A systematic investigation of quaternary ammonium ions as asymmetric phase-transfer catalysts. Synthesis of catalyst libraries and evaluation of catalyst activity

FEATURED ARTICLE

pubs.acs.org/joc

A Systematic Investigation of Quaternary Ammonium Ions as

Asymmetric Phase-Transfer Catalysts. Synthesis of Catalyst Libraries

and Evaluation of Catalyst Activity

Scott E. Denmark,* Nathan D. Gould, and Larry M. Wolf

Roger Adams Laboratory, Department of Chemistry, University of Illinois, Urbana, Illinois 61801, United States

S Supporting Information

b

ABSTRACT: Despite over three decades of research into

asymmetric phase-transfer catalysis (APTC), a fundamental

understanding of the factors that aﬀect the rate and stereo-

selectivity of this important process are still obscure. This paper

describes the initial stages of a long-term program aimed at

elucidating the physical organic foundations of APTC employ-

ing a chemoinformatic analysis of the alkylation of a protected glycine imine with libraries of enantiomerically enriched quaternary

ammonium ions. The synthesis of the quaternary ammonium ions follows a diversity-oriented approach wherein the tandem inter-

[

4 þ 2]/intra[3 þ 2] cycloaddition of nitroalkenes serves as the key transformation. A two-part synthetic strategy comprised of (1)

preparation of enantioenriched scaﬀolds and (2) development of parallel synthesis procedures is described. The strategy allows for

the facile introduction of four variable groups in the vicinity of a stereogenic quaternary ammonium ion. The quaternary ammonium

ions exhibited a wide range of activity and to a lesser degree enantioselectivity. Catalyst activity and selectivity are rationalized in a

qualitative way on the basis of the eﬀective positive potential of the ammonium ion.

’

INTRODUCTION

The requirements for a catalytic, enantioselective enolate

alkylation are the in situ generation of an enolate, alkylation in

the presence of a stereochemical controlling element, and

stereochemical stability of the product (Scheme 1). Although

auxiliary-based methods for the diastereoselective alkylation of

enolates truly revolutionized the practice of organic synthesis

Phase-transfer catalysis (PTC) is an extremely useful method

for performing nearly any type of reaction involving an ionic

starting material or intermediate. A number of characteristics of

PTC reactions make them especially attractive for industrial

applications including, ease of scalability, their intrinsic “green

3,4

beginning in the 1970s, a general and selective processes for

nature”, and potential for extension to asymmetric variants

5

1

catalytic enantioselective alkylations remains elusive. In contrast

(

APTC). One of the most useful aspects of APTC is the ability

to other carbonꢀcarbon bond-forming reactions of enolates

to catalyze a wide variety of reaction types, ranging from redox

processes to many carbonꢀcarbon bond-forming reactions. How-

ever, despite over three decades of research into APTC, a funda-

mental understanding of the factors that aﬀect catalyst activity and

stereoselectivity is still obscure. The purpose of this report is to

describe the initiation of a research program to elucidate the

structural features that govern the activity and enantioselectivity of

quaternary ammonium ion phase-transfer catalysts. The research

program consists of two separate components described in this

(

compare to, e.g., aldol and Michael reactions), the rate of

enolate alkylation is signiﬁcantly less than addition to a π-system;

thus, the use of soft-enolization techniques which are crucial for

catalytic π-addition reactions cannot be utilized for enolate

6

alkylations. The use of more reactive enolates, and strong bases

to generate them, introduces a host of compatibility issues, thus

preventing the implementation of similar stereocontrol elements

under catalytic conditions.

2

Nevertheless, notable successes have been achieved that repre-

sent creative solutions to some of the challenges described above.

For example, a lithio cycloalkanone enolate alkylation has been

and the accompanying paper. In this report we describe the

development of suitably ﬂexible synthesis of chiral, nonracemic,

quaternary ammonium ions by solution-phase parallel synthesis as

well as the collection of a large set of catalyst activity and selectivity

data. In the accompanying report (DOI 10.1021/jo2005457), the

data sets are analyzed by developing many quantitative structureꢀ

activity and selectivity-relationships (QSAR and QSSR). An

enolate alkylation reaction was chosen to initiate these studies

for three reasons: (1) it is one of the oldest and most useful

methods to form carbonꢀcarbon bonds, (2) many examples of

PTC enolate alkylations are on record, and (3) the methods to

perform enolate alkylations in a catalytic enantioselective manner

are very limited.

7,8

cleverly engineered using a triamine catalyst. Transition-metal-

catalyzed allylations are the most intensely investigated class of

asymmetric alkylation reactions and have been extensively

9

reviewed. Ligand development has been critical in facilitating

many R-aryl and R-alkenylations of carbonyl compounds under

palladium catalysis, although they are typically limited to generation

10

of quaternary centers to avoid product racemization. Two

Received: March 11, 2011

Published: March 29, 2011

r 2011 American Chemical Society

4260

dx.doi.org/10.1021/jo2005445

_|J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Scheme 1

alternative strategies have recently been reported, the asymmetric

alkylation of ketone-derived tributyltin enolates in the presence

of chromium(salen) complexes and asymmetric dimethoxy-

to a drug design process consisting of iterations of synthesis,

evaluation, and modeling. Reported herein are the ﬁrst itera-

tions of synthesis (160 catalysts) followed by evaluation (activity

and selectivity) and ﬁnally modeling (in the accompanying

11

methylation of N-acylthiazolidinethiones with an orthoester under

12

2

catalysis by a nickel(BINAP) complex. A single report of intra-

molecular alkylation of an enamine generated from a proline

derivative is on record, but extension to intermolecular versions

paper).

’ BACKGROUND

13

is unknown. Perhaps the most imaginative extension of enamine

catalysis for alkylation-type reactions is the enantioselective R-

allylation, R-enolation, R-vinylation, and R-arylation of aldehydes

1. APTC Enolate Alkylations and Known Catalyst Motifs.

Phase-transfer catalysis has already proven to be a broadly

applicable and general method to perform strong base chemistry

14

via the intermediacy of chiral enaminium radicals.

19

catalytically. In a landmark publication, the Merck Process

Group reported the enantioselective methylation of an indanone

under PTC conditions employing an N-4-trifluoromethylbenzyl

To date, PTC is the only catalytic process capable of reducing

to practice the simplest of enolate alkylations, such as methyla-

5

tion, ethylation, and benzylation (vide infra). As a synthetic tool

20

cinchona alkaloid derivative. A second critical advance ap-

APTC is complementary to the recently developed transition-

metal coupling methods that are well suited for coupling with

peared in 1989 when O’Donnell demonstrated the selective

21

2

15

alkylation of glycine imines under similar conditions (Scheme 2).

C-sp electrophiles. The mode of operation of APTC is unique

in that the base and electrophile are in separate phases, thereby

allowing the generation of highly active chiral nucleophiles in the

presence of an electrophile. The broad electrophile scope of

APTC reactions developed thus far suggests that APTC will

eventually be a general method for performing strong base

chemistry catalytically and enantioselectively.

O’Donnell’s alkylation is an operationally simple method

for R-amino acid synthesis and has since served as a bench-

mark for the development of new asymmetric phase-transfer

22

catalysts.

Scheme 2

The state of the art of APTC stands at an interesting crossroad.

On one hand, impressive advances have been achieved in the

ability to execute catalytic reactions that involve reactive carba-

nions and tremendous potential still exists. On the other hand,

without testable hypotheses about what structural features are

dominant in conferring high activity and selectivity to phase-

transfer catalysts, researchers are constrained to make and test

catalysts without guiding principles to aide in the process.

Accordingly, rationalization of activity and selectivity of APTC

is currently done ad hoc, with little understanding of the

observed trends. This dilemma is not surprising. The rates and

selectivities of APTC reactions are governed by interfacial

transport, desolvation, and a host of nonbonded interactions

which are diﬃcult to study. Still more challenging is the a priori

design of asymmetric phase transfer catalysts. In fact, the most

readily available and successful APTCs to date are the cinchona

alkaloid derivatives. Unfortunately, these structures oﬀer little

16,17

opportunity for systematic modiﬁcation.

If anything has been learned about asymmetric catalysis over

the past 30 years it is that major advances in the state of the art are

inexorably tied to the development of a solid, physical organic

foundation upon which greater and greater improvements can be

built. Therefore, we sought to establish a research strategy

capable of elucidating the fundamental origins of the structureꢀ

activity and selectivity of asymmetric phase-transfer catalysts. To

emphasize, the goals of this research were not simply to discover a new

enantioselective catalyst but rather to investigate a new method for

catalyst development. The approach we have adopted is analogous

In the intervening 20 years, many variations of cinchona alkaloid

catalysts have been reported. The ﬁrst notable advances were from

23

24

independent studies by Lygo and Corey, in which the incor-

poration of a 9-anthracenylmethyl moiety as the nitrogen substi-

tuent markedly improved the enantioselectivity (Chart 1).

Analogously, Park and Jew developed a triﬂuorinated N-benzyl

1

8

2

5a

quaternary ammonium catalysts

and dimeric, meta-bridged

25b

cinchoninium catalysts. In recent years, Maruoka and Ooi have

introduced a large number of novel, non-cinchona alkaloid catalyst

4

261

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Chart 1

systems based on a binaphthyl scaﬀold which allows introduction of

substituents in the 3,3 positions. Similarly, employing an in situ

current understanding of hydroxide-initiated PTC reactions

suggests a mechanistic continuum that is strongly dependent

0

26

31b

generation and screening process, Lygo has discovered a useful

on substrate pK and catalyst structure. To summarize, the

a

2

3c

locally C (at the nitrogen)-symmetric catalyst.

Shibasaki/

extraction mechanism is dominant at the two extremes of

2

7

28

29

Ohshima, Sasai, Arai/Nishida, and others have independently

substrate acidity (pK < 16 and pK > 23) and the interfacial

a

developed two-centered (and higher) APTC’s by incorporation of

mechanism is dominant when the pK of the substrate is between

a

quaternary ammonium ions about a C axis. In addition to extensive

16 and 23. The rational is that reactions with highly acidic

substrates are rate limited in extraction of the substrate enolate

and nonacidic substrates are rate limited by the rate of hydroxide

extraction. Thus, the interfacial mechanism is proposed to be

n

catalyst development, signiﬁcant advances in other typically strong-

base-promoted reactions have been recorded including double

alkylation of glycine imines, ketone alkylations, Michael, aldol,

Mannich, and Darzens reactions, as well as epoxidations and

40

most relevant for enolate alkylations.

19kꢀn

aziridinations.

2

. Catalyst Activity. 2.1. Mechanisms of Hydroxide-Initiated

PTC. The broad utility of PTC has motivated numerous investi-

gations from fields as far reaching as synthetic chemistry,

chemical engineering, computational chemistry, physical chem-

istry, and chemoinformatics. PTC reactions that use aqueous

bases to generate carbanions are termed hydroxide-initiated

3

0

PTC. The mechanism of hydroxide-initiated PTC reactions

is an intensely investigated subject that has been thoroughly

Figure 1. Two mechanisms of nucleophile transfer in hydroxide-

initiated PTC Reactions.

31

32

reviewed from both a chemical₃and engineering perspective.

3

34

Pioneering studies by Makosza and Starks led to two distinct

mechanistic descriptions, later termed the interfacial (Makosza)

and extraction (Starks) mechanisms. The mechanisms differ in

2.2. Guidelines for the Design of Active Catalysts. The

quaternary ammonium ion catalyst is obviously the most critical

element of a PTC reaction, and as is the case for any catalytic

the mode in which the substrate ammonium ion pair enters the

3

4

1,31b

organic phase (Figure 1). Numerous studies have provided

reaction, the selection of the “optimal” catalyst is crucial.

35,36

experimental support for both the extraction

mechanisms.

and interfacial

These studies have accumulated a number of

phenomenological observables that are used as indicators to

The fact that there are two possible mechanisms, in combination

with the possibility of many “off-cycle” pre-equilibria, all of which

are a function of substrate and catalyst combination, has made

3

7ꢀ39

31b

31a

decipher which mechanism is dominant. Reactions governed

by the extraction mechanism show little dependence on stirring

rate and first order kinetic behavior in substrate and catalyst. On

the other hand, PTC reactions governed by the interfacial

mechanism show a strong dependence on stirring rate, complex

kinetic order in substrate, and fractional order in catalyst. The

identifying simple structureꢀactivity relationships difficult.

For reactions strictly following an extraction mechanism (e.g.,

S_Ndisplacements with N₃, CN , SCN , etc.), catalyst activity

2

ꢀ

3

ꢀ

6,35,42

correlates well with lipophilicity.

No such generalizable

structureꢀactivity relationships exist for hydroxide-initiated

PTC reactions.

4

262

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

For hydroxide-initiated PTC reactions, small hydrophilic

ammonium ions are often superior catalysts, which is a com-

corroborate that, in solution, ammonium ion ester enolates tend

to orient “face-on” to one of the faces of the ammonium

43,44

51b

pounding source of confusion.

For example, triethylbenzy-

(Figure 3b). The combination of these analyses lends to a

lammonium, a small hydrophilic quaternary ammonium ion, is an

useful design mnemonic by inscribing the ammonium nitrogen

in a tetrahedron (red) where the vertices are the four carbons

bound to it. The face proximal to the oxygen is proposed to lead

to more selective reaction of the substrate bound to it.

eﬃcient catalyst for a myriad of PTC enolate alkylations (e.g.,

45

nitriles, esters, ketones, etc.). One approach has been to

correlate a macroscopic observable such as interfacial surface

tension to catalytic activity which unfortunately does not address

46

the question of what catalyst attributes confer high activity. In

the most advanced SAR to date, the ammonium ion accessibility

(

or size) and solubility were varied simultaneously, and the most

active symmetrical catalyst for alkylations was tetraethylammo-

nium bromide. A useful quantitative correlation of structure to

catalyst activity was subsequently derived, namely the structural

4

7

48

parameter, q (eq 1, where C is the number of carbons in C ).

n

i

The q parameter is deﬁned as the sum of the reciprocals of the

49

number of carbons on each chain. For hydroxide-initiated PTC

reactions, such as the alkylation of enolates, q values between 1.5

and 2.0 are optimal. This analysis is useful for simple quaternary

Figure 3. Two potentially generalizable levels of analysis for ammo-

nium ion enolate binding.

ammonium ions containing only alkyl chains (C₁) but exten-

4. Objectives of this Study. Our initial survey of the literature

revealed a few significant trends. First, while a myriad of APTC

methods have been developed for enolate π-addition reactions,

the reports on analogous alkylation reactions remains limited,

ꢀ4

sion to catalysts containing functional groups, changes in hybri-

dization, rings, branch points, or stereogenic centers has not been

reported.

19m

largely O’Donnell’s glycine imine substrate.

It is not clear

whether the lack of reports on new hydroxide-initiated PTC

enolate alkylations is because known catalysts are insufficiently

active, insufficiently selective, or both. What is clear, however, is

that a more rational, systematic manner to develop quaternary

ammonium ion catalysts is needed. Specifically, to elevate APTC

to the state of rational design available to other catalytic

enantioselective methods, two unique challenges need to be

addressed. First, a method to estimate the activity of catalysts a

priori is needed. This capacity would greatly alleviate the

synthetic investment currently required to study APTC and, in

so doing, facilitate discovery. Second, a clearer picture of

structural features that confer high selectivity to asymmetric

phase-transfer catalysts is needed.

3

. Structural Considerations for Catalyst Enantioselectiv-

ity. The rationalization of enantioselectivity and reduction to

design criteria is of preeminent importance to the study of

5

0

catalytic processes. The alkylation of O’Donnell’s imine is by

far the most studied enolate alkylation, and two models have

been proposed to rationalize the enantioselectivity observed with

cinchona alkaloid derived catalyst (Figure 2). In a thorough

computational study, O’Donnell and Lipkowitz analyzed the

enolate molecular recognition event and origin of stereoselectiv-

ity of this important transformation. Computational results

suggest selective binding of the catalyst to the Si face of the Z-

enolate (Figure 2a). Subsequently, it has been proposed that

catalysts bearing a 9-methylanthracenyl group on nitrogen react

51

52

Each of these aspects is a daunting task because they require

systematically varying the topology of a quaternary ammonium

ion catalyst, which is diﬃcult to synthesize in enantiopure form.

Fortunately, the pharmaceutical drug discovery model provides

23,24

through the E-enolate (Figure 2b).

55

close analogy for both of these challenges. Thus, the ﬁrst phase

of these studies has been initiated in analogy to a phenotypic

screening assay where catalyst activity will be evaluated as a

5

6

function of molecular “pharmacodynamic” properties. The

intermolecular enantiotopic face discrimination event is analo-

gous to molecular recognition of a substrate (or drug) by an

enzyme, except that the synthetic eﬀort is tantamount to building

the enzyme binding pocket rather than the substrate.

We therefore set out to devise a research strategy that would

provide a proof of principle as a catalyst development methodol-

ogy. The key objectives were to develop quantitative analyses of

both catalyst activity and enantioselectivity, in order to probe the

underlying hypothesis that a quantitative treatment of catalyst

activity and selectivity as a function of catalyst structure would

naturally lend to a more facile catalyst development process. To

emphasize, the motivational factors for this study were to probe

the manner in which catalyst development is done and not simply to

develop a catalyst. We therefore made two decisions before

initiating this study. First, the alkylation of O’Donnell’s imine

would serve as the benchmark reaction for catalyst screen-

ing, thereby foregoing a methodological development period.

Figure 2. Models to rationalize the enantioselectivity of cinchona-

derived PTC catalysts.

The binding site of cinchona alkaloid derived catalysts has

been probed by intermolecular NOE correlations between

ammonium borohydrides and ﬂuorides and indicates two

53

“

front-face” binding regions (Figure 3a). These studies provide

good support for strong hydrogen-bonding interactions of the

þ

ꢀ

type N CH O in solution which has also been identiﬁed in

3 3

3

the solid state by X-ray crystallographic analysis of quaternary

ammoniumꢀenolate ion pairs. High-level calculations further

54

4

263

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Second, an arbitrary, and modest, enantioselectivity threshold

of ∼80:20 was set as the point at which random catalyst screen-

ing would stop and a more systematic, QSAR-guided investiga-

tion would begin. After all, developing a QSAR model on a data

set where catalyst selectivities exceed 99:1 is, in many ways, a

moot point because the end goal has already been reached. No

limit for range of activity data was set initially, but the preliminary

survey readily identiﬁed catalysts with activity diﬀerences up to 4

orders of magnitude, which is well suited for a QSAR study. The

design and synthesis of a suitably diverse libraries of quaternary

ammonium ion catalysts and collection of activity and selectivity

data for this initial data set is described below. A qualitative

analysis of the observed trends follows which are reduced to

quantiﬁable changes in the catalysts and interpreted mechan-

Scheme 3

central position of a rigid ring system, thereby providing potential

for the controlled installation of groups in the vicinity of the

nitrogen. The fused mode, tandem inter [4 þ 2]/intra [3 þ 2]

cycloaddition with a two-carbon tether was chosen for initial

studies. In this mode, the tandem cycloaddition/hydrogenolysis

sequence generates a tricyclic cyclopentapyrrolizidin-2-one ring

system bearing a hydroxy group at C(1) and a substituent at C(5)

2

istically in the accompanying paper. Ongoing studies (data not

shown) indicate that these, quantitative models are readily

utilized as a methodological tool for further catalyst optimization.

’

RESULTS

. Research Plan and Design. 1.1. Scaffold Synthesis Method.

(

Scheme 3). Prior to elaboration of a detailed parallel synthesis

1

strategy, the local environment around the ammonium nitrogen

Without testable hypothesis about what structural features will

dominant in conferring high catalyst activity and selectivity, this

endeavor was necessarily initiated as a discovery-oriented pro-

gram. The initial focus was set on (1) identifying a suitable

scaffold (or scaffolds) that could be prepared in enantiopure

fashion and would allow for substitution in multiple positions

followed by (2) developing suitably flexible parallel synthesis

procedures for the elaboration of the scaffold to a diverse library

of quaternary ammonium ions. Concurrently the scaffold and

most easily substituted positions were evaluated in the context of

known quaternary ammonium design principles. The tandem

was considered in the context of known design principles.

1

.3. Scaffold Shape and Ammonium Ion Accessibility. By

connecting the nitroalkene and the dipolarophile with a two-

carbon tether, the ring fusions (C(7b)ꢀN, C(7b)ꢀC(5a), C-

(

7b)ꢀC(7a)) are dictated to all be cis, giving the scaffold a bowl

shape (Figure 4). Library design efforts were initiated by

inscribing the central ammonium nitrogen of the scaffold in a

regular tetrahedron (red) to determine which faces will be

accessible for Coulombic interactions. Clearly, the concave face

inscribed by C(2)ꢀC(4)ꢀC(7b) is shielded by the C(6)ꢀC(7)

methylenes rendering it sterically inaccessible. Initial molecular

[

4 þ 2]/[3 þ 2] cycloaddition of nitroalkenes, a methodology

61

4

modeling indicated that a nitrogen substituent (R ) would

occupy the face defined by C(2)ꢀC(4)ꢀC(9) to avoid steric

interaction with a group at C(8). Therefore, the two most

accessible faces of the ammonium ion tetrahedron appear to be

the two situated on the convex face of the scaffold inscribed by

C(2)ꢀC(9)ꢀC(7b) and C(4)ꢀC(9)ꢀC(7b), respectively.

extensively studied in these laboratories, was quickly identified as

a reaction that embodies all of the requisite characteristics for

57

scaffold preparation.

In the tandem cycloaddition of nitroalkenes, three key compo-

nents, a nitroalkene, a dienophile, and a dipolarophile, combine to

construct a new six-carbon unit forging four bonds, up to six

contiguous stereogenic centers, and four rings. Lastly, the newly

minted skeleton is transformed by hydrogenolysis, which proceeds

by reductive NꢀO bond cleavage followed by intramolecular

reductive amination and ﬁnally closure of the remaining ring to

construct a pyrrolizidine ring system. Of the four possible permuta-

tions of inter- and intramolecularity, the tandem inter [4 þ 2]/intra

[

3 þ 2] family wherein the dipolarophile and nitroalkene are

58

covalently tethered is the most powerful (Scheme 3). Tethering

the dipolarophile to either the R- or β-position of the dienophile

results in bridged cycloadducts and generates primary amines after

hydrogenolysis. Tethering the dipolarophile to the nitroalkene

results in either a spiro-mode (1-position) or a fused-mode (2-

position) cycloaddition and generates tertiary lactams after hydro-

genolysis. Extensive investigation of this family has revealed many

advantages, including (1) ease of preparation of the precursors, (2)

ﬂexibility in the electronic nature and conﬁguration of the compo-

nents, (3) high levels of absolute stereocontrol with chiral dieno-

Figure 4. Steric environment the pyrrolizidine scaﬀold.

1.4. Parallel Synthesis Design. The tandem inter [4 þ 2]/intra

[3 þ 2] cycloaddition of nitroalkenes in the fused mode allows for

1

the stereoselective introduction of groups at C(5) (R ) and a

hydroxy group at C(1) while preserving the relative configuration

1

2

of the scaffold ring system (Scheme 4). The variable groups R , R ,

4

and R are situated on the convex face allowing for the evaluation of

59

philes, and (4) diversity of product structure.

many different combinations of substituents. Because the C(5)-

1

.2. Scaffold Selection. A key challenge to systematic investi-

substituent (R ) is the first point of diversification its variation was

gation of APTC is the availability of catalyst scaffolds that allow

for facile modification of sites proximal to the quaternary

limited (either H or Me), thereby reducing the synthetic invest-

ment required for this initial survey. Moreover, the configuration of

the C(1) center is dictated by the geometry of the dipolarophile

thus introducing an additional element for diversification. After

51c,60

ammonium ion center.

Both the spiro- and fused-mode

cycloadditions create skeletons where the nitrogen is fixed in a

4

264

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

introduction of groups at C(1) and C(5) the scaffolds will be

elaborated by means of parallel synthesis.

Chart 2

The forward synthetic analysis from the scaﬀold is outlined in

Scheme 4. The hydroxyl group at C(1) was targeted for the next

two points of diversiﬁcation. Simple oxidation followed by

2

addition of an organometallic reagent (R -M) would allow for

the introduction of groups directly in the vicinity of the C-

(

2)ꢀC(7b)ꢀC(9) face. Also, alkylation of the resulting alcohol

3

or the original secondary alcohol) with simple alkyl halides (R -

X) would serve as a facile method for introducing groups on the

concave face. Lastly, the nitrogen substituent (R ) is installed by

4

a second alkylation to aﬀord the ﬁnal quaternary ammonium

salts. This analysis served as a general framework from which the

order of synthetic steps suitable for parallel synthesis was worked

out experimentally.

