquette, L. A. J . Org. Chem. 1996, 61, 146.
(3) Paquette, L. A.; Morwick, T. M.; Negri, J . T.; Rogers, R. D.
Tetrahedron 1996, 52, 3075.
(4) Paquette, L. A.; Kuo, L. H.; Doyon, J . Tetrahedron 1996, 52,
11625.
(5) Paquette, L. A.; Kuo, L. H.; Doyon, J . J . Am. Chem. Soc. 1997,
119, 3038.
(6) Paquette, L. A.; Kuo, L. H.; Tae, J . J . Org. Chem. 1998, 63, 2010
(preceding paper in this issue).
in 1a deters formation of 2. Since 1b is not similarly
S0022-3263(97)02292-5 CCC: $15.00 © 1998 American Chemical Society
Published on Web 02/27/1998

1,5-Asymmetric Induction in Squarate Cascades
J . Org. Chem., Vol. 63, No. 6, 1998 2023
crowded, closure to give 3 operates exclusively in many
cases.7,8 In the final analysis, it is the stereogenic center
substituted by R′ that is controlling all the others that
materialize in the eventual polycyclic product(s). When
the R′ group resides at position a in 1, 1,2-asymmetric
induction has been found to operate at a very high level.
The alternative bonding of R′ to site b (now an example
of 1,3-stereoinduction) has understandably proven to be
less effective in exhibiting chirality transfer since the
substituent finds itself in a more distal relationship to
the trigonal centers undergoing bonding.
A more subtle tactic is to introduce a stereogenic atom
in an equally proximal location to the cyclization termini,
as exemplified in 4. The expectation is that the config-
uration at the chiral center would impact on the compet-
ing rates of closure of 4a and 4b. Although the distance
of the newly formed methine carbon in 6 and 7 would
perforce represent a direct measure of this kinetic imbal-
ance, both in magnitude and in direction. Since the
transmission of chirality is necessarily a process involving
long-range 1,5-asymmetric induction, moderate stereo-
selectivity should be forthcoming because of the concert-
edness of the ring closure and the stereoalignment
demanded of these polyolefins in the cyclization transi-
tion states.10
A notably useful aspect of this particular mechanistic
cascade is the expectation that 6 and 7, once formed,
would undergo â-elimination of the chiral amide anion.6
Such an event would result in the formation of 8 and ent-
8, and culminate in their regiodirected transannular
aldolization to give the enantiomers of a single polycyclic
ketone. Chiral HPLC analysis for the determination of
percent ee and knowledge of absolute configuration would
constitute the tools necessary to realize a particularly
stringent test of π-facial discrimination during stereo-
controlled 8π bonding within octatetraenyl helices.
Resu lts a n d Discu ssion
Syn th etic Con sid er a tion s. In studies preliminary
to this investigation, observations were made that the
â-elimination of amino groups was not universally opera-
tive but notably dependent on the nature of the compan-
ion alkenyl anion.6 Particularly relevant was the finding
that the use of 5-lithiodihydrofuran as one of the nucleo-
philic reagents invariably resulted in expulsion of the
amide functionality. The conversion of 9 to 10 via two
independent (although related) means illustrates the
point.
factor is allylic in its magnitude, many examples of high
π-facial stereoselectivity have been reported for com-
pounds of this general type, particularly when highly
ordered six-ring transition states are involved.9
We now report examples in which the stereogenic
center is still further removed from the octatetraenyl
framework and bonded to nitrogen as in 5. Competitive
conrotation within either of the diastereomeric coiled
conformations labeled as 5a and 5b can be expected to
proceed at nonidentical rates. The absolute configuration
As a consequence, variants of this reaction were
targeted for detailed investigation. The requisite opti-
cally active amines were selected from various classes,
viz. acyclic, cyclic, and heterocyclic, including a C2-
symmetric example. In those cases where the starting
amine was primary, a methyl group was introduced prior
to reaction with 2,3-dibromopropene, as shown in Scheme
1. The known secondary amines were allylated directly.
In those examples where steric hindrance was present,
a change in reaction conditions from triethylamine in
(7) Paquette, L. A.; Hamme, A. T., II.; Kuo, L. H.; Doyon, J .;
Kreuzholz, R. J . Am. Chem. Soc. 1997, 119, 1242.
(8) Paquette, L. A., Kuo, L. H.; Hamme, A. T., II.; Kreuzholz, R.;
Doyon, J . J . Org. Chem. 1997, 62, 1730.
(9) Procter, G. Asymmetric Synthesis; Oxford University Press:
Oxford, England, 1996.
(10) For an analysis of conformational transmission of chirality in
the specific case of a 1,4-asymmetric induction process, consult: Lucero,
M. J .; Houk, K. N. J . Am. Chem. Soc. 1997, 119, 826.

2024 J . Org. Chem., Vol. 63, No. 6, 1998
Paquette and Tae
Sch em e 2
condition A will necessarily generate 20 and 21 in
unequal amounts. The overall response requires that
this stereoinduction be fully transmitted into the product
mixture. Reversal of the addition sequence as under
condition B gives rise only to racemic 22, since nucleo-
philic attack at either carbonyl group from either direc-
tion must occur at equal rates. Once 22 has been
generated, there remains no opportunity for the enan-
tiopure allylic amine to exert any stereochemical bias at
all during cis addition.
F igu r e 1. The enantiopure bromo amines utilized in this
investigation.
Sch em e 1
On the other hand, trans 1,2-addition to 20-22 carries
only fleeting stereochemical consequences, since the
asymmetric step materializes only later at the time of
conrotatory octatetraene cyclization. Although the latter
scenario is of immediate interest here, the relative kinetic
ordering associated with the formation of 20 and 21 offers
an equally unrivaled opportunity for gaining new per-
spectives on carbonyl addition processes.
The experimental data are given in Table 1. Entries
3, 4, 7, and 8 are characterized by notably low levels of
chirality transmission. Conversely, the most stereochem-
ically responsive chiral auxiliaries are found in entries
1, 5, 6, and 9. It is significant that replacement of the
N-methyl group in 11 (entry 1) by a more bulky benzyl
substituent (entry 4) results in increased reaction ef-
ficiency accompanied by a substantial dropoff in enan-
tioselectivity. Replacement of the phenyl group in 11 by
an R-naphthyl (entry 2) has a similar, although less
dramatic, outcome. The cyclohexyl analogue of 17 (entry
7) is still less suited to our purposes. The endo-borny-
lamine 13 (entry 3) is inferior to its pinanyl counterpart
16 (entry 6). The presence of an added chelation site
(entry 5) is beneficial and the C2-symmetric example 19
(entry 9) exhibits a useful product distribution.
CH2Cl2 to the more harsh potassium carbonate in reflux-
ing DMF was mandated.
The nine reagents generated in this manner are
grouped in Figure 1. All of the starting amines have been
previously described in enantiopure condition,11 thus
facilitating access to 11-19.
Execu tion of th e Rea r r a n gem en t Ca sca d es. It is
necessary to realize that two lines of mechanistic inquiry
are being investigated simultaneously. If a lithiated
amine is added first (condition A), the extent of cis
addition of the second alkenyl anion is certain to be
different than that observed if the lithiated dihydrofuran
is the lead reagent (condition B). No matter how the cis/
trans ratio of the cyclobutene dialkoxides might be
partitioned, the outcome of the ensuing cascade should
be rationalizable in terms of the associated partitioning
between the [3,3] sigmatropic and stepwise electrocyclic
options. The distinction is nontrivial. For the anionic
oxy-Cope alternative, the key asymmetric step is the
addition of the homochiral nitrogen-containing vinyl
anion to the squarate ester. As depicted in Scheme 2,
The crossovers in enantioselectivity observed between
conditions A and B in entries 3 and 4 are fascinating and
warrant explanation. The initial addition of 5-lithiodi-
hydrofuran to generate 22 is predictably recognized to
foster a modest level of chelation control as the second
alkenyl anion is introduced.3 Accordingly, the product
compositions generated under these circumstances will
often contain higher levels of oxy-Cope product. When
the tricyclic ketones resulting from the sigmatropic and
electrocyclic cascade exhibit optical rotations of different
sign, the observable effect will be manifested in this way.
(11) The primary amines required for the preparation of 11, 12, 16,
17, and 18 as well as the secondary amine precursor to 14 were
purchased from the Aldrich Chemical Co. endo-Bornylamine was
prepared from D-camphor (Forster, M. O. J . Chem. Soc. 1898, 73, 386),
(S)-(+)-2-(methoxymethyl)pyrrolidine from (S)-proline (Enders, D.; Fey,
P.; Kipphardt, H. Org. Synth. 1987, 65, 173), and (2S,5S)-2,5-
dimethylpyrrolidine from L-alanine (Yamazaki, T.; Gimi, R.; Welch,
J . T. Synlett 1991, 573).
Deter m in a tion of Absolu te Ster eoch em istr y. The
absolute stereochemistry of 10 was determined by an

1,5-Asymmetric Induction in Squarate Cascades
J . Org. Chem., Vol. 63, No. 6, 1998 2025
Ta ble 1. Eva lu a tion of Ch ir a l Su bstitu en ts a t Nitr ogen
on th e En a n tioselectivity of a Squ a r ea te Ester Ca sca d e
Sch em e 3
the form of “extended Newman projections”. Comparison
of the chemical shifts of the methoxy groups in these
derivatives shows those in A to be more upfield. This
effect can be attributed to shielding experienced as the
result of spatial proximity to the phenyl substituent,
thereby requiring it to be the S,S isomer.