Scheme 4

ion library members will be referred to by roman numerals

libraries IꢀV) followed by a braced number set designating the

order and number of groups introduced {XꢀX,XꢀX}. For

(

1

2

3

example, in Scheme 6 the intermediate free amine X{R ,R ,R }

1

2

3

will be converted to library V{R ,R ,R ,XꢀX} through the action

of reagents {XꢀX}.

2. Library Syntheses. 2.1. Preparation of Tandem Cycloaddi-

tion Precursors. To accurately determine the activity and enan-

tioselectivity for each catalyst (ca. three runs/catalyst), approxi-

mately 30 mg of each quaternary ammonium salt would be

required, thus mandating the synthesis of gram quantities of the

scaffolds for libraries IꢀV. Thus, the first phase of this investiga-

tion was to develop robust, scalable routes to the cycloaddition

precursors (nitroalkenes (E,E)-1 and (E,Z)-1 and chiral vinyl

ethers 2 and 3, above) in decagram quantities. Each of these

cycloaddition precursors were readily prepared either directly

following the known route or with small changes upon scale-

1

.5. Focused Libraries. To facilitate description of the libraries,

the catalysts are divided into focused sets of increasing substitu-

1

62

tion (IꢀV). The R substituent (H, Me) served to influence the

up. The preparation of the cycloaddition precursors 1ꢀ3 is

4

position of the R substituent and to differentiate the accessibility

detailed in the Supporting Information.

of the two convex faces inscribed by C(3)ꢀC(7b)ꢀC(9) and

2.2. Scaffold Preparation by Tandem Cycloaddition of Ni-

troalkenes. The centerpiece of the library synthesis was the use of

the tandem cycloaddition of nitroalkenes to cast the polycyclic

skeletons that will serve as parallel synthesis scaffolds. Known

R-hydroxy lactam 4 was prepared in racemic form by tandem

[4 þ 2]/[3 þ 2] cycloaddition of nitroalkene (E,E)-1 with

n-butyl vinyl ether via the intermediacy of nitronate 6 and nitroso

acetal 7 (Scheme 5). Reductive hydrogenolysis afforded the tricyclic

lactam (()-4a in excellent overall yield (86% over three steps).

2

C(2)ꢀC(7b)ꢀ(C9). The R substituent would play a comple-

mentary role in terms of accessibility and interaction of the

counterion. This group could be more extensively diversified

because of its later stage of introduction in the synthesis. Because

3

R is more distal to the nitrogen atom, the primary focus for this

substituent was to serve as a lipophilicity modifier. The residue

4

R was restricted primarily to groups of varying π-surfaces.

Additionally, because this group is introduced last, a larger degree

of synthetic flexibility is available, thus facilitating a systematic

investigation of the role of steric and electronic factors for these

aryl rings. The proximity of this substituent to the positively

charged nitrogen means that the character of this group should

have a larger influence (through dipole and field effects) on the

localized positive potential encompassing the nitrogen than any

of the other groups. The overall degree of modification in the

scaffold as outlined here should lead to sufficient variation in the

catalyst structural properties for a QSAR analysis (Chart 2). The

simplest members of this group (library I) bear no oxygen-

Scheme 5

containing functionality and vary only in the nitrogen substituent

4

(

R ). Oxygen functionality is introduced in libraries II and III,

whereas additional steric and conformation influencing groups

are introduced in libraries IV and V. Quaternary ammonium

4

265

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Likewise, enantiomerically enriched R-hydroxy lactams 4a

and 5 were readily prepared on scale in nonracemic form via

the tandem cycloaddition with chiral vinyl and propenyl ethers

intermediate amine hydrochloride. The free base was liberated by

partitioning the salt between 0.1 M aq NaOH and ether.

Exposure of the free amine to a slight excess of benzylic and

primary aliphatic bromides {1ꢀ6} in acetonitrile at room

temperature led smoothly to the quaternary ammonium ions

I{1ꢀ6} in 72ꢀ90%. The excess electrophile could easily be

removed by either ﬁltration through silica gel or trituration

6

2

and 3, respectively (Scheme 6). In this process, a standard

set of reaction conditions was employed, namely exposure of

a solution of nitroalkene and dienophile to TiCl (O-i-Pr)₂

2

at ꢀ78 °C to aﬀord the intermediate nitronate. Subsequent

thermal cycloaddition occurred over the course of 2ꢀ3 h

upon standing at room temperature. The resulting nitroso

acetals were immediately subjected to hydrogenolysis with

Raney nickel in methanol to aﬀord the lactams in 76%

and 89% respectively (three steps) with high enantioselec-

6

5

with ether analogous to the methods of Dehmlow and

6

Maruoka, respectively. Ultimately, it was decided that silica

gel plug ﬁltration followed by trituration with ether could

be used as the general conditions for puriﬁcation for other

ammonium salts.

6

3

tivity (er 96:4). The epimeric R-hydroxy lactam 4b (not

previously described) was readily prepared in 63% yield follow-

ing the standard protocol (er 96:4). The elaboration of these

scaﬀolds to the libraries IꢀV, respectively, is detailed in the next

section.

Scheme 7

2

.3. Preparation of Library I. The next phase of these studies

involved the development of a suitably versatile parallel synthesis

route to allow for the introduction of various substituents on the

core scaffold. The simplest scaffold (library I), containing only a

single site of variation, was first targeted to address the introduc-

tion of groups in the key N-alkylation step.

Scheme 6

2

.4. Preparation of Libraries II and III. The next level of

complexity called for the synthesis of libraries of chiral, non-

racemic ammonium salts to study their effectiveness as asym-

metric phase-transfer catalysts. The construction of libraries II

and III introduces a second parallel synthesis step, namely

O-alkylation. These two libraries embodied three objec-

tives aimed at investigating both enantioselectivity and rate

(

1) to introduce groups with variable solubilizing abilities

rate), (2) introduce variable π-surfaces and steric bulk

selectivity), and (3) investigate the effect of configuration

at C(1) (selectivity).

The diastereomeric lactams 4a and 4b were reduced to

the pyrrolizidines 11a and 11b with BH THF (>10 equiv) in

3

To complete the synthesis of the C symmetric scaﬀold,

good to excellent yield (Scheme 8). These amines were also

isolated as their borane adducts, thus protecting the amine from

s

reduction of the two oxygen functional groups was required.

Removal of the hydroxyl group was accomplished by activation as

a phenyl(thiono)carbonate 8 followed by Barton-type deoxy-

67

air oxidation and subsequent alkylation in the next step. This

sequence allowed for the preparation of multigram quantities of

borane complexes 11a and 11b as crystalline solids which was a

convenient stage for storage of material. Application of standard

genation under the action of Bu SnH/AIBN to aﬀord lactam 9 in

3

6

4

good overall yield (64%, two steps) (Scheme 7). Reduction

68

of lactam 9 with borane THF aﬀorded the fully deoxy-

Williamson ether synthesis conditions (NaH, DMF) to borane

complexes 11a and 11b allowed for facile elaboration to inter-

mediates II{1ꢀ5} and III{1ꢀ5}. The introduction of simple

hydrophobic groups (n-hexyl), hydrophilic groups (methoxy-

ethoxymethyl, MEM) as well as aromatic carbocycles and hetero-

cycles, is readily achieved in good yield under a standard set of

reaction conditions. It is noteworthy that the inclusion of a MEM

ether in these series required the use of either neutral or base-

promoted deborylation conditions because of the acid lability of

this group. Accordingly, aq HCl was replaced with aq Na CO

3

genated scaﬀold 16 for library I as its borane adduct in 80%

yield. With large quantities of borane adduct 10 in hand,

N-quaternization conditions suitable for parallel synthesis

could be investigated.

To facilitate the throughput of material, a method for the

direct conversion of the amine borane adduct to the desired

quaternary ammonium salts was sought that avoided the need for

puriﬁcation of intermediates. Heating amine borane 10 in

3

methanol in the presence of 1 M aq HCl led cleanly to the

2

3

4

266

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

under otherwise identical reaction conditions. Elaboration of

the intermediate borane adducts as described previously pro-

ceeded smoothly to aﬀord quaternary ammonium salt libraries

II{1ꢀ5,1ꢀ6} and III{1ꢀ5,1ꢀ6} in good yields over the three

step process.

Scheme 9

Scheme 8

3

following the previously developed protocol. The R substituents

in libraries IV and V were limited to unfunctionalized aliphatic

groups and electron-rich (4-MeOC H ) and electron-deficient

2

.5. Preparation of Libraries IV and V

.5.1. Introduction of R Substituents. The presence of the

2

6

4

hydroxyl group at C(1) allows for further diversification. This

site was targeted for the introduction of different aliphatic and

aromatic groups by organometallic addition to the corresponding

C(1) ketone. Accordingly, Parikhꢀvon Doering oxidation of 11a

(3,5-(CF ) C H ) aromatic groups. Also, the 9-anthrylmethyl

3 2 6 3

4 72

group at R was removed from this set, but a larger number of

groups of varying electronic makeup and size were introduced for

4

R in libraries IV and V.

69

furnished the desired ketone 13 in 98% yield (Scheme 9).

However, ketone 13 was not stable and was immediately carried

Scheme 10

on to the organometallic addition step. The increased acidity of the

7

0

R-hydrogens, likely as a consequence of nitrogen complexation,

required the use of softer, less Brønsted basic organocerium

71

reagents. The resulting additions took place (40ꢀ98% yield) to

form 15{1ꢀ7} with complete β-diastereoselection as a conse-

quence of the bowl shape of the core scaffold. Likewise, ketone 14

(

an inseparable 12:1 mixture of C(5) diastereomers from cycload-

dition with a propenyl ether) underwent the cerium-mediated

addition to afford the tertiary alcohols 16{1ꢀ7} in similar yields.

However, the inability to introduce bulky groups as well as

poor reproducibility led to a modiﬁcation of the synthetic route.

Oxidation of alcohol 5 to R-keto lactam 17 aﬀorded two

important beneﬁts over the initial route. First, the addition of

Grignard reagents to this highly active dicarbonyl compound

proceeded reproducibly and in excellent yields (95ꢀ97%) even

with bulky nucleophiles. Second, the minor C(5) diastereomer

could be removed by a single recrystallization. Reduction of the

resulting lactams 18{4,6,7} required elevated temperatures,

presumably a consequence of the additional steric bulk proximal

2.6. Summary of Library Syntheses. A library of over 160

catalysts sharing the same core scaffold has been generated that

incorporates the substituents below (Figure 5). Although this

number represents only a small fraction of the complete matrix of

∼1100 catalysts, the library represents a good approximation of

the total structural space occupied by the complete matrix on the

basis of preliminary data. The groups shown reflect the need

to evaluate the roles of steric and electronic contributions,

π-surface, lipophilicity, and polar surface area on the catalyst

structure for a given phase-transfer-catalyzed reaction.

to the site of reduction. In this way, the corresponding amine

borane adducts 15 and 16 could be isolated in respectable yields

3

(

72ꢀ85%) without increasing the number of steps in the

synthetic route.

.5.2. Parallel Synthesis toward Libraries IV and V. Exten-

2

sion of the developed parallel synthesis route to the more

hindered tertiary alcohol at C(1) found in libraries IV and V

was straightforward (Scheme 10) and readily accomplished

4

267

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Figure 5. Substituents included in the catalyst library.

3

. Collection of Kinetic Data. 3.1. Refining a Kinetic Analysis.

An important decision was an appropriate value to represent

The base-promoted benzylation of glycine benzophenone imine

the rate data (e.g., k , k_o_b_s, t₁_/₂, etc). Upon initial testing of

i

tert-butyl ester 19 has become the benchmark reaction for investi-

catalyst I{1}, a signiﬁcant induction period was observed over

the course of the ﬁrst 5% conversion. This type of behavior has

previously been reported as a function of ammonium counter-

19m

gation of new catalyst structures.

As such, a standard set of

reaction conditions has been established, namely the use 50% aq

22

4

0,76

KOH solution and toluene. Given the intrinsic biphasic nature of

PTC reactions, efficient mixing is essential to promote efficient

transport and to minimize the errors associated with precipitation.

For these reasons, a 4 mL (1 cm ꢁ3.5 cm) cylindrical vial was

chosen as the reaction vessel together with a 1.5 cm egg-shaped stir

bar. This stir bar was large enough suchthatittraversed theinterface

and yet was small enough to ensure mixing as a consistent shearing

ion and is indicative of an interfacial mechanism.

The

induction period was seen under the protocol wherein the base

was added last. However, if the order of addition of reagents was

changed such that the alkylating agent was added last, then the

converse was observed and appearance of product was rapid at

ﬁrst, but quickly declined (a kinetic “burst”). Most importantly,

regardless of the order of operations, the “time course” of the

two reactions converges to a single rate. Given the signiﬁcant

variation at the onset of the reaction, it was decided to extract

47,73,74

motion.

The scale of the reactions was dictated by the need

to conduct multiple kinetic runs using 80ꢀ120 mg of starting

ester 19 (thus dictating a minimally measurable amount of catalyst

data that would be independent of the initial rate, namely t₁_/₂

.

(

4ꢀ8 mg)) and to maintain a reaction concentration of 0.33 M.

Although this determination is labor intensive, since it would

require monitoring each reaction to greater than 50% conver-

sion, it has the added advantage of showing the entire reaction

proﬁle.

Under these conditions, orienting experiments utilizing 5 mol

%

of a catalyst at room temperature resulted in rapid reaction;

most were complete within 1ꢀ3 min! Clearly, this range of rates

was not amenable to providing reproducible results. Decreasing

the catalyst loading to 2.5 mol % and decreasing the temperature

to 2ꢀ4 °C successfully slowed the reaction to a point where a

3

.2. Stir Rate Dependence. The last critical decision to be

made was a stirring rate that would sufficiently minimize

errors from precipitation and still allow for the differences

in catalyst activities to be observed. Hydroxide-initiated PTC

reactions can exhibit a rate dependence on stirring speed in

excess of ∼2000 rpm. At high enough stirring speeds there

is little dependence which is consistent with the proposal

that the influence of mixing is to increase the surface area to

75

range of rates could be observed (Scheme 11).

Initial experiments with tetrabutylammonium bromide allowed for

an examination of sampling methods to ensure reproducibility. Samp-

ling from the bulk mixture gave inconsistent results and the sampling

needle clogged frequently. A more reliable protocol involved stopping

the agitation momentarily to allow the biphasic mixture to separate

3

9a

(

∼2 s) and sampling from the organic layer. When this protocol was

volume ratio (effectively concentration) of the biphase and

that phase-transfer reagents can decrease the interfacial sur-

used, the reaction proﬁle was reproducible, typically to within 1ꢀ2%,

4

6a

but always within 5% (see the Supporting Information).

face tension.

To determine a suitable stirring speed for application to the

standard protocol, the dependence of catalytic activity on

stirring speed was determined for a variety of variably active

quaternary ammonium ions. Reaction half-lives were deter-

mined following the experimental parameters noted above

employing a range of stir rates (1038ꢀ2500 rpm). The

resulting observed half-lives are plotted as a log/log in analogy

Scheme 11

4

268

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

5

5,56

to a doseꢀresponse graph (Figure 6).

The top two curves

represent catalysts with low catalytic activity while the two

lower curves represent catalysts among the highest activity.

Importantly, there is a suﬃcient range of stirring speeds, along

which the activities of all of the catalysts surveyed are

dependent (nonzero slope) on mixing rate. In choosing the

stir rate with which all of the kinetic experiments were to be

carried out, the important considerations were as follows: (1)

there was suﬃcient and consistent mixing, (2) the experiments

were operationally accessible, and (3) the stir rate existed in a

region where a reaction rate dependence was still observed.

The stir rate of 1600 rpm ﬁt within all three of these criteria

(

represented by the black vertical dotted line in Figure 6), and

Figure 7. Histogram of t₁_/₂data in logarithmic form.

therefore, all of the remaining kinetic data were collected at

this stir rate.

Table 1. Half-life Data for Library I

4

entry

library no.

R

t

1/2, min

1

2

3

4

5

6

I{1}

I{2}

I{3}

I{4}

I{5}

I{6}

C

6

H

5

940

730

na

1-naphthyl

2-naphthyl

9-anthryl

324

na

3 2 6 3

3,5-(CF ) C H

n-hexyl

12.8

Figure 6. Eﬀect of stir rate on reaction half-life with catalysts of varying

structure and activity.

4

.3. Catalytic Activity of Libraries II and III. The half-life

data for libraries II and III is summarized in Table 2. These

ammonium salts are generally more active catalysts than those

lacking an oxygen substituent and exhibited half-life ranges

from 6 min to 5.7 h. Catalysts containing a non-hydrogen

4

. Summary of the Kinetic Data. 4.1. Statistical Description.

With a working analytical method in hand, a large subset of the

ammonium salts was evaluated for their kinetic competence,

represented as half-lives. The half-lives were determined by

interpolation of the kinetic plot of percentage of product

formation as a function of time. The reported half-life values

represent averages over two runs with an average error of

.5%. Data from kinetic runs that resulted in errors exceeding

0% was discarded and repeated until an error of less than 20%

was observed.

To date, half-lives for 102 of the 160 catalysts have been

collected. The observed half-life data cover 4 orders of mag-

nitude of activity that was deemed suitable for the initial

investigation of the structural eﬀects of catalyst on rate. The

half-life data collected from these experiments ranged from

3

substituent at R are generally more active than the correspond-

3

ing catalysts with hydrogen at R . However, no clear depen-

3

dence of the nature of the R substituent on catalyst activity was

3

discernible. When R = H, the rate was markedly dependent

7

4

2

on the configuration at C(1) (entries 5 vs 6). However, when

3

R

6¼ H, the half-life was nearly independent of the C(1)

configuration (entries 10 vs 11, 12 vs 13, 16 vs 17, 21 vs 22, and

3 vs 24).

.4. Catalytic Activity of Libraries IV and V. The half-life data

2

4

for libraries IV and V is summarized in Table 3. The activity of

2

these catalysts appears to be largely dependent on the R

substituent within the series. Comparing catalysts with fixed

5

min to over 15 h and are summarized in logarithmic scale

1

3

4

2

R , R , and R substituents, but with varying R substituents, the

below (Figure 7). Of the 102 catalysts surveyed to date, 52

of them exhibited half-lives of 20 min or less. Twenty-

three catalysts make up the data between 20 min and 1 h,

and the half-lives of the remaining 27 catalysts range from 1

to 15 h.

2

following trend in rates holds: R = i-Pr > Me > Ph. However,

there are important exceptions. For example, catalysts containing

4

1

3

R = 3,5-(CF ) C H , R = Me, R = benzyl, the following trend

3

2 6 3

2

in half-life holds for R :Me>i-Pr > t-Bu > Ph (entries 13, 26, 35, 39).

Evidently, a compelling, but complex dependence of the bis-

4

.2. Catalytic Activity of Library I. The half-life data for library I

4

(

trifluoromethyl)phenyl group at R on the catalyst activity is

is summarized in Table 1. These ammonium ions showed fairly

poor catalytic activity exhibiting half-lives in the range of 5ꢀ12 h.

The one exception was the catalyst containing an n-hexyl

nitrogen substituent. In this case, the half-life observed was only

3

observed. Moreover, ammonium salts containing R = n-hexyl

are generally more active than catalysts with R

other substituents held constant, which suggests a dependence

on lipophilicity.

3

6¼ n-hexyl with all

1

3 min.

4

269

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Table 2. Half-life Data for Library II and III

that most strongly affect enantioselectivity were of equal

interest. Summarized below is the enantioselectivity data for

all 143 chiral, nonracemic catalysts constructed to date. The

enantiomeric ratio of benzylated product 20 was measured by

purifying a sample by silica gel chromatography from a separate

experiment at 2ꢀ4 °C with 2.5 mol % catalyst loading.

The enantiomeric ratios were determined by CSP-HPLC by a

8

0

known protocol.

.1. Enantioselectivity for Libraries II and III. The enantios-

3

4

entry library no.

R

C(1) conﬁguration t1/2, min

5

1

2

1

2

3

4

5

6

7

8

9

0

1

2

3

4

5

6

7

8

9

0

1

2

3

4

5

6

7

II{1,1}

II{1,2}

II{1,4}

II{1,5}

II{1,3}

III{1,3}

II{2,1}

II{2,2}

II{2,3}

H

C

6

H

5

R

β

181

179

298

346

245

72

electivity data for the R-series (library II, R = R = H) and the

β-series (library III, R = H) are summarized in Table 5. For the

2

1-naphthyl

9-anthryl

R-series (library II), little to no stereoinduction was observed,

regardless of the combination of substituents (entry 1). In the

3 2 6 3

3,5-(CF ) C H

3

β-series (library III), catalysts with R = H produced low to

2-naphthyl

4

moderate enrichment of the S enantiomer of 20 with R =

3

,5-(CF ) C H (entry 5, er 57:43) exhibiting the largest selec-

3 2 6 3

3

C

6

H

5

H

5

H

5

C

6

H

5

R

β

10

tivity. Relative to the R-series, catalysts with R = aliphatic con-

sistently produced the S enantiomer in slightly greater enrich-

ment (entries 8ꢀ13). Additionally, moderate enantioselectivity

was observed in catalysts containing a 2-pyridyl group (where

the point of attachment is the sp carbon of the heteroaromatic

ring) at R (entries 20, 23, and 25) with the largest selectivity

observed when R = 1-naphthyl. Similar results were observed

in catalysts with R = benzyl within this series, apart from

when R = Ph (entry 19) which resulted in decreased selec-

6

1-naphthyl

2-naphthyl

18

12.1

1

2

II{3,1} n-hexyl

III{3,1} n-hexyl

C

6

H

5

11.7

14.8

6.6

2

6

3

II{3,2} n-hexyl 1-naphthyl

III{3,2} n-hexyl 1-naphthyl

III{3,3} n-hexyl 2-naphthyl

R

β

4

9.5

3

β

13.5

12.7

4.6

4

II{3,5} n-hexyl 3,5-(CF

II{3,6} n-hexyl n-hexyl

III{3,6} n-hexyl n-hexyl

3

)

2

C

6

H

3

R

β

tivity. Overall, the enantioselectivity data for the catalysts in

library III were greater than their diastereomeric counterparts

in library II.

5.5

II{4,1} MEM

II{4,2} MEM

II{4,3} MEM

C

6

H

5

R

β

17.5

12.3

22.8

17.1

11.1

6.6

5

.2. Enantioselectivity for Libraries IV and V. The enantios-

2 3 4

1-naphthyl

2-naphthyl

electivities for catalysts with variable R , R , and R groups while

holding R constant as a methyl group (excluding entry 1) are

1

II{5,1} 2-pyridyl C

III{5,1} 2-pyridyl C

6

H

5

summarized in Table 6. See the Supporting Information for a

fully tabularized list. Entries with the highest selectivities are

6

1

II{5,2} 2-pyridyl 1-naphthyl

III{5,2} 2-pyridyl 1-naphthyl

II{5,3} 2-pyridyl 2-naphthyl

II{5,4} 2-pyridyl 9-anthryl

R

β

shown in bold type. Introducing a methyl group at R while

2

3

maintaining a hydrogen and a benzyl group at R and R ,

respectively, did not lead to significant differences in the

11

R

12.7

21.1

31.8

4

enantioselectivity apart from when the R group was a 3,5-

substituted aryl ring (entries 2ꢀ4). In this case, inversion of the

II{5,5} 2-pyridyl 3,5-(CF

3

)

2

C

6

H

3

C(1) configuration resulted in a comparable enantioselectivity

(

entries 5 and 6).

2

4

.5. Other Quaternary Ammonium Ions. Varying the substit-

The results for the catalysts with R = Me are similar to those

2

uents on a common scaffold allowed for observation of a large

range of rates. To determine how much of the “total possible”

range of activity was being sampled, tetramethylammonium

bromide (TMAB, q = 4.0) was tested and exhibited a half-life

of 1200 min (Table 4). The similarly hydrophilic, but less

accessible, tetraethylammonium bromide (q = 2.0) exhibited a

catalysts with R = H. Likewise, catalysts with a 3,5-substitution

pattern on the aryl ring of R consistently resulted in the highest

selectivities independent of the R substituent (compare entry 7

4

3

2

to entries 8ꢀ12). Increasing the size of the R substituent to

78

isopropyl and tert-butyl resulted in a reversal in the selectivity

2

(entries 13). As was the case with catalysts with R = Me, the

2

0-fold rate increase over TMAB. The more lipophilic, but

catalysts exhibiting a 3,5-substitution pattern on the aryl ring of

4

2

similarly accessible cation cetyltrimethyl (C₁₉total, q = 3.06)

ammonium bromide was only three times faster than TMAB

R with R = i-Pr and t-Bu produced the highest selectivities

2

(entry 14). Catalysts containing R = phenyl were variably

4

(entry 3). The rate increase from tetramethyl- to tetraethylam-

selective for either the S or R enantiomer, apart from when R

monium bromide was also seen in the use of tributylbenzylam-

monium bromide, which caused a further 20-fold rate increase.

To directly address this dramatic effect of ammonium accessi-

bility with catalysts more closely related to the common scaffold,

one other small set of quaternary ammonium ions was con-

structed that replaces one of the pyrrolidine rings of the scaffold

with an azetidine. As expected, the ammonium ions containing

an azetidine ring exhibited overall poor catalytic activity with half-

lives ranging from 900 to 1000 min.

was a 3,5-disubstitued aryl group (entries 15 and 16). However,

the maximum enantioselectivity was observed when R = 3,5-

4

(CF )C H (entries 19, 20, 22, and 25). This eﬀect is indepen-

3

6 3

3

dent of the oxygen substituent (R ). Additionally, the substitu-

tion pattern of the triﬂuoromethyl groups on the aryl ring (R )

4

proved to be important, as having groups at the 3- and 5- aryl

positions consistently resulted in greater enantioselectivity than

having substituents at the 2- and 4-aryl positions or just the 4-aryl

position (entry 23 and 24). Moreover, the presence of an alkyl

7

9

1

5

. Enantioselectivity of the Catalysts. In addition to deter-

group at R (Me) is imperative as the analogous catalyst with

1

mining the substituent effects on rate, those structural features

R = H aﬀorded signiﬁcantly decreased enantioselectivity (entry 1).

4

270

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Table 3. Half-life Data for Library IV and V

1

2

3

4

entry

library no.

R

C(1) conﬁguration

t

1/2, min

1

2

3

4

5

6

7

8

9

IV{2,2,1}

IV{2,2,2}

IV{3,2,1}

IV{3,2,5}

IV{5,2,5}

V{1,2,1}

V{1,2,8}

V{1,2,7}

V{1,2,6}

V{2,2,1}

V{2,2,3}

V{2,2,5}

V{2,2,8}

V{2,2,7}

V{2,2,6}

V{2,3,1}

V{2,3,2}

V{2,3,3}

V{2,3,5}

V{2,3,8}

V{2,3,7}

V{2,3,6}

V{2,7,1}

V{3,2,1}

V{3,2,5}

V{3,2,7}

V{3,3,1}

V{3,3,3}

V{3,3,5}

V{3,3,8}

V{3,2,6}

V{3,7,5}

V{3,7,8}

V{4,2,5}

V{5,2,1}

V{5,2,2}

V{5,2,3}

V{5,2,5}

V{5,2,8}

V{5,2,9}

V{5,2,10}

V{5,3,1}

V{5,3,2}

V{5,3,3}

V{5,3,5}

V{5,3,8}

V{5,3,7}

H

Me

i-Pr

C

6

H

5

C

6

H

5

R

β

93.5

51.1

19.5

26.1

16.3

32.5

51.8

53.9

9.78

24.7

20.7

23.0

899.7

19.9

15.1

28.1

23.0

75.8

30.7

236.0

26.5

28.9

21.1

17.9

132.1

122.1

28.6

41.8

20.3

62.4

42.7

86.0

60.9

249.4

95.8

10.6

12.2

10.3

9.5

H

1-naphthyl

H

6 5

C H

H

3,5-(CF

3

)

2

C

6

H

3

H

C

6

H

5

3,5-(CF

3

)

C

H

Me

H

6 5

C H

4-(MeO)C

6

H

4

4-(MeO)C

6

3,5-(t-Bu)

2

C

6

H

3

β

1

2

3

4

0

1

2

3

4

5

6

7

8

9

0

1

2

3

4

5

6

7

8

9

0

1

2

3

4

5

6

7

8

9

0

1

2

3

4

5

6

7

8

n-hexyl

R

Me

i-Pr

t-Bu

6 5

C H

2-naphthyl

3,5-(CF

4-(MeO)C

3 2 6 3

) C H

6 4

H

3,5-(t-Bu)

2 6 3

C H

n-hexyl

6 5

C H

1-naphthyl

2-naphthyl

3,5-(CF

3 2 6 3

) C H

4-(MeO)C

6 4

H

3,5-(t-Bu)

2 6 3

C H

n-hexyl

4-(MeO)C

6

H

4

C

6

H

5

C

6

H

6

H

6

H

5

C

6

3,5-(CF

3 2 6 3

) C H

3,5-(t-Bu)

2 6 3

C H

n-hexyl

6 5

C H

2-naphthyl

3,5-(CF

)

3 2

C

6

H

3

4-(MeO)C

6

H

4

C

6

H

5

n-hexyl

4-(MeO)C

6

H

4

3,5-(CF

4-(MeO)C

3,5-(CF

)

3 2

C

6

H

3

4-(MeO)C

H

6

H

4

C

6

H

5

)

3 2

C

6

H

3

C

6

H

5

6 5

C H

1-naphthyl

2-naphthyl

3,5-(CF

4-(MeO)C

(4-CF )C

(4-CN)C

)

3 2

C

6

H

3

6

H

4

17.5

16.0

14.4

9.5

6

5

6

5

3

6 4

H

6 4

H

n-hexyl

6 5

C H

1-naphthyl

2-naphthyl

18.7

13.1

4.9

3,5-(CF

4-(MeO)C

3,5-(t-Bu)

3 2 6 3

) C H

6

5

6

H

4

9.9

2

C

6

H

3

8.5

4

271

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Table 3. Continued

1

2

3

4

entry

library no.

R

C(1) conﬁguration

t

1/2, min

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

V{5,6,1}

V{5,6,5}

V{5,6,8}

V{5,7,1}

V{5,7,2}

V{5,7,3}

V{5,7,5}

V{5,7,7}

V{7,2,1}

V{7,2,5}

V{7,2,8}

V{7,2,9}

V{7,2,10}

V{6,2,1}

V{6,2,8}

Me

C

6

H

5

3,5-(CF

3

)

2

C

6

H

3

C

6

H

5

R

7.9

7.0

3,5-(CF

3

3,5-(CF

3 2 6 3

) C H

3,5-(CF

4-(MeO)C

6

H

4

6.9

4-(MeO)C

6

H

4

C

6

H

5

19.9

19.2

12.0

18.6

14.6

13.7

28.7

11.3

11.8

48.6

12.2

11.8

1-naphthyl

2-naphthyl

3,5-(CF

)

3 2

C

6

H

3

3,5-(t-Bu)

2

C

6

H

3

1-naphthyl

C

6

H

6

H

6

H

6

H

6

H

6

H

6

H

5

6 5

C H

3,5-(CF

4-(MeO)C

(4-CF )C

(4-CN)C

)

3 2

C

6

H

3

6

H

4

3

6 4

H

6 4

H

2,4,6-(CH

3

)

2

C

6

H

2

6 5

C H

2,4,6-(CH

3

)

2

6

4-(MeO)C H

4

a

Table 4. Half-life Data for other Quaternary Ammonium Ions

Table 5. Enantioselectivity of Libraries II and III

3

4

entry

library no.

R

er, S/R

entry

library no.

catalyst

R

t

1/2, min

1

2

II{1ꢀ5,1ꢀ6}

III{1,2}

III{1,3}

III{1,4}

III{1,5}

III{1,6}

III{1,1}

III{3,2}

III{3,3}

III{3,4}

III{3,5}

III{3,6}

III{3,1}

III{2,2}

III{2,3}

III{2,4}

III{2,5}

III{2,6}

III{2,1}

III{5,2}

III{5,3}

III{5,4}

III{5,5}

III{5,6}

III{5,1}

50:50 ( 3

48:52

55:45

50:50

57:43

52:48

55:45

48:52

55:45

47:53

52:48

53:47

54:46

64:36

48:52

44:56

57:43

58:42

53:47

64:36

48:52

54:46

58:42

57:43

60:40

1

2

3

4

5

6

7

Me

Et

Me

n-Bu

4

N

12,000

480

H

1-naphthyl

2-naphthyl

9-anthryl

4

N

3

NC16

H

33

2800

21

4

3

NBn

5

3,5-(CF

3

)

2

C

6

H

6

H

6

H

6

H

3

VI{1}

VI{2}

VI{3}

C

6

H

5

890

6

n-hexyl

1-naphthyl

2-naphthyl

1000

1090

7

6 5

C H

8

n-hexyl

1-naphthyl

2-naphthyl

9-anthryl

9

2

The installation of diﬀerent groups at R with π-surfaces while

maintaining R = 3,5-(CF )C H produced similar results

entries 25 and 26). When R = isopropyl or tert-butyl

entreis 13 and 14), the enantioselectivity is signiﬁcantly

diminished in comparison to catalysts that contain an aromatic

group is in this position. Together these results attest to the

10

11

12

13

3

3,5-(CF

n-hexyl

C H

3 2

)

3

6

3

2

(

6

5

1

2

4

5

6

7

8

9

0

1

C

6

H

5

1-naphthyl

2-naphthyl

9-anthryl

2

importance of π-surface rather than steric bulk at R for higher

enantioselectivity.

3 2

3,5-(CF )

n-hexyl

6 5

C H

’

DISCUSSION

. Analysis of Ammonium Ion Preparations. 1.1. Synthetic

2-pyridyl

1-naphthyl

2-naphthyl

9-anthryl

1

Strategy. Since the discovery of asymmetric phase-transfer cata-

lysis with cinchona alkaloid derivatives, a significant level of effort

has been devoted to the introduction of novel synthetic catalyst

structures. Most of the synthetic endeavors fall into one of the

following two synthetic strategies: (1) elaboration of a readily

available source of chiral material by appending a nonstereogenic

quaternary ammonium ion or (2) incorporation of a quaternary

ammonium into a molecule in such a way that it lies on a

symmetry axis. Although each of these approaches has seen some

22

23

24

25

3,5-(CF

n-hexyl

C H

3 2

)

6

5

a

This is a representative summary. For a full tabular listing, see the

Supporting Information.

success, the synthetic strategy presented herein is significantly

different and warrants discussion.

4

272

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

a

Table 6. Enantioselectivity of Libraries IV and V

1

2

3

4

entry

library no.

R

C(1) conﬁguration

er, S/R

1

2

3

4

5

6

7

8

9

IV{1,2,5}

V{1,2,X}

V{1,2,5}

V{1,2,7}

V{1,2,2}

V{1,2,5}

V{2,X,X}

V{2,3,7}

V{2,2,5}

V{2,2,7}

V{2,7,5}

V{2,7,7}

V{3,X,X}

V{4,2,5}

V{5,X,X}

V{5,3,5}

V{5,3,7}

V{5,2,X}

V{5,2,5}

V{5,6,5}

V{5,X,X}

V{5,7,5}

V{5,7,7}

V{X,2,X}

V{7,2,5}

V{6,2,5}

H

Ph

H

C

6

H

5

3,5-(CF

1ꢀ3,6 or 8

3,5-(CF

3,5-(t-Bu)

3

)

2

C

6

H

3

R

β

56:44

Me

50:50 ( 4

56:44

H

)

3 2

C

H

2

C

6

H

3

60:40

H

1ꢀ3 or 6ꢀ8

50:50 ( 5

63:37

H

3,5-(CF )

3 2

C

6

H

3

β

Me

i-Pr

t-Bu

2,3,7

n-hexyl

1ꢀ6,8

3,5-(t-Bu)

R

r

R

r

R

r

R

r

50:50 ( 4

58:42

2

C

6

H

3

C

6

H

5

3,5-(CF

3,5-(t-Bu)

3,5-(CF

3

)

2

C

6

H

3

57:43

1

2

0

1

2

3

4

5

6

7

8

9

0

1

2

3

4

5

6

2

C

6

H

3

57:43

4-(MeO)C

2,3,7

6

H

4

)

3 2

C

6

H

3

59:41

6

3,5-(t-Bu)

1ꢀ8

2

C

6

H

3

57:43

41:59 ( 5

37:63

C

6

H

5

3 2 6

3,5-(CF ) C H

3

C

6

H

5

2,3

n-hexyl

1,2,3,8

50:50 ( 6

79:21

H

5

3 2 6 3

3,5-(CF ) C H

H

C

6

C

6

C

6

H

5

3,5-(t-Bu)

2

C

6

H

3

61:39

H

2,3

50:50 ( 5

81:19

H

5

H

5

3,5-(CF

1,2,3,8

3

)

2

C

6

H

3

H

3,5-(CF

3

)

2

C

6

H

3

2

C

6

H

81:19

H

5

6,7

50:50 ( 5

81:19

H

5

4-(MeO)C

6

H

4

3,5-(CF

3

)

2

C

6

H

3

H

5

4-(MeO)C

6

H

4

3,5-(t-Bu)

2

C

6

H

3

63:37

5,7

C

6

C

6

C

6

H

5

1,8,9,10,11

50:50 ( 6

81:19

1-naphthyl

mesityl

H

5

3,5-(CF

3

)

2

C

6

H

3

5

3,5-(CF

3

2

6

H

76:24

a

This is a representative summary. For a full tabular listing, see the Supporting Information.

The approach presented here is unique in that the synthetic

tional handle for two parallel synthesis steps, a Grignard addition

eﬀort was focused on systematically varying the steric and elec-

tronic environment around a central stereogenic quaternary

ammonium ion. To this end, the synthetic investment was

divided into two diﬀerent parts. The ﬁrst part involved pre-

paration of a nitrogen containing scaﬀold (a cyclopentapyr-

rolizidine) on a signiﬁcant scale which was readily accom-

plished by application of the tandem inter [4 þ 2]/intra

and an O-alkylation, leaving only N-quaternization as the ﬁnal

parallel synthesis step. Including the previously installed R

1

substituent, a total of four variable groups were introduced.

The diversity and number of groups utilized was greatest for the

2

3

positions that could be incorporated in parallel (R 6; R , 7;

,

4

R , 11) and the least for the group that required a recast skeleton

(R , 2). A good appreciation for the shape of the catalyst(s) and

1

[

3 þ 2] cycloaddition of nitroalkenes with only minor changes

disposition of the variable groups in relation to the ammonium

nitrogen is necessary to facilitate a thorough analysis of the eﬀect

of each group (and combinations thereof) on catalytic activity

and selectivity.

to the previously published routes. Notably, the tandem

cycloaddition also served as a diversifying element in that it

was in this step that the conﬁguration at C(1) was set (libraries

1

II and III) and the R group was introduced stereoselectively,

1.2. Catalyst Shape and Position of Groups. Library I is a

logical starting point because the inherent symmetry simplifies

the analysis. Figure 8 shows a plot of the relative conformer ener-

gies as a function of a double dihedral driver about bonds

C(5a)ꢀC(6)ꢀC(7)ꢀC(7a) and C(7b)ꢀN(3)ꢀC(9)ꢀC(10)

which subsequently proved to be a critical catalyst structural

feature.

The second synthetic component involved development of

procedures amenable to parallel synthesis that introduced a

variety of groups in the vicinity of the ammonium nitrogen.

Ultimately, three operationally simple bond-forming reactions

were utilized to accomplish this goal that allowed for the catalysts

to be prepared in a parallel fashion. The tandem cycloaddition

naturally installs a hydroxyl group at C(1) which served as a func-

81

for catalyst I{1}. A full 360 degree rotation about the C(7b)ꢀ

N(3)ꢀC(9)ꢀC(10) bond constitutes a full range of motion

4

of the R substituent, in this case, a phenyl group. The R

substituent can be found in two local sparsely populated minima

corresponding to gauche conformations (( 60°), which, in the

4

273

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

case of library I, are a pair of diastereomeric conformers. A single,

4

highly populated, global minimum is found when the R sub-

stituent is projected 180° away from the N(3)ꢀC(9) bond. A

small energetic barrier separates the global and local minima, but

4

a “full range of motion” of the R substituent is prohibited

because of steric interactions with the C(8)- methyl group. For

4

the remainder of this discussion the R group will be projected in

its global minimum of ∼180°.

Figure 8. Conformer energy as a function of NꢀC(9) rotation and

C(6)ꢀC(7) bond rotation.

Figure 10. Structural clustering of kinetic data.

Given the large volume of data presented herein, a concise sta-

tistical summary provides a good indication of which data is the

most interesting and worthy of detailed analysis. The relative

contribution of the groups to the variation in the observed data is

estimated by examining the standard deviation as a function of

Substitution of the ring system in any way removes the

symmetry plane in the scaﬀold of library I; therefore, all catalysts

(except library I) consist of at least six diastereomeric confor-

mers. Because only one or two of these conformers are highly

populated, in any projections or analysis from this point on the

lowest energy conformer of the ring system will be depicted. The

remaining three variable groups each reside in diﬀerent faces of

1

4 82

each group (R ꢀR ). That is, the larger the deviation in the

observed rate (or selectivity) as a function of group positional

substitution, the greater the relationship of that substituent

2

4

1

2

1

2

(R ꢀR ) to the observed eﬀect (rate or selectivity, Figure 11).

Therefore, the variation approximates the sensitivity of the rate

and selectivity to a structural change at the indicated substituent.

Both catalyst activity and rate are inﬂuenced the most by the

the central ammonium nitrogen (R and R ). The R and R

groups are placed directly on the right and left convex faces

3

(

Figure 9), and the R group resides in the concave face of the

scaﬀold (see Figure 12).

2

4

groups R and R , with the greatest dependence on R . Also, in

3

1

both cases, the R group has the least inﬂuence. The R group was

not suﬃciently varied for interpretation by this analysis.

Figure 9. Representation of the convex faces of the ammonium scaﬀold.

2

. Summary of Results. A brief summary of the rate and enan-

tioselectivity data is necessary to facilitate the following discus-

sion. The kinetic results are summarized in bar graph format in

Figure 10. In general:

•

Catalysts that bear an oxygen substituent at C(1) are more

active than catalysts without.

•

The conﬁguration of the oxygen functional group does not

inﬂuence the catalyst activity.

Figure 11. Variation in rate and selectivity as a function of positional

group substitution.

•

Ethers are more active than the corresponding alcohols.

4

Strong electron-withdrawing groups (e.g., R = (3,5-CF ) -

3

2

C H ) on nitrogen decrease catalyst activity.

6

3

•

Less dependence on the nitrogen substituent was observed

when R was aromatic (Ph, mesityl, 1-naphthyl).

The catalytic activities of the unfunctionalized ring systems

are not as high as tetraethylammonium but better than

cetyltrimethylammonium in activity.

3. Catalyst Activity. 3.1. Effect of an Oxygen Substituent. The

presence of an oxygen at C(1) is a natural consequence of the

tandem cycloaddition utilized to construct the scaffolds. Indeed,

a β-oxygen substituent is a common structural motif employed

in many asymmetric quaternary ammonium phase-transfer

2

4

274

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

8

3

1

catalysts including the cinchona alkaloids. Inclusion of an

oxygen substituent β to the ammonium center strongly affects

the catalytic activity (compare libraries I and II), which has

when R = H, our data seems most consistent with a

selective “docking” of the enolate to the left convex face

of the ammonium. That is, the oxygen serves to increase

84,85

1

previously been reported for other catalyst systems.

A more

the positive potential on the left convex face and the R

quantitative treatment of β-oxygen stubstitution is presented in

the accompanying paper (DOI 10.1021/jo2005457), but a quali-

tative analysis, as follows, is required to analyze the differences in

catalyst activity in this data set.

Introduction of an oxygen atom two carbons removed from

the ammonium center results in two related electronic eﬀects:

group acts to sterically shield the right convex face of the

ammonium ion.

3.2. Accessibility of the Ammonium Ion. The catalytic activity

of PTC alkylation processes has been correlated with the

accessibility of the ammonium nitrogen (in terms of q), where

4

9

an optimum accessibility is found. The observation of a

maximum catalyst activity as a function of ammonium acces-

sibility has two similar explanations. In the first, the accessi-

bility of the ammonium nitrogen is related to the rate of

(

1) it inﬂuences the direction of the dipole and (2) it modulates

the magnitude of positive electrostatic potential interaction

(

þ

δ ). The electrostatic potential maps of the oxygenated and

4

7

unoxygenated catalyst scaﬀolds are compared in Figure 12. In the

C_ssymmetric ammonium salts (library I), the charge distribution

is equally dispersed between the two convex faces. In contrast,

inclusion of an alkoxy group two carbons removed from the

ammonium nitrogen results in a considerable polarization of the

positive potential toward the face to which it is attached (library

II, methoxy is included for simplicity). In library II type ammo-

nium ions the two convex faces are diﬀerentiated electronically,

placing a greater positive potential on the left face. Similarly, the₄

dipole vector in the library I type ammonium ions bisects the R

substituent but is rotated ∼60° clockwise in the library II ammo-

nium ions projecting the positive end toward the left convex face

of the ammonium ion.

exchange of anions, a kinetic phenomenon. In the second,

the accessibility of the ammonium nitrogen is related to

the ability of the catalyst to decrease the interfacial tension

and thereby facilitate enolate transfer, a thermodynamic

4

6,86

phenomenon.

The data collected herein are largely consistent with these

proposals, with the added complexity that the magnitude of

þ

the exposed positive potential (δ ) should be considered

as well. On the basis of the analysis above, the accessibility

þ

of the positive potential (δ , left convex face) should be

2

correlated with the steric bulk of the R substituent. The

sensitivity of the catalytic activity to the R substituent dis-

2

cussed above (Figure 11) advocates a strong dependence of the

overall catalyst activity on the nitrogen accessibility (steric

2

nature of R ).

The dependence of the catalytic activity on the size of the

2

4

R substituent and electron-withdrawing ability of the R sub-

stituent is summarized tabularly in Table 7 (MR = molar

87

refractivity). The dependence is weak when the electronic

4

character of R is neutral or electron rich (R = C H ,

6

5

4

-MeOC H ) but strong when R is strongly electron-

6 4

withdrawing (3,5-(CF ) C H ). In these catalysts, an

3

2

6

3

2

increase in the size of the R substituent (Me to i-Pr to

t-Bu, to Ph) leads to a signiﬁcant rate enhancement, which is

somewhat attenuated for larger aromatic groups (mesityl or

naphthyl). This disparity may be rationalized by ammonium

2

accessibility, such that catalysts with R = Ph have an

optimum accessibility.

1

Table 7. Half-lives (min) of Selected Catalysts with R = Me,

3

2

4

R = Ph, and Variable R and R

.

Figure 12. Electrostatic potential (ESP) maps (M06ꢀ2X/6-31G(d))

of scaﬀolds for Library I (unoxygenated) and Library II (oxygenated)

where C(6)ꢀC(7) and N(3)ꢀC(9) dihedrals are constrained to 0°

in the geometry optimization to exclude bias of a speciﬁc twist

conformer (Figure 11). For computational simpliﬁcation, a methyl

3

group is used at R . The directions of the dipoles are illustrated. ESPs

3

are mapped onto electron density isosurfaces (0.002 electrons/au ).

Legend is in units of kJ/mol. The diﬀerence in the maximum

electrostatic potential energies of the left and right faces of the

oxygenated scaﬀold is 3.0 kcal/mol and 0 kcal/mol for the un-

4

R

2

a

oxygenated scaﬀold (C

s

symmetric).

R

MR

C

6

H

5

3,5-(CF )

3 2

C

6

H

3

6 4

4-(MeO)C H

Me

6.88

16.08

20.85

25.28

42.97

42.45

21

900

122

96

18

i-Pr

132

174

This analysis is consistent with a stronger electrostatic inter-

action with the left convex face than the right, which in turn, is

consistent with the experimentally observed fact that the

t-Bu

phenyl

11

12

14

10

18

12

11

2

mesityl

62

R substituent has a large eﬀect on the observed rate and

selectivity. Since both rate and selectivity are highly depen-

dent on the R group and little to no selectivity is observed

1-naphthyl

29

2

a

2

MR = molar refractivity of the R substituent calculated by Chemdraw.

4

275

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

2

Rationalizing the eﬀect of the R substituent in terms of steric

factors that control the topicity of the enolate reactivity are

diﬃcult to predict, the factors that dictate the relative binding

strengths of the enolate to each of the tetrahedral faces are

more easily controlled and predicted. Initially, these factors

were the primary focus for the design and construction of the

libraries.

bulk makes intuitive sense because placing groups in this position

modulates the relative accessibility of the positive potential at

the two convex faces. However, other explanations are possible.

For example, if the major structural perturbation of replacing a

branched aliphatic group with an aromatic group is the introduc-

tion of a π-surface then the increase in catalyst activity could be

caused by an aﬃnity of catalyst for substrate (through πꢀπ

interactions). In this scenario, a shift in the equilibrium between

catalyst X (Br or HO ) and catalyst enolate ion pair

toward the enolate complex would be the origin of the net

increase in observed catalyst activity.