Determination of the percent ee values was facilitated
by the fact that 10 and ent-10 exhibit widely divergent
retention times on a CHIRALPAK AD column in a
solvent system constituted of isopropyl alcohol and hex-
anes (1:2). The values reported in Table 1 were arrived
at by chiral HPLC analysis in this fashion.
a
Generated from 11-19 by halogen-metal exchange with tert-
b
butyllithium in THF at -78 °C. A: addition of the amino-
substituted reagent precedes introduction of 5-lithiodihydrofuran.
B: the reverse of A.
established NMR method.12 Initially, its oxygenated ring
was reduced with lithium aluminum hydride at low
temperature to give principally the demethoxylated
R-alcohol 23 (Scheme 3). The appreciable deshielding of
the carbinol proton in 23 (δ 4.79) relative to that
experienced by the epimeric proton in 24 (δ 4.14), which
can be directly attributed to proximity to the ethereal
oxygen, defines the relative configuration of the reduction
products. Coupling of 23 to (S)-O-methylmandelic acid
furnished the diastereomeric esters 25 and 26, which
could be separated chromatographically. Structures A
and B illustrate the Mosher models13 of these esters in
Deu ter iu m La belin g Stu d ies. Distinction between
the operation of either the oxy-Cope or the electrocyclic
cascade requires that the lithiated allylamine carry an
appropriate stereochemical marker. Since the smallest
structural change that can be implemented is to substi-
tute deuterium for hydrogen, our attention was turned
to the preparation of 27 (Scheme 4). To this end,
propargyl bromide was subjected to bromoboration with
B-bromo-9-borabicyclo[3.3.1]nonane in anticipation of
sterically directed Markovnikov addition.14 Deuteriolysis
of the vinylborane so formed with acetic acid-O-d afforded
regio- and stereoselectively labeled 2,3-dibromopropene-
1-d, from which 27 was generated by conventional
(12) (a) Trost, B. M.; Curran, D. P. Tetrahedron Lett. 1981, 22, 4929.
(b) Trost, B. M.; Belletire, J . L.; Godleski, S.; McDougal, P. G.; Balkovec,
J . M.; Baldwin, J . J .; Christy, M. E.; Ponticello, G. S.; Varga, S. L.;
Springer, J . P. J . Org. Chem. 1986, 51, 2370.
1
displacement. While H NMR analysis showed the level
(14) Hara, S.; Dojo, H.; Takinami, S.; Suzuki, A. Tetrahedron Lett.
1983, 24, 731.
(13) Dale, J . A.; Mosher, H. S. J . Am. Chem. Soc. 1973, 95, 512.

2026 J . Org. Chem., Vol. 63, No. 6, 1998
Paquette and Tae
Sch em e 4
trogen inversion is impeded by a relatively high barrier
(∆Gq ) 7.3 kcal/mol). Conformational interconversion
is slightly less demanding (∆Gq ) 6.4 kcal/mol), with the
most stable spatial arrangement exhibiting a gauche
relationship between the nitrogen lone pair and the
R-hydrogen at the stereogenic center. As seen in the
Newman projection, the large groups are positioned anti
to each other. Should the same requirements logically
apply to 30 and 31, the helical orientation in 30 is seen
to be quite free of nonbonded steric compression. In
contrast, the state of affairs in 31 is such that the
dihydrofuran moiety must recognize an added modicum
of interaction with the R-methylbenzyl substituent, with
the result that conrotatory cyclization in this instance is
bridled to some degree.
Product partitioning can now be comparably appreci-
ated for the other cascades reported herein. The data
for entries 1 and 6, for example, are internally consistent
if the pinanyl substituent destabilizes the pro-S confor-
mation relative to the pro-R option. While the effects
are certainly subtle, they do point up the reality of
effective 1,5-asymmetric transmission in these systems.
2
of deuterium incorporation to be 80%, D NMR revealed
the double bond geometry to be exclusively Z. The
lithiated form of 27 was subsequently utilized in a
matched pair of experiments.
In the first cascade, the aminated nucleophile 27 was
the lead reagent, to be followed by 5-lithiodihydrofuran.
The second set of conditions involved reversal of this
order. As expected, the yields and percent ee values
compare closely to those earlier realized in the unlabeled
examples (Table 1). The stereochemical assignments to
28 and 29 rest reliably on the distinctive chemical shifts
of the exo (δ 2.23) and endo deuterium substituents (δ
2.34) in CHCl3 solution. These methylene positions had
earlier been defined by detailed analysis of the high-field
1H NMR spectrum of 10.
What now has become clear is the extremely high level
of adherence to the electrocyclic pathway under condition
A. In the absence of any measurable cis attack during
the second-stage addition, a feature that is directly
related to the absence of detectable levels of 29, the
percent ee therefore represents the asymmetric partition-
ing associated with the competitive ring closures of 30
and 31 (Scheme 5). The faster rate of cyclization of 30
can be understood by according proper consideration to
the conformation preferentially adopted by tertiary amines
which carry a stereogenic center at their R-position. The
detailed NMR and molecular mechanics study of the
stereodynamics of N-ethyl-N-methyl-2-aminobutane (32)
performed by Danehey and co-workers15 is representa-
tive. The D NMR behavior of 32 is consistent with that
model in which diastereomeric interconversion via ni-
Complete dissection of the mechanistic details associ-
ated with condition A (Scheme 5) allows for equally
insightful analysis of the more intricate features set in
motion under condition B. Involvement of the dihydro-
furanyl anion as the lead nucleophile gives rise to
monoalkoxides 22 and ent-22, as stated before. The early
introduction of an ethereal oxygen center in this manner
provides a sufficiently good binding site for lithium ions
so that a certain percentage of the lithioallylamine
subsequently introduced is enticed into cis addition
(Scheme 6). This event leads necessarily to the formation
of 33 and 34, respectively, and results in rapid [3,3]
sigmatropic shifting via a boatlike transition state16 with
complete transmission of stereochemistry. Thus, 33
proceeds strictly via 35 to (R)-10-endo-d, while 34 gives
rise to (S)-10-endo-d, following initial conversion to 36.
A necessary requirement of the olefin geometry in the
labeled lithiated allylamine is the ultimate placement of
the deuterium in a relationship anti to the tetrahydro-
furan ring in 10. This stereochemical feature is opposite
(16) (a) Berson, J . A.; Dervan, P. B. J . Am. Chem. Soc. 1972, 94,
7597. (b) Paquette, L. A.; Colapret, J . A.; Andrews, D. R. J . Org. Chem.
1985, 50, 201.
(15) Danehey, C. T., J r.; Grady, G. L.; Bonneau, P. R.; Bushweller,
C. H. J . Am. Chem. Soc. 1988, 110, 7269.

1,5-Asymmetric Induction in Squarate Cascades
J . Org. Chem., Vol. 63, No. 6, 1998 2027
Sch em e 5
to that demanded of the electrocyclic alternative arising
from initial trans addition.
From the four equations given below,
respective tricyclic ketones (33% or 67%, Scheme 5). To
arrive at the levels of the endo-d isomers, it becomes
necessary only to subtract from the product composition
determined in Scheme 4 the relative amounts of their
exo-d counterparts generated by the electrocyclic route.
In this final analysis, the enantiomeric endo-d isomers
in combination represent 43% of the total, and the exo-d
products the remaining 57%. Also, the two R antipodes
make up 43% of the mixture and the S enantiomers 57%.
(R)-10-exo-d ) 0.57 × 0.33 × 100 ) 19%
(S)-10-exo-d ) 0.57 × 0.67 × 100 ) 38%
(R)-10-endo-d ) 0.43 - (R)-10-exo-d × 100 ) 24%
(S)-10-endo-d ) 0.57 - (S)-10-exo-d × 100 ) 19%
it is possible to derive the percent composition of the four
possible products formed in Scheme 6. If it be assumed
that partitioning of the helical intermediates is un-
changed relative to that determined for condition A
(Scheme 5), then the limits for the two exo-d ketones are
governed by the multiplication products of the extent to
which the exo-d isomer is formed (57%, Scheme 4) and
the relative manner in which 30 and 31 advance to the
Over view . A complete dissection of an asymmetric
squarate ester cascade has been accomplished. The
analysis has been made possible by the self-immolative
manner in which a chiral amino substituent induces long-
range 1,5-asymmetric induction prior to being ejected
from an advanced intermediate. Despite the substantive
intramolecular distances that materialize between the
developing stereogenic center and the chiral nitrogen
substituent during conrotatory cyclization of the helical
octatetraene dialkoxides, stereochemical transmission

2028 J . Org. Chem., Vol. 63, No. 6, 1998
Paquette and Tae
Sch em e 6
to 0 °C, treated dropwise with methyl chloroformate (2.64 g,
2.80 mmol), warmed to room temperature, and stirred for 5
h. The separated aqueous phase was extracted with CH2Cl2
(2 × 20 mL), and the combined organic layers were dried and
concentrated to leave a colorless oil.
levels up to 35% ee were observed in several examples.
We believe the highly ordered nature of these stereoelec-
tronically demanding transition states to be responsible.
If this is correct, variants of these tandem reactions may
prove more generally applicable than heretofore appreci-
ated for the impressively abbreviated enantiocontrolled
construction of complex polycyclic systems.
This oily carbamate was dissolved in dry THF (10 mL),
added dropwise to a cold (0 °C), stirred slurry of lithium
aluminum hydride (1.32 g, 34.9 mmol) in dry THF (70 mL),
and refluxed overnight. After being cooled, the reaction
mixture was treated with 2 N NaOH (2 mL) solution, and the
resultant precipitate was filtered off and rinsed thoroughly
with ether. The filtrate was freed of solvent and used directly.
Exp er im en ta l Section
The general experimental protocols followed in this study
parallel those described earlier in ref 6.