Most consistent with the above analysis of charge polariza-

tion, the 3,5-(CF ) C H group must impart suﬃcient charge

polarization to amplify the eﬀect of the variable R group.

Interpretation of these observations in the context of a tran-

sfer rate limiting regime has led to the conclusion that an

optimum ammonium accessibility, or charge exposure is likely.

In other words, a greater charge exposure will increase the

ammonium ions association with the anionic hydroxide sur-

face. However, if the exposed δþ area or charge density is too

great, then the catalyst will not dissociate away from the

interface as readily.

The observed enantioselectivities are largely dependent

on the substituents that inﬂuence the binding of the anion

to the face of the hypothetical tetrahedron around the ammo-

nium ion with the largest concentration of positive potential

þ

ꢀ

þ

ꢀ

2

4

(

R and R , Figure 12). The observation that the enantio-

selectivity is greatest for catalysts that bear strongly electron

withdrawing groups on the nitrogen is consistent with the

need for one or both of the following potential interactions:

3

2

6

3

2

(

1) R-CH-hydrogen bonding or (2) a tighter ion pair resulting

from increased Coulombic interaction. Additionally, a π-surface

is necessary at the R substituent, presumably to engage in

2

π-stacking interactions with the phenyl rings of the reacting

1

enolate. Moreover, the presence of an alkyl group at R is

necessary to decrease the accessibility of the right-hand face

of the pyrrolizidine moiety because in its absence, the enantio-

selectivity is poor. This observation suggests that binding

to the right-hand face leads to lower selectivity because of

the pseudoenantiotopic local chirality. The aforementioned

interactions seem to operate in concert as the absence of

one of the interactions leads to signiﬁcantly diminished

enantioselectivities.

3

.3. Summary of Catalyst Activity Discussion. Three impor-

tant elements can be inferred from the discussion of rate:

(

1) The observed increase in catalyst activity upon the in-

clusion of an oxygen substituent is attributed to the

increase in the rate of ion pair formation and/or a dec-

rease in interfacial tension

’

CONCLUSIONS

2) The observed decrease in catalyst activity upon the

4

inclusion of the 3,5-(CF ) C H group at R for

A synthetic strategy for the synthesis of diverse libraries of

3

2

6

3

2

small groups at R (H, Me, i-Pr) is attributed to

a diminished tendency of the ion-pair to transport

from the interface to the organic phase for alkyla-

tion due to greater charge density at the ammonium

center.

quaternary ammonium ions has been developed. The key

feature of the synthetic strategy was to divide the preparative

work into two distinct stages: (1) scaﬀold preparation and (2)

diversity-oriented parallel synthesis. In this way, a total of 160

structurally diverse quaternary ammonium ions were prepared

that share a common scaﬀold constructed by a tandem inter

[4 þ 2]/intra [3 þ 2] cycloaddition of a nitroalkene. A method

was developed for the collection of kinetic data of a biphasic

reaction that is applicable over a wide range of catalyst activities

(half-lives ranging from days to minutes). The range of data

collected covers many orders of magnitude and therefore is well

suited for analysis by the application of quantitative structureꢀ

activity relationships. Inclusion of an oxygen atom in the

vicinity of the quaternary ammonium ion aﬀects the catalytic

activity. The catalyst enantioselectivity is strongly dependent on

the substituent attached to the same carbon as the oxygen atom.

These observations were rationalized in terms of a selective

polarization of the positive potential on one of the faces of

the tetrahedral ammonium over the other. The inclusion of a

(

3) The observed increase in catalyst activity upon the inclu-

2

4

sion of aryl groups at R with R = 3,5-(CF ) C H is

3

2 6 3

attributed to an enhanced ability of the ion-pair to trans-

port from the interface into the organic phase for alkyla-

tion because of an optimum surface exposure of the

ammonium ion (kinetic) or a change in enolate binding

equilibrium (thermodynamic).

4

. Enantioselectivity. The intermolecular forces that have

been proposed to contribute to enantioselectivity in APTC

ꢀ

reactions include (1) ROH

OꢀCRdCR hydrogen bond-

3

3 3

2

0a,c

þ

1b

ꢀ

ing,

gen bonding),

(2) R N ꢀCHRꢀH

OꢀCRdCR (R-CH hydro-

3

3 3 3

2

5

20a,c

and (3) πꢀπ interactions.

The most

selective catalysts in this study do not contain OH hydrogen-

bond-donating sites; therefore, intermolecular force (1)

can be eliminated as a stereocontrolling element. The results

here support the operation of interactions (2) and poten-

tially (3).

3

,5-substituted aromatic substituent on nitrogen proved crucial

for catalyst enantioselectivity. With strongly electron-withdraw-

ing 3,5-triﬂuoromethyl groups, the previously observed depen-

dencies on rate were ampliﬁed. The proposed dependencies of

rate on the ammonium accessibility are consistent with the data

reported herein with the added complexity that the magnitude of

the “ammonium charge” or charge density should be considered

as well. Quantitative models have been developed to describe

both the reactivity and selectivity trends discussed herein and

constitute the focus of the following paper (DOI 10.1021/

jo2005457).

The unique structural features of the catalyst scaﬀold include

the intrinsic shielding of two of the four faces of the imaginary

tetrahedron encompassing the nitrogen and the capacity to dif-

ferentiate the two remaining exposed faces (cf. Figure 12). The

second feature arises from the modular nature of the cyclo-

addition-based construction, which allows both exposed front

faces to be diﬀerentiated electronically (β-oxygenation) and

1

2

sterically (relative steric bulk of R and R ). Although those

4

276

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

’

EXPERIMENTAL SECTION

rubber septum, and an internal temperature probe (through rubber

septum). To the addition flask were added diisopropylamine (36.5 mL,

I. Preparative Studies: Tandem Cycloaddition Precursors and

Scaﬀolds

209 mmol, 1.3 equiv) and THF (300 mL). The resulting solution was

cooled to ꢀ78 °C (bath temperature) by addition of an acetone/CO2(s)

A. Scale up of Alcohol (Z)-S6

B. Preparation of Nitroalkenes and Chiral Vinyl Ethers

bath to the jacket reservoir. Then, n-butyllithium (2.56 M in hexanes,

8

1.8 mL, 1.3 equiv) was added dropwise over 20 min. To the 1-L, two-

necked flask were added ε-caprolactone (18.4 g, 161 mmol), THF

150 mL), and trimethylsilyl chloride (26.5 mL, 209 mmol, 1.3 equiv)

via syringe. The reaction vessel was brought to <ꢀ80 °C by immersion

in a hexanes/N bath. The freshly generated solution of LDA (above)

1

2

. Nitroalkene Preparations

. Preparation of Chiral Vinyl Ether 2

(

C. Scaﬀold Preparation by Tandem Cycloaddition of Nitroalk-

enes 1

2

1. Preparation of Scaﬀold for Library I

2. Preparation of Scaﬀold for Library II

3. Preparation of Scaﬀold for Library III

4. Preparation of Scaﬀold for Library IV

5. Preparation of Scaﬀold for Library V

6. Preparation of Scaﬀold for Library VI

was added dropwise at a rate that maintained an internal tempera-

ture <ꢀ75 °C (ca. 1 h). The resulting turbid solution was allowed

to warm to ꢀ25 °C and was then concentrated by rotary evaporation

(

15 mmHg, 20ꢀ25 °C). The resulting mixture was triturated with

pentane (∼100 mL) and filtered (Celite). The filtrate was again

concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). Purification

bydistillation (0.5mmHg, 43 °C) afforded ketene acetal S1 (24.8 g, 83%)

as a clear colorless oil. The data collected were consistent with those

II. Parallel Syntheses: Library Intermediates and Quaternary Ammo-

nium Bromides

2

A. Variable Group R : Organometallic Additions to Ketones

8

1

previously reported. Data for S1: bp 58ꢀ60 °C (2ꢀ3 mmHg); H

1

2

3

4

. Cerium-Mediated Additions to Keto Amine 13

. Cerium-Mediated Additions to Keto Amine 14

. Grignard Additions to Keto Amide 17

NMR (500 MHz, CDCl ) 4.12 (t, J = 5.9, 1 H, HC(3)), 3.99 (dd, J = 5.4,

3

5

.6, 2 H, H

2

C(7)), 2.02 (dd, J = 5.9, 11.6, 2 H, H

C(6)), 1.65ꢀ1.59 (m, 2 H, H

13

2

C(4)) 1.83 (ddd, J = 5.7,

10.7, 11.5, 2 H, H

2

C(5)), 0.22 (s, 9 H,

. Borane Reduction of Hydroxy Amides 19

0

3

H C(5 )); C NMR (126 MHz, CDCl ) 160.0 (C(2)), 83.1 (C(3)),

3

B. Variable Group R : Williamson Ether Synthesis

0

71.5 (C(7)), 31.3 (C(4)), 26.2 (C(6)), 23.3 (C(5)), 0.16 (C(5 )) IR

1. General Procedure I for the Preparation of Library

(neat, NaCl plate) 2931 (s), 2837 (m), 1682 (s), 1455 (m), 1355 (m);

Intermediates

MS (ESI, Q-tof) 187 (21) [M þ 1], 171 (38), 147 (14), 132 (12), 129

2

3

4

. Preparation of Amino Borane Intermediates II{2ꢀ5}

3

(

13), 117 (30), 75 (100), 69 (32), 55 (23), 54 (23); TLC R 0.77

f

hexanes/Et O, 4:1) [UV, KMnO ].

2 4

. Preparation of Amino Borane Intermediates III{2ꢀ5}

. Preparation of Amino Borane Intermediates

IV{2ꢀ5,2ꢀ3}

5

. Preparation of Amino Borane Intermediates

3

V{2ꢀ7,2ꢀ7}

4

C. Variable Group R : Deborylation and N-Quaternization

1

2

. Parallel Synthesis Steps IIꢀIII: General Procedure II

. Preparation of Quaternary Ammonium Bromides

I{1ꢀ6}

Preparation of 6,7-Dihydro-5H-oxepin-2-one (S2). To a 1.0-L, one-

3

4

5

6

7

. Preparation of Quaternary Ammonium Bromides

necked, round-bottomed flask fitted with a reflux condenser, a magnetic

stir bar, and a nitrogen inlet adapter, fitted with a rubber septum, were

added sequentially palladium(0)bisdibenzylideneacetone (309 mg,

II{1ꢀ5,1ꢀ6}

. Preparation of Quaternary Ammonium Bromides

III{1ꢀ5,1ꢀ6}

0.537 mmol, 0.01 equiv), allyl methyl carbonate (12.2 mL, 107 mmol,

2.0 equiv), and acetonitrile (268 mL). Silyl ketene acetal S2 (1.0 g,

21.5 mmol) was then added via syringe. The resulting solution was

. Preparation of Quaternary Ammonium Bromides

IV{2ꢀ7,1ꢀ7,1ꢀ11}

. Preparation of Quaternary Ammonium Bromides

V{1ꢀ7,1ꢀ7,1ꢀ11}

placed in an oil bath preheated to 40 °C and allowed to stir. After being

stirred for 30 min at 40 °C, the reaction mixture was removed from

the oil bath and allowed to cool to room temperature. The solution

was concentrated to approximately one-quarter of the total volume

(∼70 mL). This solution was passed through a plug of silica gel

. Preparation of Quaternary Ammonium Bromides

VI{1ꢀ3}

For general experimental details and literature preparations, see the

Supporting Information.

I. Preparative Studies: Tandem Cycloaddition Precursors

and Scaffolds.

(20 mm ꢁ 5 cm) using hexanes/Et O (1:4, 3 ꢁ 50 mL). The filtrate

2

was concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C) to

afford 5.13 g (97%) of S2 as a pale-yellow oil. This material was carried

on to the next step without further purification. The data collected were

8

9

consistent with those previously reported. Data for S2: bp 70 °C

1

(

0.1 mmHg); H NMR (500 MHz, CDCl

3

) 6.41 (td, J = 4.4, 12.1, 1 H,

C(7)), 2.50 (dd,

HC(4)), 5.98 (d, J = 12.4, 1 H, HC(3)),4.28 (m, 2 H, H

J = 5.4, 10.2, 2 H, H C(5)), 2.12 (td, J = 6.6, 12.9, 2 H, H C(6));

2

13

C

2

NMR (126 MHz, CDCl

3

) 169.2 (C(2)), 144.0 (C(4)), 121.9 (C(3)),

0.20 (pentane/MTBE,

67.3 (C(7)), 30.1 (C(5)), 27.0 (C(6)); TLC R

:1) [KMNO₄].

f

1

A. Scale up of Alcohol (Z)-6

Preparation of Trimethyl(4,5,6,7-tetrahydrooxepin-2-yloxy)silane

(

S1). A jacketed 500 mL, three-necked, liquid addition flask fitted with

a nitrogen inlet adaptor and two rubber septa was fitted to a 1-L,

two-necked, round-bottomed flask fitted with a magnetic stir bar, a

4

277

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

Preparation of (Z)-Methyl 6-Hydroxy-2-hexenoate ((Z)-S3). To a

00 mL, one-necked, round-bottomed flask fitted with a rubber

150 (42), 149 (45), 93 (100), 91; TLC R

[UV, KMnO ].

f

0.57 (hexanes/EtOAc, 4:1)

1

4

septum, a large a magnetic stir bar, and a nitrogen inlet adaptor was

added ester S2 (4.5 g, 40.1 mmol) via syringe. Methanol (80 mL)

was then added followed by potassium carbonate (554 mg, 4 mmol,

0

.1 equiv). The suspension was stirred vigorously for 20 min, during

which time the solution gradually became opaque. The resulting

suspension was poured into a 500 mL separatory funnel containing

water (100 mL) and diluted with dichloromethane (100 mL). The

layers were separated, and the aqueous layer was extracted with

dichloromethane (5 ꢁ 50 mL). The combined organic extracts were

Preparation of Methyl (2E,6E)-7-Nitro-2,6-octadienoate ((E,E)-1). To

a 1-L, three-necked, round-bottomed flask fitted with two rubber

septa and a nitrogen inlet adaptor and an internal temperature probe

4

dried (MgSO ), filtered (cotton plug), and concentrated via rotary

evaporation (15 mmHg, 20ꢀ25 °C). The resulting clear oil was filtered

through a small plug of silica gel (1.8 cm ꢁ2 cm) with 25 mL of EtOAc to

remove any remaining base impurities. Concentration of the resulting

solution yielded 5.19 g (89%) of geometrically pure (Z)-S3 as deter-

(

through a septum) was added alcohol (E,E)-S3 (21 g, 147 mmol)

along with CH Cl and DMSO (147 mL each). Triethylamine (12 mL,

74 mmol, 6.0 equiv) was added, and the flask was immersed in a

ice/NaCl(s) bath. Once the internal temperature reached ꢀ5 °C,

SO pyridine complex was added in a single portion (29.4 g, 184

2

8

1

mined by H NMR analysis. The data collected were consistent with

90

1

3

those previously reported. Data for (Z)-S3: H NMR (500 MHz,

CDCl ) 6.24 (td, J = 8.3, 11.5, 1 H, HC(4)), 5.87 (td, J = 1.4, 11.5,

H, HC(5)), 3.72 (s, 3 H, H C(7)), 3.61 (t, J = 5.9, 2 H, H C(1)),

.74 (dt, J = 1.4, 8.2, 2 H, H C(C(3)), 1.73 (td, J = 8.3, 11.5, 1 H,

mmol, 1.25 equiv). After 2 h at <0 °C, an additional portion of

3

SO pyridine complex (29.4 g, 184 mmol, 1.25 equiv) was added.

3

1

2

3

2

After another 2 h, the reaction was quenched by pouring into a

-L separatory funnel containing 200 mL of satd aq NH Cl. The

2

1

4

13

H C(2)); C NMR (126 MHz, CDCl ) 167.5 (C(6)), 149.8 (C(4)),

2

3

organic phase was diluted with 150 mL of Et O and rinsed with

2

120.2 (C(5)), 61.1 (C(1)), 51.3 (C(7)), 31.2 (C(2)), 25.1 (C(3)); IR

NH Cl (3 ꢁ 100 mL) followed by CuSO (2 ꢁ 100 mL), brine

4

(neat) 3423 (br), 2950 (m), 2870 (m), 2343 (w), 2361 (w), 1723 (s),

(100 mL) and satd aq NaHCO

3

(100 mL). The combined organic

1

645 (m), 1439 (s), 1408 (m), 1202 (s), 1173 (s); MS (ESI, Q-tof)

extracts were dried (MgSO

4

), filtered over a silica gel plug (3 ꢁ 1 cm),

67 (100), 145 (35), 127 (15); TLC R 0.22 (hexanes/EtOAc, 4:1)

f

and concentrated to afford the crude aldehyde as a light yellow oil.

The aldehyde was carried on directly to the nitroalkene without

further purification as previously described to give 22.9 g (80%) of

nitroalkene (E,E)-1 as a light-yellow oil. The data collected were

4

[UV, KMnO ].

9

0

1

consistent with those previously reported. Data for (E,E)-1: H

NMR (500 MHz, CDCl ) 7.14 (t, J = 7.7, 1 H, HC(6)), 6.19

td, J = 11.4, J = 7.6, 1 H, HC(3)), 5.85 (d, J = 11.9, 1 H,HC(2)),

.70 (s, 3 H, H C(9)), 2.85 (m, 2 H), 2.38 (m, 2 H), 2.16 (s,

3

(

3

1

3

1

3

H, H C(8)); C NMR (126 MHz, CDCl ) 166.5 (C(8)),

48.4 (C(2)), 146.2 (C(6)), 133.9 (C(3)), 122.5 (C(7)), 51.6

C(9)), 26.5 (C(4)), 22.4 (C(5)), 12.6 (C(1)); IR (neat) 2992

3

(

B. Preparation of Nitroalkenes and Chiral Vinyl Ethers

. Nitroalkene Preparations. Preparation of Methyl (2Z,6E)-7-

(w), 2953 (w), 2843 (w), 2255 (m), 1721 (s), 1662 (m), 1523 (s),

1438 (m), 1391 (m), 1334 (s), 1284 (m), 1213 (m), 1173 (m),

1

Nitro-2,6-octadienoate ((E,Z)-1). To a 500 mL, two-necked, round-

bottomed flask fitted with a rubber septum and a nitrogen inlet

adaptor was added alcohol (Z)-S3 (3.0 g, 20.8 mmol) along with

1042 (w), 971 (w), 906 (s); TLC R 0.6 (hexanes/EtOAc, 4:1)

[UV, KMnO₄].

f

CH

mmol, 6.0 equiv) was added, and the flask was immersed in anice/NaCl(s)

bath. Once the internal temperature reached ꢀ5 °C, SO pyridine

2 2

Cl and DMSO (100 mL each). Triethylamine (17.4 mL, 125

3

complex was added in a single portion (5.5 g, 31 mmol, 1.5 equiv). After

2

3

h at <0 °C, an additional portion of SO pyridine complex (5.5 g,

3

1 mmol, 1.5 equiv) was added. After another 2 h, the reaction was

quenched by pouring into a 500 mL separatory funnel containing

4

100 mL of satd aq NH Cl. The organic phase was diluted with another

1

00 mL of CH Cl and rinsed with NH

2

4

Cl (3 ꢁ 50 mL) followed by

CuSO (1 ꢁ 50 mL), brine (100 mL), and H O (100 mL). The com-

2. Preparation of Chiral Vinyl Ether 2. Preparation of (1R,2S)-2-

Phenylcyclohexyloxyethene (2). To a 250 mL, one-necked, round-

bottomed flask fitted with a drying tube containing calcium chloride

with a reflux condenser attached open to air and a magnetic stir bar

were added sequentially n-butyl vinyl ether (110 mL, 851 mmol,

4

2

bined organic extracts were dried (MgSO

4

), filtered over a silica gel plug

(

3 ꢁ 1 cm), and concentrated to afford 2.98 g (89%) of crude

intermediate aldehyde as a light yellow oil. The aldehyde was carried on

directly to nitroalkene (E,Z)-1 without further purification as previously

described. The data collected were consistent with those previously

reported. Data for (E,Z)-1: H NMR (500 MHz, CDCl

(

9

1

30 equiv), alcohol (1R,2S)-2-phenylcyclohexanol (5.0 g, 28.4 mmol),

and palladium(II) acetateꢀ1,10-phenanthroline (115 mg, 0.284 mmol,

0.01 equiv). The flask was immersed in an oil bath and heated to

65 °C over 4 h. After being stirred for 3 days, the reaction mixture

was allowed to cool to room temperature and concentrated to 1/2

of the total volume (∼50 mL) by rotary evaporation (15 mmHg,

20ꢀ25 °C). The resulting solution was filtered over a plug of silica

9

0

1

3

) 7.10

t, J = 7.8, 1 H, HC(3)), 6.19 (dd, J = 8.2, 10.9, 1 H, HC(6)), 5.85

d, J = 11.4, 1 H, HC(7)), 3.70 (s, 3 H, H C(9)), 2.84 (dq, J = 1.6,

.6, 2 H, H C(4)), 2.38 (m, 2 H, H C(5)), 2.16 (s, 3 H, H C(1));

C NMR (126 MHz, CDCl ) 166.4 (C(8)), 147.1 (C(2)), 134.6

C(6)), 121.1 (C(3)), 51.2 (C(7)), 27.3 (C(9, 1)), 12.5 (C(4,5));

3

7

2

3

1

3

(

IR (neat) 3054 (w), 2987 (w), 2954 (w), 2254 (m), 1719 (s), 1650

m), 1521 (s), 1440 (m); MS (EI, 70 eV) 197 (5), 182 (12), 168 (25),

gel (40 mm ꢁ 4 cm), rinsed with hexanes/EtOAc/Et N (49:49:2, 3 ꢁ

3

(

30 mL), and concentrated by rotary evaporation (15 mmHg,

4

278

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

2

0ꢀ25 °C). The resulting pale yellow oil was purified by silica gel

column chromatography (4 cm ꢁ10 cm, gradient elution, hexanes/

EtOAc/Et N, 49:0:1, 48:1:1, 47:2:1, 44:5:1, 250 mL each) to afford 4.3 g

74%) of 2 as a colorless oil along with 1.2 g of recovered (1R,2S)-

(500 MHz, CDCl

3

) 4.67 (d (br), J = 7.2, 1 H, HC(2)), 3.91 (ddd, J =

3.8, J = 8.5, J = 12.1, 1 H, HC(8)), 2.94 (ddd, J = 7.9, J = 8.1, J = 11.9, 1

H, HC(8)), 2.71ꢀ2.66 (m, 1 H, HO), 2.64 (dd, J = 7.5, J = 14.9, 1 H,

HC(7a)), 2.28 (ddd, J = 4.9, J = 7.5, J = 12.5, 1 H, HC(5a)), 2.13 (ddd,

J = 3.8, J = 8.0, J = 11.9, 1 H, HC(5)), 1.80 (m, 1 H, HC(5)), 1.72 (m,

3

(

2

-phenylcyclohexanol. The data collected were consistent with those

90

1

previously reported. Data for 2: bp 85 °C (0.1 mmHg); H NMR

2 H, HC(7), HC(5a)), 1.50 (m, 1 H, HC(7)), 1.33 (s, 3 H, H C(8)),

3

13

(500 MHz, CDCl

3

) 7.31ꢀ7.16 (m, 5 H,), 6.05 (dd, J = 14.1, 6.5,

1.27 (ddd, 1 H, J = 5.7, J = 10.6, 10.6, HC(6)); C NMR (126 MHz,

CDCl ) 176.4 (C(2)), 75.5 (C(7b)), 72.8 (C(1)), 51.0 (C(7a)), 49.1

0

1

H, HC(1 )), 4.11 (dd, J = 14.0, 1.1, 1 H, HC(2 )), 3.84 (dt,

3

0

J = 10.3, 4.5, 1 H, HC(1)), 3.77 (dd, J = 6.6, 1.1, 1 H, HC(2 )), 2.65 (dt,

J = 11.8, 3.5, 1 H, HC(2)), 2.25ꢀ2.20 (m, 1 H), 1.94ꢀ1.86 (m, 2 H),

(C(5a)), 42.0 (C(4)), 31.4 (C(6)), 30.9 (C(7)), 24.8 (C(5)), 22.8

(C(8)); IR (NaCl plate) 3283 (s, br), 2961 (s), 2860 (s), 1697 (s),

1677 (s), 1653 (s); MS (ESI, Q-tof) 182 (13), 181 (99), 167 (11), 166

(100), 163 (10), 162 (15), 138 (56), 111 (17), 110 (16), 107 (26), 96

(27), 82 (33), 81 (26), 67 (21), 56 (29), 55 (37), 53 (18); TLC R 0.20

2 4

(hexane/EtOAc, 4:1) [I , KMnO ].

1

3

1

.78ꢀ1.74 (m, 1 H), 1.58ꢀ1.34 (m, 4 H); C NMR (126 MHz,

0

CDCl

3

) 151.1 (C(1 )), 143.7 (C(7)), 128.2 (C(9)), 127.6 (C(8)),

0

1

3

2

1

26.2 (C(10)), 87.4 (C(2 )), 81.8 (C(1)), 50.3 (C(2)), 34.1 (C(6)),

f

2.1 (C(3)), 25.9 (C(4)), 24.8 (C(5)); IR (neat) 3031 (w), 2936 (s),

859 (m), 1632 (s), 1559 (w), 1495 (w), 1449 (m), 1356 (w), 1183 (s),

119 (m), 1076 (s), 818 (s); MS (EI, 70 eV) 202 (100), 160 (3),

59 (28), 158 (11), 91 (100), 81 (17), 67 (8), 55 (6); TLC R 0.90

f

(

hexanes/Et

C. Scaffold Preparation by Tandem Cycloaddition of Nitroalkenes

. Preparation of Scaffold for Library I.

2 2

O, 9:1) [I ].

1

Preparation of (1S,3S,5aS,7aS,7bR)-Octahydro-1-oxo[(phenoxy-

thiocarbonyl)oxy]-7b-methyl-2H-cyclopenta[gh]pyrrolizin-2-one

(

(()-8). To a 50 mL, single-necked, round-bottomed flask fitted with a

nitrogen inlet, a rubber septum, and a magnetic stir bar was added

racemic R-hydroxy lactam 4a (215 mg, 1.2 mmol) followed by

Preparation of rel-(1S,3S,5aS,7aS,7bR)-Octahydro-1-hydroxy-7b-

dimethylaminopyridine (72 mg, 0.59 mmol, 0.5 equiv) and CH Cl₂

2

methyl-2H-cyclopenta[gh]pyrrolizin-2-one ((()-4a). To a 50 mL,

three-necked, round-bottom flask fitted with two rubber septa, a

magnetic stir bar, a nitrogen inlet adaptor, and an internal temperature

probe was added nitroalkene (E,E)-1 (1.0 g, 5 mmol) followed by

(17 mL). Lastly, pyridine (202 μL, 2.4 mmol, 2.0 equiv) was added

followed by phenylchlorothionoformate (330 μL, 2.4 mmol, 2 equiv),

both via syringe. The resulting light-yellow solution was stirred

for 4 h at room temperature, during which time it gradually became

darker. The solution was concentrated by rotary evaporation

(15 mmHg, 20ꢀ25 °C), and the residue was purified by silica gel

column chromatography (2 cm ꢁ10 cm, hexane/EtOAc, 4:1).

Hot filtration followed by recrystallization hexanes/CH Cl (25:1,

CH

by immersion in acetone/CO2(s) bath. Then a Me

6.3 mL of a 2 M solution in toluene, 12.5 mmol, 2.5 equiv) via syringe

2

Cl

2

(20 mL). The internal temperature was brought to ꢀ60 °C

3

Al was added

(

followed by n-butyl vinyl ether (1.25 mL, 12.6 mmol, 2.5 equiv).

The resulting bright yellow solution solution was allowed to stir at

for 6 h, during which time the yellow color gradually faded. The

reaction was quenched by the cautious addition of silica gel (ca. 2 g) via

a long stem funnel until no bubbling was observed. The mixture was

then poured into a funnel containing more silica gel (∼3 g, prewetted

with EtOAc) and rinsed with ethyl acetate (200 mL). The result-

2

∼20 mL) afforded 294 mg (78%) of 8 as colorless rhomboids. Data for

1

8: mp 100ꢀ101 °C; H NMR (500 MHz, CDCl ) 7.42 (m, (second

3

0

order), 2 H, HC(4 )), 7.30 (dt (second order), J = 1.2, 7.2, 1 H,

0

HC(5 )), 7.14 (m, 2 H, HC(3 )), 6.12 (dd, J = 1.0, 7.3, 1 H,

HC(1)), 4.02 (ddd, J = 3.8, 8.6, 12.2, 1 H, HC(4)), 3.01 (m, 2 H,

HC(4), HC(7a)), 2.34 (m, 1 H, HC(5a)), 2.18 (et, J = 3.8, 8.1,

8.1, 11.9, 1 H, HC(7)), 1.86 (ddd, J = 7.6, 7.6, 12.8, 1 H,

HC(7 or 6)), 1.78 (m, 2 H, HC(5), HC(6)), 1.59 (et, J = 4.8,

ing clear solution was dried (MgSO

4

) to remove any adventitious

water, filtered, and concentrated by rotary evaporation (15 mmHg,

2

0ꢀ25 °C) to give a clear viscous oil. The intermediate nitronate was

8.2, 8.2, 13.0, 1 H, HC(5)), 1.39 (s, 3 H, H C(8)), 1.37 (d, J = 5.7,

3

1

3

0

diluted with 50 mL of toluene and transferred to a 100 mL round-

bottomed flask. The flask was fitted with a nitrogen inlet adaptor, and

NaHCO

1 H, HC(6)); C NMR (126 MHz, CDCl ) 194.5 (C(1 )),

3

0

169.8 (C(1)), 153.4 (C(2 )), 129.5 (C(4 )), 126.6 (C(5 )), 121.8

(C(3 )), 82.6 (C(2)), 75.6 (C(9)), 49.5 (C(3)), 48.7 (C(6)), 42.5

0

3

was added (420 mg, 5 mmol, 1.0 equiv) along with a stir

bar. The suspension was allowed to stir at room temperature for

(C(8)), 31.4 (C(7)), 30.9 (C(4)), 25.7 (C(5)), 23.2, (C(10)); IR

(CH Cl , film) 2966 (w), 2252(m), 1706 (s), 1490 (m), 1278 (s),

1208 (s); TLC R 0.78 (hexanes/EtOAc, 1:1) [UV, KMnO ]; HRMS

f 4

7

2

h, filtered, and concentrated by rotary evaporation (15 mmHg,

0ꢀ25 °C). The resulting mixture of nitroso acetals was diluted in

2

EtOAc/MeOH (9/1, 25 mL) and added to a test tube (6 cm ꢁ14 cm)

containing a spatula tip (∼100 mg) of Raney Ni (previously washed

17 19 3

C H O NS (317.10857) calcd 317.10857, found 317.10814.

Anal. Calcd for C H O NS (317.11): C, 64.33; H, 6.03; N, 4.41.

1

7

19

3

with H

2

O, MeOH, and EtOAc, 2 ꢁ 15 mL each) along with a magnetic

Found: C, 64.44; H, 6.00; N, 4.50.

stir bar. The tube was placed in a steel autoclave, which was then pre-

ssurized with H (350 psi) and allowed allowed to stir for 2 days. After

days, the autoclave was carefully vented in a fume hood and the

2

solution was filtered through a plug of Celite (5 cm ꢁ5 cm) with

EtOAc (200 mL). The resulting clear filtrate was concentrated by

rotary evaporation (15 mmHg, 20ꢀ25 °C) and purified by silica gel

column chromatography (2 cm ꢁ6 cm, CH Cl /EtOAc, 19:1, 10:1,

Preparation of (3S,5aS,7aR,7bR)-Octahydro-7b-methyl-2H-cyclo-

penta[gh]pyrrolizin-2-one ((()-9).

2

1

:1, 100 mL each) to afford 778 mg (86%) of the racemic R-hydroxy

lactam 4a as a white powder. The data collected were consistent with

those previously reported. Data for 4a: mp 108ꢀ115 °C; H NMR

Thionocarbonate 8 (140 mg, 0.59 mmol) in benzene (40 mL)

was added to a one-necked, round-bottomed ﬂask via syringe

9

0

1

4

279

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

which was ﬁtted with a reﬂux condenser and a magnetic stir bar. Atop

the reﬂux condenser was ﬁtted a nitrogen inlet adaptor and a rubber

septum. The round-bottomed ﬂask was immersed in a preheated (100 °C)

oil bath and allowed to come to reﬂux (∼20 min). During this time,

tributyltin hydride (1.08 mL, 0.77 mmol, 1.3 equiv) and AIBN (206 μL,

2. Preparation of the Scaffold for Library II.

0.12 mmol, 0.2 equiv) were added to a separate 25 mL, one-necked, conical

ﬂask along with 15 mL of benzene and a magnetic stir bar. Once the round-

bottomed ﬂask reached reﬂux, the tributyltin hydride AIBN solution was

transferred dropwise via cannula over 3.5 h, maintaining a consistent gentle

reﬂux. The resulting solution was allowed to reﬂux for an additional 2 h and

then cooled to room temperature. Potassium ﬂuoride (3.0 g, 51.6 mmol)

was added to the resulting light-yellow solution, and the resulting mixture

was allowed to stir for an additional 3 h. The suspension was ﬁltered (cotton

plug), and the ﬁltrate was concentrated by rotary evaporation (15 mmHg,

Preparation of (1S,3S,5aS,7aS,7bR)-Octahydro-1-hydroxy-7b-methyl-

H-cyclopenta[gh]pyrrolizin-2-one (4a). To a 250 mL, three-necked,

round-bottom flask fitted with two rubber septa, a magnetic stir bar, a

nitrogen inlet adaptor, and an internal temperature probe were added

nitroalkene (E,E)-1 (3.68 g, 18.45 mmol) and chiral vinyl ether 2 (5.6 g, 28

mmol, 1.5 equiv) via syringe. The flask was then evacuated using high

2

0ꢀ25 °C). The residue was puriﬁed by silica gel column chromatography

3 cm ꢁ10 cm, hexane/EtOAc, 1:0, 9:1, 4:2, 7:3, 3:2 1:1, 50 mL each) with

a plug of KF to aﬀord 82.1 mg (83%) of lactam 9 as a colorless oil. Data for

vacuum (∼0.1 mmHg) for 30 min and then was backfilled with N

2

and

(

charged with CH Cl (110 mL). The internal temperature was set to

2

ꢀ

85 °C (hexanes/N

2

bath). This yellow solution was sirred for 15 min, and

1

9

3

2

: bp 80 °C (0.1 mmHg); H NMR (500 MHz, CDCl

3

) 3.88 (ddd, J =

93

then freshly prepared TiCl

2

2 2 2

(O-i-Pr) solution (1.2 M in CH Cl , 3.0

.6, 8.6, 12.0, 1 H, HC(4)), 2.94 (dd, J = 8.0, 16.6, 2 H, HC(1), HC(4)),

.28 (dd, J = 7.8, 13.9, 1 H, HC(5a)), 2.15 (m, 3 H, HC(7a), HC(1),

equiv) was added dropwise via syringe at a rate that the internal temperature

did not rise above ꢀ70 °C (ca. 15 min). After addition of the Lewis acid, the

cooling bath was replaced with an acetone/CO2(s) bath, and the resulting

bright yellow solution was stirred for another 5 h while an internal

temperature eꢀ75 °C was maintained. Throughout the reaction, the

yellow color gradually faded and a white precipitate formed. After 5 h, the

reaction was quenched with triethylamine (6.1 equiv, 1.0 M in MeOH) via

syringe while an internal temperature of less than ꢀ40 °C was maintained.

The cooling bath was then removed, and the reaction mixture was allowed

to warm to 0 °C (ca. 15 min). The resulting white suspension was then

diluted with ethyl acetate (100 mL) and poured into a 1 L separatory funnel

HC(7 or 6)), 1.94 (ddd, J = 7.2, 7.2, 20.6, 1 H, HC(5)), 1.76 (ddd, J =

.8, 13.5, 13.6, 1 H, HC(7)), 1.62 (m, 2 H, HC(5), HC(7 or 6)), 1.42

6

13

(m, 1 H, HC(7 or 6)), 1.30 (s, 3 H, H

3

C(8)); C NMR (126 MHz,

CDCl ) 176.6 (C(2)), 79.1 (C(7b)), 48.5 (C(7a)), 44.1 (C(5a)),

3

42.2 (C(4)), 40.3 (C(1)), 33.3 (C(5)), 32.2 (C(6)), 31.3 (C(7)), 23.0

(

1

C(8)); IR (neat) 3019 (s), 2957 (m), 2865 (w), 2400 (m), 1676 (m),

þ

521 (m), 1400 (w), 1215 (s); MS (ESI, Q-tof) 167 (M þ 1, 12), 166

þ

(

M , 100); mol formula C H NO (165.23); HRMS C H NO

165.1232) calcd 166.1232, found 166.1231; TLC R

].

10

15

10 15

f

0.17 (hexanes/

EtOAc, 1:1) [KMnO

4

containing satd aq NH Cl (250 mL) and tert-butyl methyl ether (100 mL).

4

The layers were separated, and the organic phase was washed with satd aq

NH

50 mL). The aqueous layers were combined and back-extracted with tert-

butylmethyl ether (3 ꢁ 100 mL). The combined organic layers were dried

over NaHCO /MgSO (1/1), filtered (cotton plug), and concentrated by

4

Cl (2 ꢁ 50 mL), brine (1 ꢁ 100 mL), and satd aq NaHCO

3

(1 ꢁ

1

3

4

rotary evaporation (15 mmHg, 20ꢀ25 °C). The resulting residue was

filtered through a pad of silica gel (3 ꢁ 3 cm), eluting with ethyl acetate

(100 mL) to remove any remaining amine impurities. The resulting clear

Preparation of (5aS,7aR)-Octahydro-7b-methylcyclopenta[gh]-

pyrrolizine Borane (10). To a 100 mL, one-necked, round-bottomed

flask fitted with a rubber septum and a nitrogen inlet adaptor were

solution was concentrated (15 mmHg, 20ꢀ25 °C) to a pale-yellow residual

oil in a 1 L, one-necked, round-bottom flask. tert-Butylmethyl ether

3

(200 mL) was added followed by NaHCO

a magnetic stir bar. The flask was fitted with an a nitrogen inlet adaptor,

evacuated, backfilled with N , and allowed to stir at room temperature for

(1.5 g, 1.0 equiv) and a large

3

added lactam 15 (1.09 g, 6.6 mmol) and THF (60 mL). The flask was

cooled to 0 °C in an ice bath, and BH

3

THF complex (3.0 equiv,

.0 M solution, 7.0 mL) was added dropwise over 10 min (bubbling

2

3

1

12 h. The suspension was stirred at room temperature for 12 h and then

filtered through Celite (3 ꢁ 3 cm) and concentrated by rotary evaporation

(15mmHg, 20ꢀ25 °C). The resulting mixture of nitroso acetals was diluted

in EtOAc/MeOH (9/1, 25 mL) and added to a test tube (6 cm ꢁ14 cm)

containing a spatula tip (∼200 mg) of Raney Ni (previously washed with

observed). The cooling bath was removed, and the resulting clear

solution was stirred for 8 h. The reaction was quenched by the

addition of 30 mL of MeOH, and the reaction was concentrated

by rotary evaporation (15 mmHg, 20ꢀ25 °C). The resulting thick,

glassy oil was purified by silica gel column chromatography (2 cm

H

O, MeOH, and EtOAc, 2 ꢁ 15 mL each) along with a magnetic stir bar.

2

ꢁ

12 cm, hexanes/EtOAc, 10:1, 5:1, 3:1, 200 mL each) to afford 794

The tube was placed in a steel autoclave, which was then pressurized with H

2

1

mg (80%) of borane complex 10 as a white wax. Data for 10: H NMR

500 MHz, CDCl ) 3.27 (ddd, J = 7.0, 7.0, 11.9, 2 H, HC(2 and 4)),

.14 (m, 2 H, HC(2 and 4)), 2.30 (m, 2 H, HC(5a), HC(7a)), 2.14

ddd, J = 6.9, 6.9, 15.2, 2 H, HC(1), HC(5)), 1.86 (m, 2 H, CH ), 1.56

), 1.47 (s, 3 H, HC(8)), 1.26 (m,

(350 psi) and set on a stir-plate. After being stirred for 2 days at room

temperature, the autoclave was carefully vented in a fume hood and the

solution was through a plug of Celite (5 cm ꢁ5 cm) with EtOAc (200 mL).

The resulting clear solution was concentrated by rotary evaporation (15

mmHg, 20ꢀ25 °C) and purified by silica gel column chromatography (30

(

3

(

2

ddd, J = 9.8, 9.8, 10.3, 3 H, CH

2

9

2

13

J = 6.6, 13.7, 2 H, CH

2

), (0.8ꢀ2.5, br, 3 H, (H

3

B) ); C NMR

mm ꢁ 8 mm) (CH

2 2

Cl /EtOAc, 19:1, 10:1, 1:1, 200 mL each) to afford

(

126 MHz, CDCl ) 87.3 (C(7b)), 62.5 (C(2)), 52.2 (C(5a)), 31.9

C(1)), 28.5 (C(8)), 24.9 (C(6)); IR (NaCl plates, thin film) (m),

2.76 g (76%, three steps) of R-hydroxy lactam 4a as a white powder. The

62c,91

3

spectroscopic data are in accord with those previously reported.

Data

3

)

1

3

020 (s)2971 (w), 292 (w), 2400 (m), 2361 (m), 2326 (m), 1517 (m),

for 4a: mp 110ꢀ114 °C (hexanes/EtOAc); H NMR (500 MHz, CDCl

1

475 (w), 1425 (w), 1361 (w), 1215 (s), 929 (m), 756 (s), 669 (s);

4.69 (d, J = 7.2, 1 H, HC(1)), 3.93 (ddd, J = 12.0, 8.5, 3.6, 1 H, HC(4)),

2.99ꢀ2.93 (m, 1 H, HC(4)), 2.70ꢀ2.64 (m, J = 7.4, 2 H, HOC(1),

HC(7)), 2.33ꢀ2.27 (m, 1 H, HC(5a)), 2.18ꢀ2.12 (m, 1 H, HC(5)),

þ

MS (ESI, Q-tof) 165 (M , 10), 164 (83), 163 (19), 153 (13), 152 (100);

mol formula C10 20BN (165.08); HRMS C10 19BN (164.1611)

calcd 164.1611, found 164.1611; TLC R 0.54 (CH Cl /hexanes, 3:1)

I₂, CAM].

H

1.85ꢀ1.70 (m, 3 H, HC(6), H C(7)), 1.56ꢀ1.42 (m, 1 H, HC(5)), 1.35

13

f

2

[

3

(s, 3 H, H C(10)), 1.33ꢀ1.27 (m, 1 H, HC(6)); C NMR (126 MHz,

4

280

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

CDCl

3

) 176.6 (C(1)), 75.7 (C(9)), 72.9 (C(2)), 51.2 (C(3)), 49.3

C(6)), 42.2 (C(8)), 31.6 (C(7)), 31.1 (C(5)), 24.9 (C(4)), 23.0

C(10)); IR (CDCl , film) 3390 (s), 2961 (s), 2860 (m), 1705 (s), 1325

s); MS (ESI, Qtof) 182 (100), 102 (14); [R]

<ꢀ70 °C was maintained (ca. 15 min). After addition of the Lewis acid,

(

the cooling bath was replaced with an acetone/CO

bath, and the

(s)

2

3

resulting bright yellow solution was stirred for another 5 h while an

internal temperature eꢀ72 °C was maintained. During the course of

the reaction, the yellow color gradually faded and a white precipitate

formed. After 5 h, the reaction was quenched with triethylamine

24

D

ꢀ35.2 (c = 1.07, CH

2 2

Cl )

90

lit. (99:1) =ꢀ36.9 (c=1.00, CH Cl ); TLC R 0.20 (hexane/EtOAc, 4:1)

2

f

[

I , KMnO ].

2 4

(6.1 equiv, 1 M in MeOH) via syringe while an internal temperature

<

ꢀ40 °C. The cooling bath was then removed and the reaction mixture

was allowed to warm to 0 °C (ca. 15 min). The resulting white

suspension was then diluted with ethyl acetate (50 mL) and poured in

a 500 mL separatory funnel containing a biphasic mixture of satd aq

NH

4

Cl and MTBE (100 mL each). The biphasic solution was separated

and the organic extract was extracted with satd aq NH Cl solution (2 ꢁ

4

50 mL), H

2

O (2 ꢁ 50 mL) and brine (2 ꢁ 50 mL). The combined

Preparation of (1S,3R,5aS,7aS,7bR)-Octahydro-1-hydroxy-7b-methyl-

H-cyclopenta[gh]pyrrolizin Borane (11a). To a 50 mL, two-necked,

aqueous layers were back extracted with MTBE (3 ꢁ 75 mL). The com-

2

3

bined organic layers were dried over NaHCO /MgSO (1/1), filtered

3

4

round-bottomed flask fitted with a rubber septum, a nitrogen inlet

adaptor, a magnetic stir bar, and an internal temperature probe (inserted

through a rubber septum) were added lactam 4a (181 mg, 1.0 mmol)

and THF (1 mL). The flask was cooled to 0 °C in an ice bath, and

(cotton plug), and concentrated by rotary evaporation (15 mmHg,

2

0ꢀ25 °C). The resulting residue was filtered through a pad of silica gel

(3 ꢁ 3 cm), eluting with ethyl acetate (100 mL) to remove any remain-

ing amine impurities. The resulting clear solution was concentrated by

rotary evaporation (15 mmHg, 20ꢀ25 °C) to a pale-yellow residual oil

in a 1-L one-necked, round-bottomed flask. Hexanes (200 mL) was

3

BH THF complex (5.0 equiv, 1.0 M solution, 5 mL) was added drop-

3

wise over 10 min (bubbling observed). The cooling bath was removed,

and the resulting clear solution was stirred for 2 h. The reaction was

quenched by the addition of 20 mL of MeOH and the mixture was con-

centrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). This process

was repeated three more times to afford the crude product as a white

solid. Purification by silica gel column chromatography ((2 cm ꢁ7 cm),

hexanes/EtOAc, 95:5, 85:15, 3:1, 150 mL each) afforded ∼185 mg of

added followed by NaHCO

magnetic stir bar. The flask was fitted with a nitrogen inlet adaptor

evacuated and backfilled with N and immersed in a preheated 40 °C oil

3

(4.22 g, 50.0 mmol, 5.0 equiv) and a large a

2

bath. The suspension was stirred at 40 °C (bath temperature) for 30 h

and then filtered through Celite (3 ꢁ 3 cm) and concentrated by rotary

evaporation (15 mmHg, 20ꢀ25 °C). The resulting mixture of nitroso

acetals was diluted in EtOAc/MeOH (9:1, 25 mL) and added to a test

tube (6 cm ꢁ14 cm) containing a spatula tip (∼200 mg) of Raney Ni

1

1a. Recrystallization from hexanes/MTBE (5:1) afforded 175 mg

(

96%) of the borane complex 11a as white needles. Data for 11a: mp

1

64ꢀ165 °C (MTBE/hexanes); H NMR (500 MHz, CDCl

3

) 4.75 (dd,

(

previously washed with H O, MeOH, and EtOAc, 2 ꢁ 15 mL each)

2

J = 8.1, 15.3, 1 H, HC(1)), 3.44 (dd, J = 6.5, 10.9, 1 H, HC(2)), 3.28

along with a magnetic stir bar. The tube was placed in a steel autoclave,

which was then pressurized with H (350 psi). After 2 days, the autoclave

(

3

ddd, J = 2.5, 6.7, 12.1, 1 H, HC(4)), 3.16 (dt, J = 6.6, 11.7, 1 H, HC(4)),

2

.03 (m, 1 H, HC(2)), 2.39 (m, 2 H HC(5a), HC(7a)), 2.07ꢀ2.00 (m, 1

was carefully vented in a fume hood, and the solution was filtered through

a plug of Celite (5 cm ꢁ5 cm) with EtOAc (200 mL). The resulting clear

filtrate was concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C)

H, HC(7)), 1.99ꢀ1.90 (m, 1 H, HC(6)), 1.89ꢀ1.73 (m, 2 H, HC(6),

HC(5)), 1.74ꢀ1.58 (M, 2 H, HC(7), HC(5)), 2.10ꢀ1.25 (s, broad, 3 H,

13

H B), 1.48 (s, 3 H, H C(8)); C NMR (126 MHz, CDCl ) 87.8

3

and purified by silica gel column chromatography (2 cm ꢁ8 cm, CH

2 2

Cl /

(

C(7b)), 69.4 (C(1)), 66.0 (C(2)), 63.1 (C(4)), 55.1 (C(7a)), 53.1

C(5a)), 32.1 (C(5)), 28.0 (C(7)), 26.4 (C(6)), 25.2 (C(8)); IR (NaCl

plate, film) 3018 (m), 2969 (m), 2872 (w), 2367 (m), 2363 (m), 2333

EtOAc, 19:1, 10:1, 1:1, 100 mLeach). Recrystallization from hot hexanes/

EtOAc (10:1, ∼50 mL) afforded 1.15 g (63%, 3 steps) of analytically pure

R-hydroxy lactam 4b as colorless needles. Data for 4b: mp 112ꢀ117 °C;

(

m), 2276 (w), 1215 (s); MS (ESI, Q-tof) 180 (42), 167 (21), 152

1

24

D

H NMR (500 MHz, CDCl ) 4.68 (dd, J = 1.2, 7.3, 1 H, HC(1)), 3.90

3

100), 96 (59); [R]

]. Anal. Calcd for C10

1.13; N, 7.73. Found: C, 66.05; H, 10.95; N, 7.52.