The unpurified N-methylamine was dissolved in CH2Cl2 (50
mL) and treated with triethylamine (4.9 mL, 34.9 mmol) and
then 2,3-dibromopropene (3.0 mL, 23.3 mmol). The reaction
mixture was stirred overnight, washed with water (50 mL),
dried, and concentrated in vacuo. Chromatography of the
Gen er a l P r oced u r e for Con ver tin g P r im a r y Am in es
to 2-Br om oa llyla m in es. A solution of the primary amine
(23.3 mmol) in CH2Cl2 (20 mL) was admixed with potassium
carbonate (6.43 g, 46.5 mmol) in water (20 mL) and with tetra-
n-butylammonium iodide (50 mg). This mixture was cooled

1,5-Asymmetric Induction in Squarate Cascades
J . Org. Chem., Vol. 63, No. 6, 1998 2029
residue on silica gel (elution with 10:1 hexanes/ethyl acetate)
gave the desired bromoamine.
28.5, 23.1 ppm; MS m/z (M+) calcd 233.0415, obsd 233.0396;
[R] -64.6° (c 0.35, CHCl3). Anal. Calcd for C9H16BrNO: C,
46.17; H, 6.89. Found: C, 46.26; H, 6.84.
(2S,5S)-1-(2-Br om oa llyl)-2,5-d im eth ylp yr r olid in e (19):
84% yield; colorless oil; 1H NMR (300 MHz, CDCl3) δ 5.88 (m,
1H), 5.48 (s, 1 H), 3.33 (s, 2 H), 3.15-3.00 (m, 2 H), 2.10-1.90
(m, 2 H), 1.45-1.30 (m, 2 H), 0.94 (d, J ) 6.2 Hz, 6 H); 13C
NMR (75 MHz, CDCl3) 133.5, 116.4, 55.7, 55.1, 31.0, 17.3 ppm;
MS m/z (M+) calcd 217.0466, obsd 217.0451; [R] 83.5° (c 0.22,
CHCl3). Anal. Calcd for C9H16BrN: C, 49.56; H, 7.39.
Found: C, 49.27; H, 7.36.
B. Mor e F or cin g Con d ition s. (S)-N-(2-Br om oa llyl)-r-
m eth yld iben zyla m in e (14). A mixture of the secondary
amine (3.0 mL, 14.3 mmol), K2CO3 (5.95 g, 42.9 mmol), n-Bu4-
NI (50 mg), and 2,3-dibromopropene (1.85 mL, 14.3 mmol) in
DMF (35 mL) was refluxed for 2.5 h. The reaction mixture
was cooled to room temperature, filtered off, and diluted with
brine (50 mL) and ether (100 mL). The separated organic layer
was washed with brine (50 mL), dried, and concentrated to
give 4.4 g (93%) of colorless liquid 14: 1H NMR (300 MHz,
CDCl3) δ 7.62-7.38 (m, 10 H), 6.11 (s, 1 H), 5.74 (s, 1 H), 4.19
(q, J ) 6.9 Hz, 1 H), 3.81 (d, J ) 14.0 Hz, 1 H), 3.72 (d, J )
14.0 Hz, 1 H), 3.54 (d, J ) 15.2 Hz, 1 H), 3.33 (d, J ) 15.2 Hz,
1 H), 1.58 (d, J ) 6.9 Hz, 3 H); 13C NMR (75 MHz, CDCl3)
142.1, 139.5, 133.1, 128.6, 128.2, 128.0, 127.8, 126.8, 118.0,
57.6, 56.5, 53.4, 14.4 ppm; MS m/z (M+) calcd 329.0779, obsd
(S)-N-(2-Br om oa llyl)-N,r-d im et h ylb en zyla m in e (11):
1
77% yield; colorless oil; H NMR (300 MHz, CDCl3) δ 7.40-
7.23 (m, 5 H), 5.90 (s, 1 H), 5.57 (s, 1 H), 3.71 (q, J ) 6.8 Hz,
1 H), 3.28 (d, J ) 14.9 Hz, 1 H), 3.07 (d, J ) 14.9 Hz, 1 H),
2.23 (s, 3 H), 1.40 (d, J ) 6.8 Hz, 1 H); 13C NMR (75 MHz,
CDCl3) 143.3, 132.6, 128.1 (2 C), 127.6 (2 C), 126.9, 117.8, 62.6,
40.1, 38.0, 18.0 ppm; MS m/z (M+) calcd 253.0466, obsd
253.0486; [R] -12.3° (c 0.21, CHCl3).