. Preparation of Scaffold for Library III.

ꢀ1.74 (c = 1.0, CH

2 2 f

Cl ); TLC R 0.33 (hexanes/

20BNO (181.08): C, 66.33; H,

(ddd, J = 3.7, 8.5, 12.0, 1 H, HC(4)), 3.05 (br, 1 H, HOC(1)), 2.93 (ddd,

EtOAc, 3:1) [I

1

3

2

H

J = 1.2, 8.1, 12.0, 1 H, HC(4)), 2.63 (q, J = 7.3, 1 H, HC(7a)), 2.25 (m, 1

H, HC(5a)), 2.12 (m, 1 H, HC(5)), 1.79 (m, 1 H, HC(6)), 1.72 (m, 2 H,

H C(7)), 1.49 (m, 1 H, HC(5)), 1.32 (s, 3 H, H C(8)), 1.27 (m, 1 H,

2

3

1

3

HC(6)); C NMR (126 MHz, CDCl ) 176.2 (C(1)), 75.8 (C(9)), 73.0

3

(

C(2)), 51.2 (C(3)), 49.3 (C(6)), 42.2 (C(8)), 31.5 (C(4)), 31.1

C(7)), 24.8 (C(8)), 23.0 (C(10)); IR (KBr plate) 3302 (s, br), 2952

s), 2864(s), 1700(s), 1653(s);MS(ESI, Q-tof) 181 (100), 166(91), 38

53), 107 (26), 82 (28), 55 (35); TLC R

f

0.33 (EtOAc/CH

Cl ). Anal. Calcd for C10

181.23): C, 66.27; H, 8.34; N, 7.65. Found: C, 65.97; H, 8.32; N, 7.65.

2 2 2

Cl , 1:3) [I ,

2

4

KMnO

(

4

]; [R]

D

þ50.6 (c = 1.03, CH

2

H15NO

2

Preparation of (1R,3S,5aS,7aS,7bR)-Octahydro-1-hydroxy-7b-methyl-

2H-cyclopenta[gh]pyrrolizin-2-one (4b). To a 250 mL, three-necked,

round-bottomed flask fitted with two rubber septa,

a mag-

netic stir bar, a nitrogen inlet adaptor, and an internal temperature

probe were added nitroalkene (E,Z)-1 (2.0 g, 10.0 mmol) and chiral

vinyl ether 2 (3.1 g, 15.1 mmol, 1.5 equiv) via syringe. The resulting

yellow oil was then evacuated using high vacuum (∼0.1 mmHg) for 30

min. The flask was backfilled with N₂and charged with CH Cl

2

(

60 mL). The solution was cooled to ꢀ85 °C by immersion in a

hexanes/N bath. This cooled yellow solution was sirred for 15 min, and

then freshly prepared TiCl (O-i-Pr) solution (1.2 M in CH Cl , 3.0

equiv) was added dropwise via syringe while an internal temperature

Preparation of (1R,3R,5aS,7aS,7bR)-Octahydro-1-[N-(3,5-Dinitro-

phenyl)carbamoxy]-7b-methyl-2H-cyclopenta[gh]pyrrolizin-2-one (S4).

A 25 mL round-bottomed flask (A) was fitted with a reflux condenser,

2

4

281

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

a nitrogen inlet adaptor, a rubber rubber septum, and a magnetic stir bar.

The apparatus was opened, and 3,5-dinitrobenzoylazide (25 mg, 0.105

mmol, 1.2 equiv) was added as a solid. The apparatus was evacuated and

4. Preparation of Scaffold for Library IV.

2

backfilled with N three times and toluene (5.0 mL) was added. The

clear solution was immersed in a preheated (115 °C) oil bath and stirred

at reflux for 30 min. In a separate, two-neck, 5 mL conical flask (B) fitted

with a triangular a magnetic stir bar, rubber septum, and a nitrogen

inlet adaptor the starting alcohol was added (17.2 mg, 0.087 mmol)

followed by toluene (1.0 mL). The resulting solution (B) was

transferred to flask (A) via cannula. The resulting light yellow

solution was heated to reflux for 2 h and then allowed to cool to

room temperature. The reaction mixture was concentrated in vacuo,

and the crude product was purified by column chromatography

Preparation of (3R,5aS,7aS,7bR)-Octahydro-7b-methyl-2H-cyclo-

penta[gh]pyrrolizin-1-one Borane (13). To a 100 mL, three-necked,

3

round-bottomed flask fitted with a nitrogen inlet adapter, a magnetic stir

bar, and an internal temperature probe were added 11a (500 mg, 2.76

mmol), CH Cl (11.5 mL), and DMSO (2.3 mL). The resulting solution

2

was cooled to ꢀ12 °C in an ice/salt bath. To this solution was added diiso-

(

hexane/EtOAc (3/1, 1/1)) to afford 36 mg (75%) of enantioen-

propylethylamine (2.41 mL, 13.8 mmol, 5.0 equiv) followed by a solution

riched (S4) as a white solid. The sample was compared to racemate

of SO pyridine complex (1.32 g, 8.28 mmol, 3.0 equiv) in DMSO

3

9

0,93

1

as previously described.

Data for S4: H NMR (500 MHz,

(4.6 mL) at rate such that the internal temperature remained <ꢀ5 °C

(∼5 min). After the solution was stirred for an additional 20 min at

ꢀ10 °C, the cooling bath was removed, and the solution was allowed to

warm to room temperature (∼10 min) and then cooled to 0 °C by

CDCl ) 9.65 (s, 1 H, HN) 8.71 (d, J = 2.0, 2 H), 8.67 (dd, J = 2.0,

3

2

.0, 1 H), 5.03 (s, 1 H), 3.93 (ddd, 1 H, J = 3.9, 8.7, 12.6), 3.09 (m, 1

H), 2.40 (dd, J = 4.8, 9.0, 1 H,), 2.28 (m, 2 H), 2.13 (m, 1 H), 1.87 (m,

2

1

H), 1.71 (m, 2 H), 1.43 (s, 3 H), 1.28 (m, 5 H), 0.85 (dd, J = 11.2,

3.1, 1 H); C NMR (126 MHz, d7-DMF) 170.7 (C(2)), 152.9

reimmersion in the ice/salt bath. The solution was diluted with Et O

2

1

3

(10mL) andquenchedwithH O (10 mL). This mixture waspouredinto a

2

(

4

(

1

3

C(10)), 148.8 (C(11)), 141.8 (C(13)), 117.6 (C(14)), 111.8

C(12)), 82.4 (C(7b)), 48.4 (C(1)), 50.1 (C(7a)), 48.1 (C(5a)),

250 mL separatory funnel, and the layers were separated. The aqueous

extract was washed with Et O (2 ꢁ 25 mL). The organic extracts were

2

1.9 (C(4)), 31.3 (C(6)), 30.9 (C(7)), 23.0 (C(9)); MS (EI, 70 eV)

05(82), 390(100), 374(12), 187(24), 91(25), 75(40); TLC R 0.32

]; CSF-SFC: t = 11.27 min (96%) and

washed with satd aq CuSO solution (2 ꢁ 25 mL), satd aq NaHCO

4

3

f

solution (1 ꢁ 25 mL), and brine (25 mL) and the combined organic

hexanes/EtOAc, 3:1) [I

2

R

extracts dried over MgSO , filtered through Celite (20 mm ꢁ 5 cm), and

4

5.73 (4%) (Chiralpak AD, 125 bar, 40 °C, 15% MeOH in CO ,

rinsed with TBME (3 ꢁ 10 mL). The filtrate was concentrated by rotary

2

.0 mL/min, 220 nm).

evaporation (15 mmHg, 20ꢀ25 °C) to afford 498 mg (99%) of 13 as a

1

white solid without further purification. Data for 13: H NMR (500 MHz,

CDCl

HC(4)), 3.48 (d, J = 17.6, 1 H, HC(2)), 3.16 (ddd, J = 6.5, 1 H, HC(4)),

.63ꢀ2.57 (m, 2 H, HC(7a), HC(5a)), 2.40 (J = 6.0, 7.4, 9.4, 13.6, 1 H,

HC(5)), 2.16ꢀ2.03 (m, 3 H, HC(7), HC(7)), 1.65 (s, 3 H, H C(8)),

.65ꢀ1.58 (m, 2 H, HC(5)), 1.48ꢀ1.41 (m, 1 H, HC(6)), 0.8ꢀ2.5 (br, 3

3

) 3.89 (d, J = 17.6, 1 H, HC(2)), 3.55 (ddd, J = 7.6, 7.6, 12.2, 1 H,

2

3

1

3

13

H, (H B) ); C NMR (126 MHz, CDCl ) 212.1 (C(1)), 86.5 (C(7b)),

3

7

0.0 (C(2)), 64.9 (C(4)), 58.4 (C(7a)), 51.8 (C(5a)), 35.1 (C(6)), 32.3

Preparation of (1R,3R,5aS,7aS,7bR)-Octahydro-1-hydroxy-7b-methyl-

H-cyclopenta[gh]pyrrolizine Borane (11b). To a 100 mL, single-

necked, round-bottomed flask fitted with a rubber septum, a magnetic

(C(7)), 29.9 (C(5)), 24.4 (C(8)); IR (NaCl plate, thin film) 2964 (s),

2

3

2

1

865 (w), 2381 (s), 2333 (s), 2273 (s), 1755 (s), 1468 (m), 1419 (w),

383 (m), 1195 (s), 1164 (s), 1071 (m), 913 (s), 732 (s); MS (EI, 70 eV)

79 (12), 178 (100), 165 (28), 150 (24), 137 (48), 121 (15), 108 (16), 95

stir bar, and a nitrogen inlet adaptor were added lactam 4b (170 mg,

0

.866 mmol) and THF (60 mL). The flask was cooled in an ice bath,

þ

(27); mol formula C10H18BNO (179.07); HRMS C10H17ONB

and BH THF complex (30 equiv, 1.0 M solution, 9.0 mL) was added

3

(178.1403) calcd 178.1403, found 178.1406; TLC R 0.35 (hexanes/

dropwise over 10 min (bubbling observed). The cooling bath was

removed, and the clear solution was stirred for 8 h. The reaction was

quenched by the addition of 30 mL of MeOH, and the mixture was

concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). The

resulting white solid was redissolved in 15 mL of MeOH (15 mL) and

again concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C).

Purification by silica gel column chromatography ((2 cm ꢁ7 cm),

hexanes/EtOAc, 1:0, 10:1, 8:1, 5:1, 3:1, 100 mL each) afforded 130.2

f

EtOAc, 3:2) [I , CAM].

2

5

. Preparation of Scaffold for Library V (via 20 and 21).

mg (77%) of borane complex 11b as a white solid. Data for 11b: mp

1

8

3

8ꢀ91 °C; H NMR (500 MHz, CDCl

3

) 3.89, (br, 1 H, HC(1)),

.54, (d, J = 8.7, 1 H, HO(C(1)), 3.33, (ddd, J = 12.7, 2.9, 1.3, 1 H,

Preparation of (1S,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-5-

methyl-7b-methyl-2H-cyclopenta[gh]pyrrolizin-2-one ((þ)-5). To a

100 mL, three-necked, round-bottomed flask fitted with two rubber

septa, a magnetic stir bar, a nitrogen inlet adaptor, and an internal

temperature probe were added nitroalkene (E,E)-1 (1.0 g, 5.02 mmol)

and chiral propenyl ether 3 (1.63 g, 7.53 mmol, 1.5 equiv) via syringe.

The resulting yellow oil was then evacuated under high vacuum (∼0.1

HC(2), 3.30, (dd, J = 12.6, 5.22, 1 H, HC(4)), 3.19ꢀ3.10, (m, 2 H

HC(7a), HC(5a)), 2.32ꢀ2.25, (m, 2 H, HC(2), HC(4)), 2.02ꢀ1.90,

(

7

3

m, 2 H, HC(6), HC(7)), 1.83ꢀ1.75, (m, 1 H, HC(6)), 1.55ꢀ1.45,

m, 3 H, HC(5), HC(6), HC(7)), 1.51, (s, 3 H H

3

C(8)), 0.8ꢀ2.5,

1

3

br, 3 H, (H B)); C NMR (126 MHz, CDCl ) 87.5, (C(1))

3

5.2, (C(7b)) 68.4, (C(2)) 64.1, (C(4)) 60.9, (C(7a)) 53.0, (C(5a))

1.5, (C(5)) 30.7, (C(6)) 28.0, (C(7)) 25.2, (C(8)); IR (NaCl

mmHg) for 30 min. The flask was backfilled with N

CH Cl

temperature) using hexanes/N bath. This yellow solution was sirred

2

and charged with

plate, film) 3053 (m), 2986 (m), 2871 (w), 2386 (m), 2306 (m),

253 (m), 1421 (w), 1382 (w), 1265 (s); MS (ESI, Q-tof) 180 (100);

TLC R 0.33 (hexanes/EtOAc, 3:1) [I ]. Anal. Calcd for C10 20BNO

181.08): C, 66.33; H, 11.13; N, 7.73. Found: C, 66.32; H, 10.90;

N, 7.66.

2

(22 mL). The solution was cooled to ꢀ85 °C (internal

2

f

2

H

for 15 min, and then freshly prepared TiCl

CH Cl , 12.6 mL, 15.1 mmol, 3.0 equiv) was added dropwise via syringe

while an internal temperature eꢀ70 °C was maintained (ca. 15 min).

2 2

(O-i-Pr) solution (1.2 M in

(

2

4

282

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

After addition of the Lewis acid, the cooling bath was replaced with an

acetone/CO₂₍_s₎bath, and the resulting bright yellow solution was stirred

for another 5 h while an internal temperature eꢀ75 °C was maintained.

During the course of the reaction, the yellow color gradually faded and a

white precipitate formed. After 5 h, the reaction was quenched with

triethylamine (30.6 mL, 6.1 equiv, 1 M in MeOH) via syringe while an

internal temperature of <ꢀ40 °C was maintained. The cooling bath was

then removed, and the reaction mixture was allowed to warm to 0 °C (ca.

condenser, a nitrogen inlet adaptor, a rubber rubber septum, and a

magnetic stir bar. The apparatus was opened, and 3,5-dinitrobenzoyl

azide (25 mg, 0.11 mmol, 1.1 equiv) was added as a solid. The apparatus

2

was evacuated and backfilled with N three times, and toluene (4.8 mL)

was added. The clear solution was immersed in a preheated (115 °C) oil

bath and stirred at reflux for 30 min. In a separate, two-neck, 5 mL conical

flask (B) fitted with a triangular a magnetic stir bar, rubber septum, and a

nitrogen inlet adaptor was added the starting alcohol (18.9 mg, 0.096

mmol) followed by toluene (1.0 mL). The resulting solution (B) was

transferred to flask A via cannula. The resulting light yellow solution

was heated to reflux for 2 h and then allowed to cool to room

temperature. The reaction mixture was concentrated in vacuo, and

the crude product was purified by column chromatography (hexane/

EtOAc (3/1, 1/1)) to afford 38 mg (98%) of enantioenriched S6 as a

white solid. The sample was compared to racemate as previously

15 min). The resulting white suspension was then diluted with ethyl

acetate (50 mL) and poured onto a biphasic mixture of satd aq NH Cl

4

and ethyl acetate (75 mL) in a 500 mL separatory funnel. The aqueous

extract was washed with ethyl acetate (3 ꢁ 75 mL). The organic extracts

were washed with satd aq NH Cl solution (2 ꢁ 50 mL), H O (2 ꢁ

4

2

5

0 mL), and brine (2 ꢁ 50 mL). The combined organic extracts were

3 4

dried over NaHCO /MgSO (1/1), filtered (cotton plug), and con-

9

0

1

centrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). The resulting

residue was filtered through a pad of silica gel (3 ꢁ 3 cm), eluting with

ethyl acetate (100 mL) to remove any remaining amine impurities. The

resulting clear solution was concentrated by rotary evaporation (15

mmHg, 20ꢀ25 °C) to a pale-yellow residual oil in a 1 L, one-necked,

round-bottomed flask. TBME (100 mL) was added followed by

described. Data for S6: H NMR (500 MHz, CDCl ) 9.90 (s, 1 H,

3

NH), 8.61 (s, 1 H, HC(14)), 8.56 (s, 2 H, HC(12)), 5.89 (d, J = 6.7,

1 H, HC(1)), 4.12 (dd, J = 7.4, 11.9, 1 H, HC(4)), 2.71 (dd, J = 9.5,

14.1, 1 H), 2.69 (dd, J = 10.7, 23.5, 1 H), 1.96 (dd, J = 7.0, 7.0, 1 H),

1.93ꢀ1.81 (m, 2 H), 1.63ꢀ1.50 (m, 3 H), 1.46 (s, 3 H, H C(9)), 1.14

3

1

3

3 3

(d, J = 6.7, 3 H, H C(15)); C NMR (126 MHz, CDCl ) 220.6, 171.5,

NaHCO

3

(5 g, 50.0 mmol, 10.0 equiv). The flask was equipped with a

152.6, 148.8, 141.2, 118.3, 118.2, 112.6, 75.9, 75.0, 58.3, 50.9, 42.7,

31.3, 26.3, 24.3, 17.7; TLC R 0.34 (hexanes/EtOAc, 3:1) [I ]; CSF-

HPLC t = 7.98 min (1%) and 9.88 (99%) (Chiralpak AD, 125 bar,

nitrogen inlet adaptor and a magnetic stir bar and then evacuated and

f

2

backfilled with N (3ꢁ). The suspension was stirred at room tempera-

2

R

ture for 12 h, filtered through Celite (3 ꢁ 3 cm, cotton plug), and

concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). The result-

ing mixture of nitroso acetals was diluted in EtOAc/MeOH (9:1, 25 mL)

and added to a test tube (6 cm ꢁ14 cm) containing a spatula tip (∼200

2

40 °C, 15% MeOH in CO , 3.0 mL/min, 220 nm).

mg) of Raney Ni (previously washed with H

2

O, MeOH, and EtOAc, 2 ꢁ

15 mL each) along with a magnetic stir bar. The tube was placed in a steel

autoclave, which was then pressurized with H (350 psi). After 2 days,

2

the autoclave was carefully vented in a fume hood and the solution was

filtered through a plug of Celite (5 cm ꢁ5 cm, cotton plug) with EtOAc

Preparation of (3R,5aS,7aS,7bR)-Octahydro-5-methyl-7b-methyl-

H-cyclopenta[gh]pyrrolizin-1-one (14). To a 50 mL, three-necked,

round-bottomed flask fitted with a nitrogen inlet adapter, a magnetic stir

bar, and an internal temperature probe were added 12 (505 mg, 2.59

(100 mL). The resulting clear solution was concentrated by rotary

2

evaporation (15 mmHg, 20ꢀ25 °C) and purified by silica gel column

chromatography (2 cm ꢁ 8 cm, hexanes/EtOAc, 9:1, 4:1, 7:3, 2:3, 1:4,

1

00 mL each) affording 870 mg (89%) of a mixture of epimeric R-

2 2

mmol), CH Cl (10.8 mL), and DMSO (2.2 mL). The resulting

1

hydroxy lactams 5 and S5 as a white solid (11:1 by H NMR analysis).

Data for 5: H NMR (500 MHz, CDCl

solution was cooled to ꢀ12 °C in an ice/NaCl bath. To this solution

1

3

) 4.63 (dd, J = 6.8, 1.5, 1 H,

was added diisopropylethylamine (2.25 mL, 12.9 mmol, 5.0 equiv)

HC(4)), 4.05 (dd, J = 11.8, 7.4, 1 H, HC(4)), 2.73 (br, s, 1 H, OH),

3

followed by the dropwise addition of a solution of SO pyridine

3

2

.59ꢀ2.46 (m, 2 H), 1.86ꢀ1.73 (m, 3 H), 1.67ꢀ1.55 (m, 1 H),

complex (1.24 g, 7.76 mmol, 3 equiv) in DMSO (4.3 mL) while an

internal temperature <ꢀ5 °C was maintained. After the mixture was

stirred for 20 min at ꢀ10 °C, the cooling bath was removed and the

mixture was allowed to warm to room temperature. This mixture was

poured into a 250 mL separatory funnel, and the layers were separated.

1.53ꢀ1.40 (m, 2 H), 1.34 (s, 3 H, H C(8)), 1.07 (d, J = 6.7, 3 H,

3

13

H

3 3

C(9)); C NMR (126 MHz, CDCl ) 176.0 (C(2)), 75.4 (C(7b)),

72.4 (C(1)), 58.1 (C(5a)), 51.9 (C(7a)), 50.5 (C(4)), 42.1 (C(5)), 30.8

(C(6)), 25.2 (C(7)), 24.8 (C(9)), 17.5 (C(8)); IR (NaCl plate, film)

3

1

8

338 (br), 2959 (s), 2870 (s), 1685 (s), 1462 (m), 1412 (m), 1334 (m),

269 (w), 1223 (s), 1142 (m), 1115 (w), 1097 (w), 1044 (w), 964 (w),

80 (w); MS (ESI, Q-tof) 195.1 (67), 180.1 (100), 167.1 (12), 152.1

The aqueous extract was washed with Et O (2 ꢁ 25 mL). The organic

2

extracts were washed with satd aq CuSO

4

solution (2 ꢁ 25 mL), satd aq

NaHCO

3

solution (1 ꢁ 25 mL), and brine (25 mL), and the combined

(

56), 138.1 (14), 111 (43), 96.1 (15), 81.1 (24); TLC R 0.20 (hexanes/

f

organic extracts were dried over MgSO , filtered through Celite (20 mm ꢁ

4

EtOAc, 1:1) [I

2 2

]. Anal. Calcd for C11H17NO (195.26): C, 67.66; H,

5

cm, cotton plug), and rinsed with TBME (3 ꢁ 10 mL). The filtrate is

8.78; N, 7.17. Found: C, 67.41; H, 9.05; N, 7.19.

concentrated in vacuo (15 mmHg, 20ꢀ25 °C) to afford 498 mg (99%)

1

of 14 as a white solid. Data for 14: H NMR (500 MHz, CDCl

3

) 3.98

(d, J = 18.1, 1 H, HC(2)), 3.49 (d, J = 18.1, 1 H, HC(2)), 3.39 (dd, J =

6

8

.4, 12.8, 1 H, HC(4)), 3.23 (dd, J = 12.4, 12.4, 1 H, HC(4)), 2.59 (d, J =

.1, 1 H, HC(7a)), 2.28ꢀ2.23 (m, 1 H, HC(7)), 2.15ꢀ2.07 (m, 2 H,

HC(6), HC(5a)), 2.09ꢀ1.96 (m, 2 H, HC(7), HC(5)), 1.66 (s, 3 H,

H C(8)), 1.46ꢀ1.38 (m, 1 H, HC(6)), 1.10 (d, J = 6.5, 3 H, H C(9)),

3

13

0

8

3

.8ꢀ2.5 (br, 3 H, (H

3

B) ); C NMR (126 MHz, CDCl

3

) 212.4 (C(1)),

7.3 (C(7b)), 72.0 (C(4)), 70.3 (C(2)), 60.9 (C(5a)), 57.9 (C(7a)),

8.1 (C(5)), 32.54 (C(7 or 6)), 32.45 (C(7 or 6)), 24.8 (C(8)), 17.5

(C(9)); IR (NaCl plate, film) 2965 (s), 2964(s), 2963 (s), 2962 (m),

Preparation of (1R,3R,5R,5aS,7aS,7bR)-Octahydro-1-[N-(3,5-Dinitro-

phenyl)carbamoxy]-7b-methyl-2H-cyclopenta[gh]pyrrolizin-2-one

1757 (s), 1457 (w), 1383 (w), 1215 (s), 1163 (s), 745 (s); MS (EI, 70

eV) 192 (100), 180 (50), 166 (32), 151 (71), 149 (67), 136 (32), 123

(19), 95 (44), 81 (100); mol formula C H BNO (193.09); HRMS

(

S6). A 25 mL round-bottomed flask (A) was fitted with a reflux

1

1 20

4

283

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

þ

C

0

11

H

19BNO (192.1560) calcd 192.1560, found 192.1560; TLC R

f

time gas evolution was observed. After the mixture was stirred for 2 h

at room temperature, the reaction was quenched with methanol

(5 mL), and the resulting solution was concentrated by rotary

evaporation (15 mmHg, 20ꢀ25 °C) to afford crude S7 as a mixture

.45 (CH Cl /Et O, 19:1) [I , CAM].

2 2 2 2

Preparation of Scaffold for Library V via (R-Keto Lactam 17).

1

of diastereomers (5/9, H NMR). The crude mixture was purified by

silica gel column chromatography (1.8 cm ꢁ 8 cm column, gradient

2 2

elution, CH Cl /ether, 99:1, 49:1, 24:1, 47:3, 9:1, 25 mL each) to

1

afford 22 mg (35%) of S7 as a white solid. Data for S7: H NMR (500

MHz, CDCl ) 3.88 (dd, J = 4.4, 10.6, 1 H, HC(1)), 3.74 (d, J = 10.8, 1

3

H, OH)), 3.56 (d, J = 12.7, 1 H, HC(2)), 3.37 (dd, J = 4.4, 12.7, 1 H,

HC(2)), 3.24 (dd, J = 5.2, 12.7, 1 H, HC(4)), 2.85 (dd, J = 12.2, 12.2, 1

H, HC(4)), 2.39 (dd, J = 9.1, 9.1, 1 H, HC(7a)), 2.11ꢀ2.04 (m, 1 H,

HC(7)), 1.91ꢀ1.84 (m, 2 H, HC(5a), HC(5)), 1.77ꢀ1.68 (m, 1 H,

Preparation of (3R,5S,5aS,7aS,7bR)-Octahydro-5-methyl-7b-methyl-

H-cyclopenta[gh]pyrrolizin-1,2-one (17). To a 50 mL, three-necked,

round-bottomed flask fitted with a nitrogen inlet adapter, a magnetic stir

bar, and an internal temperature probe were added 5 (491 mg, 2.51

2

HC(6)), 1.66ꢀ1.63 (m, 1 H, HC(6)), 1.60 (s, 3 H, H

3

C(8)),

C(9)),

1

0

.41ꢀ1.32 (m, 1 H, HC(7)), 1.01 (d, J = 5.8, 3 H, H

92 13

.8ꢀ2.5 (br, 3 H, (H B) ); C NMR (126 MHz, CDCl ) 88.5

3

2 2

mmol), CH Cl (10.5 mL), and DMSO (2.1 mL). The resulting

(

C(7b)), 76.0 (C(1)), 71.1 (C(4)), 69.2 (C(2)), 62.1 (C(5a)), 61.4

C(7a)), 34.8 (C(5)), 31.2 (C(7)), 29.6 (C(6)), 26.0 (C(8)), 16.5

C(9)); IR (CDCl

solution was cooled to ꢀ12 °C in an ice/salt bath. To this solution

was added diisopropylethylamine (2.2 mL, 12.6 mmol, 5.0 equiv)

3

) 3492 (s), 2964 (s), 2922 (s), 2869 (s), 2369 (s),

3

followed by the dropwise addition of a solution of SO pyridine

3

2

318 (s), 2270 (s), 1454 (s), 1412 (m), 1381 (m), 1348 (m), 1250

complex (1.2 g, 7.54 mmol, 3.0 equiv) in DMSO (5.2 mL) while an

internal temperature <ꢀ5 °C was maintained. After being stirred for 20

min at ꢀ10 °C, the mixture was allowed to warm to room temperature.

(

m), 1186 (s), 1160 (m) 1121 (m), 1048 (s), 1023 (m), 996 (m), 973

m), 925 (m), 869 (m), 844 (w), 818 (w), 802 (w); MS (EI, 70 eV)

1

94.2 (78), 181 (24), 166 (100), 150 (3), 138 (12), 124 (10), 110

2 2

The mixture was then cooled to 0 °C and diluted with CH Cl (25 mL)

(

44), 96 (11), 84 (14); mol formula C H BNO (195.11); HRMS

1

22

and H O (25 mL). This mixture was poured into a 250 mL separatory

2

C

11

H

21BNO (194.1716) calcd 194.1716, found 194.1720; TLC R

Cl /ether, 9:1) [I ].

f

funnel, and the layers were separated. The aqueous extract was washed

0

.61 (CH

2

with CH

2

Cl

2

(2 ꢁ 25 mL). The organic extracts were washed with satd

aq CuSO solution (2 ꢁ 25 mL), satd aq NaHCO solution (1 ꢁ

4

3

2

5 mL), and brine (25 mL) and the combined organic extracts dried over

MgSO

4

, filtered (cotton plug), and rinsed with TBME (3 ꢁ 10 mL). The

resulting pale-yellow oil was purified by silica gel column chromatogra-

phy (10 mm ꢁ 10 cm column, gradient elution, hexanes/EtOAc, 3:1,

1:1, 1:3, 1:9, 100 mL each) followed by recrystallization (hexanes/

TBME, 1:1, 25 mL) to afford 485 mg (82%) of 17 as white needles and as

1

a single diastereomer. Data for 17: mp 83ꢀ84 °C (MTBE/hexanes); H

NMR (500 MHz, CDCl ) 4.36 (ddd, J = 1.2, 7.7, 12.3, 1 H, HC(4)), 2.98

Preparation of (1S,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-5-methyl-

7b-methyl-2H-cyclopenta[gh]pyrrolizine Borane (12). To a 250 mL

round-bottomed flask equipped with a nitrogen inlet adapter, a rubber

3

(dd, J = 7.8, 12.0, 1 H, HC(4)), 2.71 (dd, J = 5.6, 9.0, 1 H, HC(7a)),

3

2

.13ꢀ2.04 (m, 1 H, HC(7)), 2.06ꢀ1.97 (m, 3 H, HC(6), HC(5a),

HC(5)), 1.77ꢀ1.70 (m, 1 H, HC(7)), 1.45 (s, 3 H, H C(8)), 1.40ꢀ1.32

septum, and a magnetic stir bar were added sequentially 5 (931 mg,

3

(

7

3

m, 1 H, HC(6)), 1.20 (d, J = 5.9, 3 H, H C(9)), 0.8ꢀ2.5 (br, 3 H,

4.77 mmol), THF (23.9 mL, 0.2 M), and BH THF complex

3

92

13

H

3

B) ) ; C NMR (126 MHz, CDCl

4.4 (C(7b)), 57.0 (C(5a)), 55.8 (C(7a)), 51.9 (C(4)), 41.0 (C(5)),

2.8 (C(6)), 30.7 (C(7)), 25.1 (C(8)), 19.8 (C(9)); IR (NaCl plate,

3

) 203.6 (C(1)), 160.4 (C(2)),

(23.9 mL, 1.0 M solution, 10.0 equiv) dropwise over 5 min during

which time gas evolution was observed. After the mixture was stirred

for 4 h at room temperature, the reaction was quenched with methanol

(40 mL), and the resulting mixture was concentrated by rotary

evaporation (15 mmHg, 20ꢀ25 °C). The resulting colorless oil was

purified by silica gel column chromatography (2 cm ꢁ 15 cm column,

film) 3496 (w), 3404 (w), 2961 (s), 2873 (s), 1760 (s), 1713 (s), 1462

s), 1401 (s), 1332 (w), 1226 (s), 1180 (s), 1084 (m), 678 (w); MS (EI,

0 eV) 193 (26), 165 (26), 150 (32), 122 (6), 107 (11), 93 (5), 81

100); TLC R 0.35 (hexanes/EtOAc, 3:1) [I ]. Anal. Calcd for

(

7

(

gradient elution, CH Cl /ether, 99:1, 49:1, 24:1, 47:3, 9:1, 100 mL

2 2

f

2

C

11

H

15N (193.24): C, 68.37; H, 7.82; N, 7.25. Found: C, 68.57; H,

each) to afford 875 mg (94%) of 12 as a white solid. Data for 12: mp

1

8.02; N, 7.31.

86.5ꢀ87.5 °C; H NMR (500 MHz, CDCl

3

) 4.81 (ddd, J = 7.2, 7.2,

0.5, 1 H, HC(1)), 3.46 (dd, J = 6.6, 10.3, 1 H, HC(2)), 3.17 (dd, J =

.2, 12.5, 1 H, HC(4)), 3.02 (dd, J = 10.5, 10.5, 1 H, HC(2)), 2.82 (dd,

1

6

J = 12.5, 12.5, 1 H, HC(4)), 2.34 (dd, J = 8.2, 17.0, 1 H, HC(7a)),

2

1

H

.11ꢀ2.01 (m, 1 H, HC(5)), 1.94ꢀ1.85 (m, 2 H, HC(7), HC(5a)),

.81ꢀ1.76 (m, 1 H, HC(6)), 1.76ꢀ1.67 (m, 2 H, HC(7), HC(6)),

3

.56 (br, s, 1 H, OH), 1.50 (s, 3 H, H C(8)), 1.03 (d, J = 6.4, 3 H,

9

2

13

3

C(9)), 0.8ꢀ2.5 (br, 3 H, (H

3 3

B) ); C NMR (126 MHz, CDCl )

8

8.1 (C(7b)), 70.0 (C(4)), 68.8 (C(1)), 65.7 (C(2)), 62.0 (C(5a)),

55.1 (C(7a)), 34.6 (C(5)), 29.7 (C(7)), 26.4 (C(6)), 25.7 (C(8)), 16.4

(C(9)); IR (CHCl film) 3447 (br), 2962 (s), 2963 (s), 2964 (s), 2965

Preparation of (1R,3R,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-5-methyl-

b-methyl-2H-cyclopenta[gh]pyrrolizine Borane (12) and

7

(

2

3

1S,3R,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-5-methyl-7b-methyl-

H-cyclopenta[gh]pyrrolizine Borane (S7). To a 25 mL, two-

(m), 2378 (s), 2377 (s), 2375 (s), 1474 (m), 1457 (m), 1381 (w),

1214(s), 1166 (s), 1118 (m), 1088 (m), 141 (w), 1010 (w), 995 (w), 946

3

þ

necked, round-bottomed flask equipped with a nitrogen adapter, a

(w), 755 (s); MS (EI, 70 eV) 194 ([M -1],8), 181 (16), 166 (100), 110

2

4

rubber septum, and a magnetic stir bar were added sequentially 14

(60), 96 (12), 82 (8); TLC R

ꢀ11.14 (c = 1.0, CHCl ). Anal. Calcd for C11

67.71; H, 11.37; N, 7.18. Found: C, 67.8; H, 11.31; N, 7.06.

f

0.36 (CH

2

Cl

2

/ether, 9:1) [I

H

2 D

]; [R]

(

63 mg, 0.326 mmol), THF (1.6 mL, 0.2 M), and BH

1.6 mL, 1.0 M solution, 5.0 equiv) dropwise over 5 min during which

3

HF complex

3

22BNO (195.1095): C,

3

4

284

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

. Preparation of Scaffold for Library VI.

FEATURED ARTICLE

6

necked, round-bottomed flask fitted with a nitrogen inlet adaptor, two

rubber septa, an internal temperature probe, and a magnetic stir bar was

added amino alcohol S8 (420 mg, 2.71 mmol) as a solution in THF

(25 mL). To this solution was added triphenylphosphine (0.711 g, 2.71

mmol, 1.0 equiv). The resulting slightly cloudy solution was allowed

room temperature until all of the phosphine was dissolved (15 min) and

was then brought to ꢀ1 °C by use of a salt/ice bath. Diisopropyl

azodicarboxylate (DIAD, 0.53 mL, 2.71 mmol, 1.0 equiv) was then

added dropwise over 5 min while an internal temperature of e3 °C was

maintained. During the addition of the first three-fifths of DIAD, the

orange color quickly dissipated but persisted during the addition of the

last two-fifths. The reaction was allowed to stir for an additional 25 min

during which time it gradually turned from orange to a very light-yellow.

At this point, the reaction was judged to be complete (TLC) and was

Preparation of rel-(2S,2aR,5aS)-6-[6a-Methylperhydrocyclo-

penta[b]pyrrole]methanol (S8). To a 500 mL, three-necked, round-

bottomed flask fitted with a nitrogen inlet adaptor, a magnetic stir bar, a

rubber septum, and an internal temperature probe were added n-butyl vinyl

2 2

ether (1.4 mL, 10.5 mmol, 2.5 equiv), CH Cl (100 mL), and trimethy-

laluminum (10.5 mL, 10.5 mmol, 2.5 equiv, 1.0 M). The resulting clear

solution was allowed to stir for 10 min as the vessel was cooled to ꢀ77 °C

by submersion in an i-PrOH/CO2(s) bath. Nitroalkene (590 mg, 4.2 mmol)

3

quenched by the addition of BH THF solution (1.0 M THF, 13.6 mL,

3

1

3.6 mmol) over 10 min while an internal temperature of e5 °C was

2 2

was dissolved in CH Cl (100 mL) and was transferred to the trimethy-

maintained. The reaction was allowed to stir for an additional 30 min,

laluminum solution by cannulation while an internal temperature <ꢀ70 °C

was maintained (ca. 45 min). The resulting orange solution was allowed to

stir for an additional 3 h during which time the color gradually faded. The

mixture was quenched by the slow, cautious addition of silica gel via a long

stem addition funnel until no bubbling was observed (∼2 g). The mixture

filtered through more silica gel (∼1 g, prewetted with EtOAc) and rinsed

and the remaining borane was quenched by carefully pouring the mixture

into a 250 mL separatory funnel containing H O (50 mL), which

2

bubbled vigorously. The milky white mixture was extracted with

dichloromethane (3 ꢁ 50 mL). The combined organic layers were

4

dried (MgSO ), filtered (cotton plug), and concentrated by rotary

evaporation (15 mmHg, 20ꢀ25 °C). Purification of S9 by silica gel

4

with ethyl acetate (80 mL). The resulting clear filtrate was dried (MgSO ),

chromatography (3 cm ꢁ16 cm, gradient elution, hexanes/EtOAc, 93:7,

filtered, and concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C) to

90:10, 85:15) afforded 329 mg of S9 as a white solid which was then

yield 3.8 g (98%) of the intermediate nitronates as as a clear oil.

sublimed (70 °C, cold water, 0.2 mmHg) to afford 324.0 mg (79%) of

The clear oil was transferred to a 250 mL, single-necked, round-

bottomed ﬂask with benzene (150 mL). The ﬂask was ﬁtted with a reﬂux

condenser, a nitrogen inlet adaptor, and a magnetic stir bar and

immersed in a preheated (85 °C) oil bath. After 12 h, the starting

material was consumed (TLC), and the light-yellow solution was

allowed to cool to room temperature. The solvent was removed by

rotary evaporation (15 mmHg, 20ꢀ25 °C) to give the intermediate

nitrosoacetals as a viscous, light-yellow oil. The crude nitroso acetals

were dissolved in methanol in methanol (20 mL) and transferred to a

test tube (3 ꢁ 14 cm) containing Raney nickel (∼50 mg, previously

analytically pure S9 as a white, waxy solid. Data for S9: mp 154ꢀ156 °C;

1

H NMR (500 MHz, CDCl ) 3.90 (t, J = 10.25, 1 H, HC(1)), 3.19ꢀ3.14

3

(

m, 1 H, HC(7)), 3.06ꢀ3.02 (m, 1 H, HC(7)), 2.93 (dd, J = 6.1, 1 H,

HC(1)), 2.50ꢀ2.46 (m, 2 H, HC(2), HC(5)), 2.10ꢀ1.60 (m, 9 H,

3

13

HC(1)), 1.50 (s, 3 H, HC(9)), 0.8ꢀ2.5 (br, 3 H, (H B) ); C NMR

3

(

3

(

126 MHz, CDCl

6.9 (C(5)), 31.3 (C(3)), 31.1 (C(6)), 29.4 (C(4)), 19.9 (C(9)); IR

KBr plate) 2962 (s), 2968 (s), 2358 (s), 2260 (s), 1772 (w), 1734 (w),

3

) 87.0 (C(8)), 64.3 (C(1)), 63.9 (C(7)), 51.7 (C(2)),

1

716 (w), 1699 (w), 1683 (w), 1652 (w), 1635 (w), 1558 (m), 1521 (m),

540 (m), 1474 (m), 1456 (s), 1376 (s), 1167 (s); MS (EI, 70 ev) 137

washed with H

stir bar. The test tube was placed in a steel autoclave which was then

pressured with H (300 psi) and placed on a stir plate. The reaction was

2

O, MeOH, and EtOAc, 2 ꢁ 10 mL each) and a magnetic

f

(2), 149 (22), 150 (100), 151 (53), 152 (5); TLC R 0.31 (hexane/

EtOAc, 4:1) [CAM]. Anal. Calcd for C H BN (151.06): C, 71.56; H,

9 18

2

12.08; N, 9.27. Found: C, 71.37; H, 12.08; N, 9.22.

stirred under H for 28 h and was then ﬁltered through a pad of Celite,

2

II. Parallel Syntheses: Library Intermediates and Quatern-

washing with 150 mL of methanol. The ﬁltrate was concentrated by

rotary evaporation (15 mmHg, 20ꢀ25 °C) and puriﬁed by silica gel

column chromatography (2 cm ꢁ5 cm, gradient elution: CH Cl /

2

ary Ammonium Bromides. A. Variable Group R : Organometal

Ketone Additions

2

1

. Cerium-Mediated Additions to Keto Amine 13.

MeOH/NH OH, 20:1:0.01, 10:2:.01, 150 mL each) to aﬀord 456 mg

4

(

1

70%, three steps) of amino alcohol S8 as a white solid. Data for S8: mp

1

54ꢀ160 °C; H NMR (500 MHz, CDCl

3

) 3.95 (dd, J = 11.3, 2.0, 1.0, 1

H, HC(1)), 3.67 (dd, J = 11.5, 3.9, 1 H, HC(4)), 2.96 (p, J = 5.60, 1 H,

HC(4)), 2.84ꢀ2.79 (m, 1 H, HC(4)), 2.21 (tt, 1 H), 1.96ꢀ1.80 (m, 2 H,

HC(5), HC(6)), 1.79ꢀ1.71 (m, 1 H, HC(5)), 1.62ꢀ1.51 3 H, HC(6),

1

3

H

2

C(7)) 1.45ꢀ1.40 (m, 1 H, HC(5a)), 1.29, (s, 3 H, H

3

C(8));

C

NMR (126 MHz, CDCl ) 72.4 (C(2a)), 62.1 (C(1)), 50.7 (C(5a)),

3

5

0.4(C(2)), 47.8 (C(4)), 34.8 (C(5)), 30.7 (C(6)), 27.9 (C(7)), 27.6

Preparation of (1S,3S,5aS,7aS,7bR)-Octahydro-1-hydroxy-1-methyl-7b-

methyl-2Hcyclopenta[gh]pyrrolizine Borane (15{2}). To a 25 mL,

two-necked, round-bottomed flask equipped with a nitrogen inlet

(

C(8)); IR (NaCl plate) 329 (s, br), 2949 (s), 2858 (s), 1692 (m), 1682

m), 1633.6 (m), 1455 (m); MS (EI, 70 eV) 155 (8), 97 (10), 96(100),

3

8

0

3 (19), 82(32), 42 (9), 36 (20); GC (method 1) t

R

5.96 min; TLC R

].

f

adapter, a rubber septum, and a magnetic stir bar was added CeCl₃7

H O (559 mg, 1.50 mmol, 1.5 equiv). The flask was immersed in a

2

3

.26 (CH Cl /CH OH,/NH OH, 20:3:0.1) [I , KMnO

2

3

4

2

4

preheated oil bath (140 °C) under vacuum (<0.1 mmHg) for 2 h. The

flask was backfilled with nitrogen, allowed to cool to room tem-

perature, and then immersed in an ice bath. To the solid was added

THF (4.0 mL) via syringe, and the resulting milky, heterogeneous

mixture was allowed to stir vigorously at room temperature for 2 h. The

mixture was then sonicated and stirred intermittently for 15 min

intervals over a 2 h time period. The flask was then immersed in an

ice bath, and methylmagnesium bromide (2.7 M, 556 μL, 1.5 mmol, 1.5

equiv) was added via syringe. The ice bath was removed, and the solution

Preparation of rel-(1S,3aS,5aR,5bR)-1-Boranyl-5b-methylcyclo-

penta[ef]azoniabicyclo[3.2.0]heptane (S9). To a 100 mL, three-

4

285

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

1

was stirred at room temperature for 1.5 h. The flask was immersed in an

ice bath, and ketone 13 (179 mg, 1.0 mmol) was added as a solution in

THF (1.0 mL) via syringe. After the mxiture was stirred for 30 min while

for 15{3}: H NMR (500 MHz, CDCl

3

) 3.98 (ddd, J = 6.6, 11.1, 11.5, 1

H, HC(4)), 3.39 (d, J = 13.8, 1 H, HC(2)), 3.35 (d, J = 9.0, 1 H, HC(2)),

3.34ꢀ3.31 (m, 1 H, HC(4)), 2.46ꢀ2.43 (m, 2 H, HC(5a), HC(5)),

2.20ꢀ2.17 (m, 1 H, HC(7a)), 2.00ꢀ1.90 (m, 2 H, HC(7), HC(6)), 1.80

(m, 1 H, HC(6)), 1.74ꢀ1.67 (m, 1 H, HC(7)), 1.67ꢀ1.60 (m, 1 H,

4

immersed in an ice bath, the reaction was quenched with satd aq NH Cl

solution (5.0 mL). The biphasic mixture was poured into a 125 mL

separatory funnel containing satd aq NH Cl solution (20 mL) and Et O

HC(9)), 1.53ꢀ1.49 (m, 1 H, HC(5)), 1.48 (s, 3 H, H C(8)), 1.40 (br, s,

4

2

3

(

20 mL). The layers were separated, and the aqueous extract was washed

with Et

O (2 ꢁ 25 mL). The organic extracts were washed with satd aq

NaHCO (1 ꢁ 25 mL) and brine (1 ꢁ 25 mL), and the combined

1 H, OH), 0.91 (d, J = 2.4, 3 H, H

3

C(10), 0.90 (d, J = 2.5, 3 H,

0

3

13

2

H

3

C(10 )), 0.8ꢀ2.5 (br, 3 H, (H

3 3

B) ); C NMR (126 MHz, CDCl )

89.0 (C(7b), 82.4 (C(1)), 74.2 (C(2)), 63.2 (C(4)), 62.5 (C(7a)), 51.4

(C(5a)), 39.7 (C(9)), 35.5 (C(6)), 30.0 (C(5)), 27.9 (C(7)), 25.5

(C(8)), 16.9 (C(10)), 16.7 (C(10)); IR (thin film) 3433 (s), 2411 (m),

2319 (s), 2270 (s), 1309 (w), 1280 (S), 1162 (m), 1070 (w), 988 (m),

975 (w), 849 (w), 703 (w); MS (ESI, Q-tof) 222 (89), 209 (50),

194 (34), 178 (18), 166 (16), 124 (21), 110 (65), 96 (100), 81 (20);

3

4

organic extracts were dried over MgSO , filtered (cotton plug), and

concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). The

resulting pale-yellow oil was purified by silica gel column chromatogra-

phy (2 cm ꢁ 8 cm, gradient elution, CH Cl /EtOAc, 1:0, 99:1, 49:1,

2

1

9:1, 50 mL each) to afford 15{2} (134 mg, 68%) as a white solid. Data

1

for 15{2}: H NMR (500 MHz, CDCl

3

) 3.88 (ddd, J = 6.3, 10.6, 10.3, 1

13

C H

26BNO (223.16); HRMS C13

H25BNO (222.2029) calcd 222.2029,

H, HC(4)), 3.37 (d, J = 13.4, 1 H, HC(2)), 3.33 (d, J = 13.6, 1 H,

HC(2)), 3.31ꢀ3.27 (m, 1 H, HC(4)), 2.35ꢀ2.34 (m, 1 H, HC(5a),

HC(5)), 2.16ꢀ2.13 (m, 1 H, HC(7a)), 1.96ꢀ1.89 (m, 2 H, HC(7),

HC(6)), 1.80ꢀ1.68 (m, 2 H, HC(7), HC(6)), 1.54ꢀ1.49 (m, 1 H,

HC(5)), 1.50 (s, 3 H, H C(8), 1.41 (s, 3 H, H C(9)), 0.8ꢀ2.5 (br, 3 H,

found 222.2030; TLC R 0.40 (hexanes/EtOAc, 4:1) [I , CAM].

f

2

3

93

13

(

3 3

H B) ); C NMR (126 MHz, CDCl ) 89.2 (C(7b)), 78.1 (C(1)), 74.7

(C(2)), 63.7 (C(7a)), 63.6 (C(4)), 52.2 (C(5a)), 34.6 (C(6)), 30.5 (C(8

or 9)), 29.6 (C(5)), 27.3 (C(7)), 25.8 (C(8 or 9)); IR (thin film) 3503 (s),

2

1

383 (s), 2329 (s), 2277 (s), 1301 (m), 1258 (m), 1178 (m), 1124 (w),

070 (m), 1018 (w), 953 (w), 850 (w); MS (EI, 70 eV) 194.2 (93), 181

Preparation of (1S,3S,5aS,7aS,7bR)-Octahydro-1-hydroxy-1-phenyl-

(

34), 166 (50), 138 (6), 110 (57), 96 (100), 81 (18); mol formula

7

b-methyl-2H-cyclopenta[gh]pyrrolizine Borane (15{5}). To a 25 mL,

þ

3

C H BNO (195.18); HRMS C H BNO (194.1716) calcd 194.1716,

found 194.1714; TLC R 0.25 (CH Cl /EtOAc, 19:1) [I , CAM].