(R)-N-(2-Br om oa llyl)-N,r-d im eth yl-1-n a p h th a len em e-
th yla m in e (12): 69% overall yield; colorless oil; 1H NMR (300
MHz, CDCl3) δ 8.46 (d, J ) 7.7 Hz, 1 H), 7.87 (m, 1 H), 7.78
(d, J ) 8.2 Hz, 1 H), 7.63 (d, J ) 7.0 Hz, 1 H), 7.56-7.44 (m,
3 H), 5.88 (s, 1 H), 5.54 (s, 1 H), 4.44 (q, J ) 6.7 Hz, 1 H), 3.37
(d, J ) 15.0 Hz, 1 H), 3.22 (d, J ) 15.0 Hz, 1 H), 2.31 (s, 3 H),
1.53 (d, J ) 6.7 Hz, 3 H); 13C NMR (75 MHz, CDCl3) 139.8,
134.0, 132.4, 131.7, 128.6, 127.6, 125.5, 125.3, 125.2, 124.5,
124.4, 117.6, 63.0, 60.1, 38.1, 16.9 ppm; MS m/z (M+) calcd
303.0622, obsd 303.0620; [R] -34.1° (c 0.22, CHCl3).
(1R ,2S ,4R )-N -(2-B r o m o a lly l)-N -m e t h y l-2-b o r n a n -
a m in e (13): 72% overall yield; colorless oil; 1H NMR (300
MHz, CDCl3) δ 5.92 (s, 1 H), 5.54 (s, 1 H), 3.29 (d, J ) 15.6
Hz, 1 H), 3.06 (d, J ) 15.6 Hz, 1 H), 2.51 (m, 1 H), 2.22 (s, 3
H), 2.10-1.95 (m, 2 H), 1.71 (m, 1 H), 1.57 (t, J ) 4.6 Hz, 1
H), 1.30-1.15 (m, 2 H), 1.10 (dd, J ) 12.4, 4.0 Hz, 1 H), 0.95
(s, 3 H), 0.87 (s, 3 H), 0.83 (s, 3 H); 13C NMR (75 MHz, CDCl3)
132.9, 117.0, 69.7, 65.1, 50.0, 48.9, 44.3, 41.9, 36.5, 28.7, 27.0,
20.1, 18.7, 16.6 ppm; MS m/z (M+) calcd 285.1092, obsd
329.0787; [R] -33.6° (c 0.35 CHCl3). Anal. Calcd for C18H20
BrN: C, 65.46; H, 6.10. Found: C, 65.39; H, 6.06.
-
P r ototyp ica l Squ a r a te Ester Ca sca d e. I. Con d ition
A. To a cold (-78 °C), magnetically stirred solution of 11 (687
mg, 2.70 mmol) in dry THF (5 mL) was added tert-butyllithium
(3.20 mL of 1.7 M in pentane, 5.40 mmol) dropwise. After 20
min at this temperature, a solution of dimethyl squarate (256
mg, 1.80 mmol) in THF (5 mL) was slowly introduced, to be
followed 30 min later with 5-lithio-2,3-dihydrofuran [prepared
from dihydrofuran (0.27 mL, 3.60 mmol) and tert-butyllithium
(2.12 mL of 1.7 M in pentane, 3.60 mmol)] as rapidly as
possible. The reaction mixture was continuously agitated for
2 h at -78 °C, allowed to warm to room temperature overnight
(13 h), and diluted at 0 °C with deoxygenated NH4Cl solution
(10 mL). After 1.5 h at 20 °C, saturated NaHCO3 solution (20
mL) was introduced, the separated aqueous layer was ex-
tracted with ether (2 × 30 mL), and the combined organic
layers were dried and evaporated. Chromatography of the
residue on silica gel (elution with 2:1 hexanes/ethyl acetate)
gave 133 mg (29%, 35% ee) of 10 as a colorless liquid. This
product was spectroscopically identical to known racemic
material.6
II. Con d ition B. A cold (-78 °C) solution of 2,3-dihydro-
furan (0.15 mL, 1.98 mol) in dry THF (5 mL) was treated
dropwise with tert-butyllithium (1.17 mL of 1.7 M in pentane,
1.98 mmol). After 10 min at -78 °C, 30 min at 0 °C, and an
additional 10 min at -78 °C, a solution of dimethyl squarate
(256 mg, 1.80 mmol) in THF (5 mL) was slowly introduced
followed 30 min later by the lithiated amine [prepared from
11 (916 mg, 3.60 mmol) and tert-butyllithium (2.12 mL of 1.7
M in pentane, 7.20 mmol) in THF (5 mL) at -78 °C]. The
reaction mixture was stirred at -78 °C for 2 h and at room
temperature for 13 h prior to being returned to 0 °C and
treated with deoxygenated saturated NH4Cl solution (10 mL).
The preceding workup was next followed to provide 111 mg
(24% of 10 (18% ee).
285.1094; [R] 23.7° (c 0.20, CHCl3). Anal. Calcd for C14H24
-
BrN: C, 58.74; H, 8.45. Found: C, 58.61; H, 8.41.
(1R ,2R ,3R ,5S )-N -(2-Br om oa llyl)-N -m e t h yl-3-p in a n -
a m in e (16): 70% overall yield; colorless oil; 1H NMR (300
MHz, CDCl3) δ 5.90 (s, 1 H), 5.53 (s, 1 H), 3.33 (d, J ) 14.9
Hz, 1 H), 3.18 (d, J ) 14.9 Hz, 1 H), 3.11 (m, 1 H), 2.31 (s, 3
H), 2.26 (m, 1 H), 2.10-1.90 (s, 3 H), 1.80-1.70 (m, 2 H), 1.19
(s, 3 H), 1.10 (d, J ) 7.1 Hz, 3 H), 0.98 (s, 3 H), 0.86 (d, J )
9.8 Hz, 1 H); 13C NMR (75 MHz, CDCl3) 133.2, 117.2, 63.2,
61.9, 48.0, 41.5, 39.9, 39.0, 37.9, 33.3, 28.0, 27.1, 23.3, 21.7
ppm; MS m/z (M+) calcd 285.1092, obsd 285.1088; [R] -60.5°
(c 0.63, CHCl3). Anal. Calcd for C14H24BrN: C, 58.74; H, 8.45.
Found: C, 58.74; H, 8.44.
(S )-N -(2-B r o m o a lly l)-N ,r-d i m e t h y lc y c lo h e x a n e -
m eth yla m in e (17): 80% overall yield; colorless oil; 1H NMR
(300 MHz, CDCl3) δ 5.88 (s, 1 H), 5.50 (s, 1 H), 3.23 (d, J )
15.3 Hz, 1 H), 3.10 (d, J ) 15.3 Hz, 1 H), 2.30 (m, 1 H), 2.20-
2.10 (m, 1 H), 2.15 (s, 3 H), 1.75-1.60 (m, 4 H), 1.30-1.10 (m,
4 H), 0.95-0.75 (m, 2 H), 0.88 (d, J ) 6.6 Hz, 3 H); 13C NMR
(75 MHz, CDCl3) 133.4, 116.7, 63.2, 62.6, 41.5, 36.2, 30.9, 30.5,
26.7, 26.5, 26.3, 10.0 ppm; MS m/z (M+) calcd 259.0935, obsd
259.0933; [R] -30.7° (c 0.61, CHCl3). Anal. Calcd for C12H22
BrN: C, 55.39; H, 8.52. Found: C, 55.44; H, 8.49.
-
(R)-N-(2-Br om oa llyl)-N-m eth yl-1-in d a n a m in e (18): 75%
1
overall yield; colorless oil; H NMR (300 MHz, CDCl3) δ 7.48
(m, 1 H), 7.30-7.20 (m, 3 H), 6.00 (s, 1 H), 5.61 (s, 1 H), 4.52
(t, J ) 7.5 Hz, 1 H), 3.29 (d, J ) 14.9 Hz, 1 H), 3.19 (d, J )
14.9 Hz, 1 H), 3.05-2.80 (m, 2 H), 2.32 (s, 3 H), 2.25-2.10 (m,
1 H), 2.10-1.95 (m, 1 H); 13C NMR (75 MHz, CDCl3) 143.3,
143.2, 132.7, 127.4, 126.2, 125.0, 124.5, 117.4, 69.0, 61.3, 37.8,
30.4, 24.3 ppm; MS m/z (M+) calcd 265.0466, obsd 265.0457;
[R] -32.1° (c 0.85, CHCl3). Anal. Calcd for C13H16BrN: C,
58.66; H, 6.06. Found: C, 58.58; H, 6.06.
Gen er a l P r oced u r e for Con ver tin g Secon d a r y Am in es
to 2-Br om oa llyla m in es. A. Mild Con d ition s. The amine
(28.0 mmol) was dissolved in CH2Cl2 (70 mL) and treated
sequentially with triethylamine (8.61 mL, 56.0 mmol) and 2,3-
dibromopropene (3.97 mL, 28.0 mmol). The reaction mixture
was stirred overnight and processed in the previously de-
scribed manner.
P r ep a r a tion of th e (S)-Ma n d ela te Ester s in 25 a n d 26.
A solution of 10 (355 mg, 1.41 mmol, 35% ee) in dry THF (10
mL) was treated with a solution of lithium aluminum hydride
in THF (5.6 mL of 1.0 M, 5.6 mmol) at -78 °C. After 10 min,
the reaction mixture was warmed to 0 °C for 1 h, at room
temperature for 30 min, and quenched with 3 N sodium
hydroxide solution. After filtration to remove solids, the
filtrate was concentrated and the residue was chromato-
graphed on silica gel. Elution with 1:1 hexanes/ethyl acetate
gave pure 23 (195 mg, 62%) as a white solid.
(S)-1-(2-Br om oallyl)-2-(m eth oxym eth yl)pyr r olidin e (15):
1
92% yield; colorless oil; H NMR (300 MHz, CDCl3) δ 5.81 (s,
1 H), 5.47 (s, 1 H), 3.68 (d, J ) 15.0 Hz, 1 H), 3.40-3.20 (m,
2 H), 3.31 (s, 3 H), 3.18 (d, J ) 15.0 Hz, 1 H), 3.08 (m, 1 H),
2.73 (m, 1 H), 2.25 (m, 1 H), 1.95-1.53 (series of m, 4 H); 13C
NMR (75 MHz, CDCl3) 132.4, 117.3, 76.5, 63.5, 62.6, 59.0, 54.2,
This alcohol was dissolved in CH2Cl2 (20 mL) and treated
sequentially with dicyclohexylcarbodiimide (269 mg, 1.31

2030 J . Org. Chem., Vol. 63, No. 6, 1998
Paquette and Tae
mmol), (S)-O-methylmandelic acid (159 mg, 0.96 mmol), and
4-(dimethylamino)pyridine (10 mg). After 3 h at room tem-
perature, the solid precipitate was filtered off, the filtrate was
concentrated, and the residue was chromatographed on silica
gel (elution with 8:1 hexanes/ethyl acetate) to give 233 mg
(72%) of 25 and 26 (ratio 2:1) that proved to be partially
separable under these conditions.
(S )-N -[(Z)-2-Br om oa llyl-3-d ])-N ,r-d im e t h ylb e n zyl-
a m in e (27). A cold (0 °C) solution of bromo-9-BBN (52 mL of
1.0 M in CH2Cl2, 52 mmol) was treated dropwise with
propargyl bromide (3.0 mL of 80% in toluene, 32 mmol), stirred
at room temperature for 1 h, and returned to 0 °C prior to the
introduction of CH3CO2D (10 mL). After 1 h at this temper-
ature, 3 N sodium hydroxide solution (160 mL) and 30%
hydrogen peroxide were added. The reaction mixture was
stirred at 20 °C for 30 min prior to separation of the aqueous
phase and extraction of the latter with CH2Cl2 (2 × 50 mL).
The combined organic layers were dried and concentrated to
leave the labeled 2,3-dibromopropene, which was immediately
reacted with freshly prepared N-methylamine (5.24 g, 38
mmol) in CH2Cl2 (220 mL) containing triethylamine (42.2 mL,
158 mmol). After a 5 h reaction time, the predescribed workup
was applied and 5.45 g (61%) of 27 was obtained as a colorless
For 25: IR (film, cm-1) 3590, 1756, 1651; 1H NMR (300 MHz,
CDCl3) δ 7.50-7.40 (m, 2 H), 7.40-7.25 (m, 3 H), 5.89 (s, 1
H), 5.15 (s, 1 H), 4.84 (s, 1 H), 4.81 (s, 1 H), 4.55 (s, 1 H), 3.70-
3.60 (m, 1 H), 3.55-3.45 (m, 1 H), 3.42 (s, 3 H), 3.40 (s, 3 H),
2.80 (s, 1 H), 2.75-2.65 (m, 1 H), 2.50-2.35 (m, 1 H), 2.04 (d,
J ) 15.2 Hz, 1 H), 1.75-1.65 (m, 1 H), 1.45-1.35 (m, 1 H); 13
C
NMR (75 MHz, CDCl3) 169.9, 158.3, 154.1, 136.1, 128.5, 128.4,
126.8, 107.1, 101.5, 95.1, 83.9, 82.9, 79.0, 69.3, 57.5, 56.6, 39.3,
35.4, 33.6 ppm; MS m/z (M+) calcd 372.1573, obsd 372.1530.
For 26: IR (film, cm-1) 3590, 1756, 1651; 1H NMR (300 MHz,
CDCl3) δ 7.45-7.35 (m, 2 H), 7.35-7.20 (m, 3 H), 5.90 (s, 1
H), 5.12 (s, 1 H), 4.79 (s, 2 H), 4.56 (s, 1 H), 3.60-3.45 (m, 1
H), 3.54 (s, 3 H), 3.36 (s, 3 H), 3.25-3.15 (m, 1 H), 2.78 (s, 1
H), 2.30-2.20 (m, 2 H), 1.83 (d, J ) 13.6 Hz, 1 H), 1.35-1.10
(m, 2 H); 13C NMR (75 MHz, CDCl3) 169.3, 157.8, 153.7, 136.2,
128.6, 128.4, 127.2, 106.8, 101.5, 94.8, 83.5, 81.9, 78.6, 68.9,
57.0, 56.5, 38.8, 35.0, 33.1 ppm; MS m/z (M+) calcd 372.1573,
obsd 372.1570.
2
oil; H NMR (76.8 MHz, CHCl3) δ 5.53 (80% d1; 100% Z).
Ack n ow led gm en t. Financial support was provided
by the National Science Foundation. In addition, we
2
thank Dr. Charles Cottrell for recording the H NMR
spectra and Dr. Kurt Loening for assistance with
nomenclature.
J O9722921