11

22

11 21

two-necked, round-bottomed flask equipped with a nitrogen inlet adapter,

a rubber septum, and a magnetic stir bar was added CeCl₃7H O (559

f

2

3

2

mg, 1.50 mmol, 1.5 equiv). The flask was immersed in a preheated oil

bath (140 °C) and stirred under vacuum (<0.1 mmHg) for 2 h. The flask

was backfilled with N , allowed to cool to room temperature, and then

2

immersed in an ice bath. To the solid was added THF (4.0 mL), and the

resulting milky, heterogeneous mixture was allowed to stir vigorously at

room temperature for 2 h. The mixture was then sonicated and stirred

intermittently for 15-min intervals over a 2 h time period. The flask was

then immersed in an ice bath, and phenylmagnesium bromide (3.0 M,

500 μL, 1.5 mmol, 1.5 equiv) was added. The ice bath was removed, and

the solution was stirred at room temperature for 1.5 h. The flask was

immersed in an ice bath, and ketone 13 (179 mg, 1.0 mmol) was added

as a solution in THF (1.0 mL) via syringe. After the mixture was stirred

for 30 min while immersed in an ice bath, the reaction was quenched

Preparation of (1S,3S,5aS,7aS,7bR)-Octahydro-1-hydroxy-1-isopro-

pyl-7b-methyl-2H-cyclopenta[gh]pyrrolizine Borane (15{3}). To a

3

25 mL, two-necked, round-bottomed flask equipped with a nitrogen

inlet adapter, a rubber septum, and a magnetic stir bar was added

CeCl 7H O (559 mg, 1.50 mmol, 1.5 equiv). The flask was immersed

3

2

3

in a preheated oil bath (140 °C) and stirred under vacuum (<0.1

mmHg) for 2 h. The flask was backfilled with nitrogen, allowed to cool to

room temperature, and then immersed in an ice bath. To the solid was

added THF (4.0 mL), and the resulting milky, heterogeneous mixture

was allowed to stir vigorously at room temperature for 2 h. The mixture

was then sonicated and stirred intermittently for 15 min intervals over a

with satd aq NH Cl solution (5.0 mL). The biphasic mixture was poured

4

into a 125 mL separatory funnel containing satd aq NH

(20 mL) and Et O (20 mL). The layers were separated, and the aqueous

extract was washed with Et

O (2 ꢁ 25 mL). The organic extracts were

washed with satd aq NaHCO (1 ꢁ 25 mL) and brine (1 ꢁ 25 mL), and

4

Cl solution

2

3

4

the combined organic extracts were dried over MgSO , filtered (cotton

2

h time period. The flask was then immersed in an ice bath, and

plug), and concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C).

The resulting pale-yellow oil was purified by silica gel column chroma-

tography (1.8 cm ꢁ8 cm, gradient elution, CH Cl /EtOAc, 1:0, 99:1,

isopropylmagnesium chloride (1.9 M, 811 μL, 1.5 mmol, 1.5 equiv) was

added via syringe. The ice bath was removed, and the solution was stirred

at room temperature for 1.5 h. The flask was immersed in an ice bath,

and ketone 13 (179 mg, 1.0 mmol) was added as a solution in THF

2

49:1, 19:1, 50 mL each) to afford 15{5} (131 mg, 51%) as a white solid.

1

Data for 15{5}: H NMR (500 MHz, CDCl

3

) 7.42ꢀ7.36 (m, 4 H,

(

1.0 mL) via syringe. After the mixture was stirred for 30 min while

immersed in an ice bath, the reaction was quenched with satd aq NH Cl

solution (5.0 mL). The biphasic mixture was poured into a 125 mL

separatory funnel containing satd aq NH Cl solution (20 mL) and Et

20 mL). The layers were separated, and the aqueous extract was washed

HC(10), HC(11)), 7.33ꢀ7.28 (m, 1 H, HC(12)), 4.28ꢀ4.20 (m, 1 H,

HC(4)), 3.81 (d, J = 13.9, 1 H, HC(2)), 3.68 (d, J = 13.9, 1 H, HC(2)),

3.45ꢀ3.40 (m, 1 H, HC(4)), 2.75 (dd, J = 3.0, 8.3, 1 H, HC(7a)), 2.45ꢀ

2.36 (m, 2 H, HC(5a), HC(5)), 2.12ꢀ2.02 (m, 2 H, HC(7), HC(6)),

1.94ꢀ1.78 (m, 3 H, HC(7), HC(6), OH), 1.63ꢀ1.58 (m, 1 H, HC(5)),

4

2

O

(

92

13

with Et

NaHCO

2

O (2 ꢁ 25 mL). The organic extracts were washed with satd aq

1.58 (s, 3 H, H

3

C(8)), 0.8ꢀ2.5 (br, 3 H, (H

3

B) ); C NMR (126 MHz,

3

(1 ꢁ 25 mL) and brine (1 ꢁ 25 mL), and the combined

CDCl ) 145.1 (C(9)), 129.0 (C(10 or 11)), 128.0 (C(10 or 11)), 124.9

3

organic extracts were dried over MgSO , filtered (cotton plug), and

(C(12)), 89.2 (C(7b)), 82.1 (C(1)), 74.9 (C(2)), 64.3 (C(7a)), 63.3

(C(4)), 52.2 (C(5a)), 35.2 (C(6)), 29.7 (C(5)), 27.3 (C(7)), 25.9

(C(8)); IR (thin film) 3433 (s), 2411 (m), 2319 (s), 2270 (s), 1309

(w), 1280 (S), 1162 (m), 1070 (w), 988 (m), 975 (w), 849 (w), 703 (w);

4

concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). The

resulting pale-yellow oil was purified by silica gel column chromatogra-

phy (1.8 cm ꢁ 8 cm, gradient elution, CH

2 2

Cl /EtOAc, 1:0, 99:1, 49:1,

þ

19:1, 50 mL each) to afford 15{3} (89 mg, 40%) as a white solid. Data

MS (EI, 70 eV) 254.2 (62), 243.2 ([M ], 66), 228 (20), 210 (10), 178 (8),

4

286

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

1

58 (48), 149 (15), 138 (11), 123 (46), 110 (100); mol formula

inlet adapter, a rubber septum, and a magnetic stir bar was added

CeCl₃7H O (2.90 g, 7.77 mmol, 1.5 equiv). The flask was immersed in

C H BNO; 257.18; HRMS C H NO (243.1623) calcd 243.16232,

found 243.16251; TLC R

16

24

16 21

3

2

f

2 2 2

0.20 (CH Cl /EtOAc, 19:1) [I , CAM].

a preheated oil bath (140 °C) and stirred under vacuum (∼0.1 mmHg).

After 2 h, the flask was backfilled with nitrogen, allowed to cool to room

temperature, and then immersed in an ice bath. To the solid was added

THF (22 mL), and the resulting milky, heterogeneous mixture was

allowed to stir vigorously at room temperature for 2 h. The mixture was

then sonicated and stirred intermittently for 15 min intervals over a 2 h

time period. The flask was then immersed in an ice bath, and isopro-

pylmagnesium chloride (1.8 M, 4.3 mL, 1.50 mmol, 1.5 equiv) was

added via syringe. The ice bath was removed, and the solution was stirred

at room temperature for 1.5 h. The flask was immersed in an ice bath,

and ketone 14 (1.0 g, 5.18 mmol) was added as a solution in THF

Cerium-Mediated Additions to Keto Amine 14

Cerium-Mediated Additions to 14. Preparation of (1S,3S,5S,-

5

2

aS,7aS,7bR)-Octahydro-1-hydroxy-1-methyl-5-methyl-7b-methyl-

H-cyclopenta[gh]pyrrolizine Borane (16{2}). To a 100 mL, two-

3

(

4.0 mL). After the mixture was stirred for 30 min while immersed in an

ice bath, the reaction was quenched with satd aq NH Cl solution

20 mL). The biphasic mixture was poured into a 250 mL separatory

funnel containing satd aq NH Cl solution (30 mL) and Et O (40 mL).

necked, round-bottomed flask equipped with a nitrogen inlet adapter, a

rubber septum, and a magnetic stir bar was added CeCl 7H

3 2

O (2.30 g,

4

3

(

8.70 mmol, 1.5 equiv). The flask was immersed in a preheated oil bath

(

140 °C) and stirred under vacuum (∼0.1 mmHg). After 2 h, the flask

4

2

The layers were separated, and the aqueous extract was washed with

Et

O (2 ꢁ 50 mL). The organic extracts were washed with satd aq

NaHCO (1 ꢁ 50 mL) and brine (1 ꢁ 50 mL), and the combined

was backfilled with nitrogen, allowed to cool to room temperature, and

then immersed in an ice bath. To the solid was added THF (25 mL), and

the resulting milky, heterogeneous mixture was allowed to stir vigorously

at room temperature for 2 h. The mixture was then sonicated and stirred

intermittently for 15 min intervals over a 2 h time period. The flask was

then immersed in an ice bath, and methylmagnesium bromide (3.0 M,

2

3

4

organic extracts were dried over MgSO , filtered (cotton plug), and

concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). The

resulting pale-yellow oil was purified by silica gel column chro-

matography (2 cm ꢁ 10 cm, gradient elution, CH Cl /ether, 1:0,

3.0 mL, 1.50 mmol, 1.5 equiv) was added . The ice bath was removed,

2

9

9:1, 49:1, 24:1, 47:3, 9:1, 17:3, 100 mL each) to afford 16{3}

and the solution was stirred at room temperature for 1.5 h. The flask was

immersed in an ice bath, and ketone 14 (1.12 g, 5.80 mmol) was added as

a solution in THF (4.0 mL). After the mixture was stirred for 30 min

while immersed in an ice bath, the reaction was quenched with satd aq

1

(931 mg, 79%) as a pale yellow oil. Data for 16{3}: H NMR

500 MHz, CDCl ) 3.60 (d, J = 13.5, 1 H, HC(2)), 3.51 (dd, J = 7.9,

(

3

1

2.2, 1 H, HC(4)), 3.31 (d, J = 13.5, 1 H, HC(2)), 3.16 (dd, J = 10.7,

2.1, 1 H, HC(4)), 2.42ꢀ2.32 (m, 1 H, HC(5)), 2.18 (dd, J = 7.0, 9.1,

H, HC(7a)), 2.09ꢀ1.98 (m, 1 H, HC(7)), 1.89ꢀ1.83 (m, 3 H,

NH

50 mL separatory funnel containing satd aq NH

and Et O (40 mL). The layers were separated, and the aqueous extract

4

Cl solution (20 mL). The biphasic mixture was poured into a

2

4

Cl solution (30 mL)

HC(6), HC(6), HC(5a)), 1.82ꢀ1.69 (m, 2 H, HC(7), HC(9)),

2

1

3

.49 (s, 3 H, H

3

C(8)), 1.37 (br, s, 1 H, OH), 0.99 (d, J = 6.6,

C(10)), 0.87 (d,

was washed with Et

with satd aq NaHCO

combined organic extracts were dried over MgSO , filtered (cotton

2

O (2 ꢁ 50 mL). The organic extracts were washed

H, H C(12)), 0.91 (d, J = 6.8, 3 H, H

3

(1 ꢁ 50 mL) and brine (1 ꢁ 50 mL), and the

9

2

13

J = 6.9, 3 H, H

3

C(11)), 0.8ꢀ2.5 (br, 3 H, (H

3

B) ); C NMR (126

4

MHz, CDCl ) 89.4 (C(7b)), 80.2 (C(1)), 77.1 (C(2)), 74.1 (C(4)),

plug), and concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C).

3

60.9 (C(7a)), 60.0 (C(5a)), 38.3 (C(9)), 34.5 (C(5)), 31.3 (C(6)), 25.4

The resulting pale-yellow oil was purified by silica gel column chroma-

(

C(7)), 25.3 (C(8)), 17.6 (C(12)), 17.0 (C(10)), 16.8 (C(11)); IR

CHCl ) 3507 (br), 2964 (s), 2866 (m), 2380 (s), 2326 (s), 2268 (s),

tography (2 cm ꢁ12 cm, gradient elution, CH

2 2

Cl /ether, 1:0, 99:1, 49:1,

24:1, 9:1, 4:1, 100 mL each) to afford 16{2} (1.1 g, 87%) as a white solid.

3

1

1465 (m), 1171 (m), 1000 (w); MS (EI, 70 eV) 236 (18), 234 (24), 223

23), 208 (30), 192 (7), 180 (15), 124 (36), 110 (100), 96 (12), 81 (10);

mol formula C14 28BNO (237.19); HRMS C14 27BNO (236.2186)

calcd 236.2186, found 236.2183; TLC R 0.40 (CH Cl /ether, 19:1)

Data for 16{2}: mp 84ꢀ85 °C (MTBE/hexanes); H NMR (500 MHz,

CDCl ) 3.43 (d, J = 12.0, 1 H, HC(2)), 3.28 (dd, J = 6.8, 12.6, 1 H, HC(4)),

.25 (d, J = 12.3, 1 H, HC(2)), 2.97 (dd, J = 12.2, 12.2, 1 H, HC(4)), 2.13

et, J = 4.8, 8.1, 9.6, 12.7, 1 H, HC(5)), 2.06 (ddd, J = 8.2, 8.2, 7.8, 1 H,

HC(7a)), 2.02ꢀ1.96(m, 1H, HC(7)), 1.85(dd, J=7.0, 7.0, 1H, HC(5a)),

C(9)), 1.52

s, 3H, H C(8)), 1.38(d,J= 4.7, 1 H, OH), 0.99 (d, J = 6.5, 3 H, H3C(10)),

(

3

H

3

(

f

2

[I , CAM].

1

(

.83ꢀ1.70 (m, 3 H, HC(7), HC(6), HC(6)), 1.61 (s, 3 H,H

3

92

13

0

7

.8ꢀ2.5 (br, 3 H, (H B) ); C NMR (126 MHz, CDCl ) 88.9 (C(7b)),

3

5.5 (C(1)), 74.9 (C(2)), 73.4 (C(4)), 61.9 (C(7a)), 61.8 (C(5a)), 34.0

(

C(5)), 31.0 (C(9)), 29.7 (C(6)), 27.1 (C(7)), 26.0 (C(8)), 16.5

C(10)); IR (CHCl ) 3969 (br), 2962 (s), 2870 (s), 2380 (s), 2331 (s),

3

2276 (s), 1459 (s), 1381 (s), 1340 (m), 1180 (s), 1134 (s), 1064 (s), 1049

(

w), 1019 (w), 960 (s), 871 (w), 826 (m); MS (EI, 70 eV) 208 (4), 195

16), 180 (48), 124 (28), 110 (100), 96 (16), 81 (8); mol formula

Preparation of (1S,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-1-

phenyl-5-methyl-7b-methyl-2H-cyclopenta[gh]pyrrolizine Borane

(16{5}). To a 50 mL, two-necked, round-bottomed flask equipped with

3

C H BNO (209.14); HRMS C H BNO (208.1873) calcd 208.1873,

1

2

24

12 23

found 208.1875; TLC R

f

0.33 (CH

2

Cl

2

/Et

2

O, 19:1) [I

2

, CAM].

a nitrogen inlet adapter, a rubber septum, and a magnetic stir bar was

added CeCl

3

7H

2

O (1.74 g, 4.66 mmol, 1.5 equiv). The flask was

3

immersed in a preheated oil bath (140 °C) and stirred under vacuum

(

∼0.1 mmHg). After 2 h, the flask was backfilled with nitrogen, allowed

to cool to room temperature, and then immersed in an ice bath. To the

solid was added THF (14 mL), and the resulting milky, heterogeneous

mixture was allowed to stir vigorously at room temperature for 2 h. The

mixture was then sonicated and stirred intermittently for 15 min

intervals over a 2 h time period. The flask was then immersed in an

ice bat,h and phenylmagnesium bromide (2.9 M, 1.6 mL, 1.50 mmol, 1.5

equiv) was added via syringe. The ice bath was removed, and the solution

Preparation of (1S,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-1-isopro-

pyl-5-methyl-7b-methyl-2H-cyclopenta[gh]pyrrolizine Borane (16{3}).

3

To a 100 mL, two-necked, round-bottomed flask equipped with a nitrogen

4

287

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

was stirred at room temperature for 1.5 h. The flask was immersed in an

ice bath, and ketone 14 (600 mg, 3.11 mmol) was added as a solution in

THF (2.0 mL) via syringe. After the mixture was stirred for 30 min while

12.7, 1 H, HC(7)), 1.99 (dd, J = 7.3, 15.3, 1 H, HC(5)), 1.86 (ddd, J =

6.4, 6.4, 12.8, 1 H, HC(6)), 1.78 (ddd, J = 2.1, 7.4, 9.8, 1 H, HC(5a)),

1.64ꢀ1.56 (m, 1 H, HC(7)), 1.40 (s, 3 H, H

3

C(8)), 1.11 (d, J = 7.1, 3 H,

C(10)),

immersed in an ice bath, the reaction was quenched with satd aq NH

4

Cl

H

3

C(11)), 1.09ꢀ1.02 (m, 1 H, HC(6)), 0.97 (s, 9 H, H

3

9

3

13

solution (15 mL). The biphasic mixture was poured into a 250 mL

separatory funnel containing satd aq NH Cl solution (30 mL) and Et O

0.8ꢀ2.5 (br, 3 H, (H B) ); C NMR (126 MHz, CDCl ) 179.3

3

(C(2)), 83.2 (C(1 or 7b)), 76.6 (C(1 or 7b)), 59.3 (C(5a)), 51.4

(C(4)), 49.6 (C(7a)), 37.8 (C(5)), 37.3 (C(10)), 33.9 (C(6)), 29.8

4

2

(

40 mL). The layers were separated, and the aqueous extract was washed

with Et

O (2 ꢁ 50 mL). The organic extracts were washed with satd aq

NaHCO (1 ꢁ 50 mL) and brine (1 ꢁ 50 mL), and the combined

2

3

(C(7)), 25.0 (C(11)), 23.9 (C(8)), 21.3 (C(9)); IR (CDCl , film) 3404

(br), 2958 (s), 2865 (s), 2248 (w), 1678 (s), 1464 (m), 1395 (m), 1365

(m), 1340 (m), 1230 (w), 1181 (w), 1147 (w), 1115 (w), 1087 (w),

1057 (w), 1015 (w), 994 (w), 910 (w), 766 (w); MS (EI, 70 eV) 251.2

(1), 236.2 (1.8), 218 (1.9), 195.1 (100), 166.1 (2.7), 152.1 (3.1), 140.1

(2.8), 110.1 (8.7); mol formula C H NO (251.36); HRMS

3

4

organic extracts were dried over MgSO , filtered (cotton plug), and

concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). The

resulting pale-yellow oil was purified by silica gel column chromatogra-

phy (1.8 cm ꢁ 10 cm, gradient elution, hexanes/TBME, 9:1, 4:1, 7:3,

1

5

25

2

1

:1, 1:4, 50 mL each) to afford 16{5} (828 mg, 98%) as a white solid.

C

15

H

25NO

2

(251.1885) calcd 251.1885, found 251.1883; TLC R

f

1

Data for 16{5}: mp 81.5ꢀ82 °C (MTBE/hexanes); H NMR (500

0.25 (hexanes/TBME, 3:1) [I ].

2

MHz, CDCl ) 7.37 (d, J = 4.3, 1 H, HC(10), HC(11)), 7.31 (m, 1 H,

3

HC(12)), 4.10 (d, J = 13.7, 1 H, HC(2)), 3.70 (d, J = 13.6, 1 H, HC(2)),

3.68 (dd, J = 8.3, 12.0, 1 H, HC(4)), 3.29 (dd, J = 10.8, 12.1, 1 H,

HC(4)), 2.75 (dd, J = 6.1, 9.3, 1 H, HC(7a)), 2.64ꢀ2.55 (m, 1 H,

HC(5)), 2.34ꢀ2.37 (m, 1 H, HC(5a)), 1.99ꢀ1.93 (m, 3 H, H C(7),

2

HC(6)), 1.93ꢀ1.83 (m, 1 H, HC(6)), 1.80 (br, s, 1 H, OH), 1.55 (s, 3 H,

9

2

H

3

C(8)), 1.02 (d, J = 6.6, 3 H, H

3

C(13)), 0.8ꢀ2.5 (br, 3 H, (H

3

B) );

1

3

C NMR (126 MHz, CDCl

C(12)), 125.1 (C(10)), 89.6 (C(7b), 79.3 (C(1)), 78.8 (C(2)),

4.5 (C(4)), 62.6 (C(5a)), 59.8 (C(7a)), 34.7 (C(5)), 31.5 (C(7)),

5.1 (C(8)), 24.1 (C(6)), 17.7 (C(13)); IR (CHCl film) 3475 (br),

3

) 146.0 (C(9)), 128.9 (C(11)), 128.0

(

Preparation of (1S,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-

1-(2,4,6-trimethylphenyl)-5-methyl-7b-methyl-2H-cyclopenta[gh]-

7

2

1

7

1

3

pyrrolizine Borane (18{6}). To a 25 mL, two-necked, round-bot-

3

960 (s), 2871 (s), 2377 (s), 2327 (s), 2271 (s), 1494 (w), 1447 (m),

266 (w), 1172 (s), 1131 (m), 1067 (w), 1004 (m), 954 (w), 876 (w),

00 (s); MS (EI, 70 eV) 268.2 (24), 257.2 (23), 242.1 (11), 224.1 (3),

92.1 (4), 178.1 (4), 158 (14), 137.1 (11), 124.1 (26), 110.1 (100);

tomed flask equipped with a nitrogen inlet adapter, a rubber septum,

and a magnetic stir bar was added 17 (100 mg, 0.517 mmol) followed by

THF (2.7 mL). The flask was immersed in an ice bath, and mesitylmag-

nesium bromide (0.39 M, 2.0 mL, 1.5 equiv) was added dropwise via

syringe. After the mixture was stirred for 10 min, the cooling bath was

removed, and the solution was stirred at room temperature for 20 min.

The reaction flask was immersed in an ice bath, and the reaction was

mol formula C H BNO (271.21); HRMS C H NO (257.1780)

17

26

17 23

calcd 257.1780, found 257.1775; TLC R

, CAM].

. Grignard Additions to Keto Amide 17.

f

0.58 (hexanes/TBME, 1:1)

[I

2

3

quenched with satd aq NH Cl solution (15 mL). The biphasic mixture

4

was poured into a 125 mL separatory funnel containing satd aq NH

solution (10 mL) and Et O (30 mL). The layers were separated, and the

aqueous extract was washed with Et

O (2 ꢁ 25 mL). The organic extracts

were washed with H O (1 ꢁ 25 mL) and brine (1 ꢁ 25 mL), and the

4

Cl

2

4

combined organic extracts were dried over MgSO , filtered (cotton plug),

and concentrated by rotary evaporation (15 mmHg, 20ꢀ25 °C). The

resulting pale-yellow oil was purified by silica gel column chromatography

(1.8 cm ꢁ 8 cm, gradient elution, hexanes/EtOAc, 9:1, 4:1, 7:3, 1:1,

25 mL each) to afford 18{6} (157 mg, 97%) as a white solid. Data for

Preparation of (1R,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-

-(tert-butyl)-5-methyl-7b-methyl-2H-cyclopenta[gh]pyrrolizin-2-one

18{4}). To a 25 mL, two-necked, round-bottomed flask equipped with

a nitrogen inlet adapter, a rubber septum, and a magnetic stir bar was

added 17 (100 mg, 0.518 mmol) followed by THF (5.0 mL). The flask

was immersed in an acetone/CO₂₍_s₎bath and stirred for 20 min. Then,

t-BuLi (1.6 M, 648 μL, 2.0 equiv) was added dropwise over 5 min via

syringe. After the mixture was stirred for 10 min, the acetone/CO2(s)

bath was replaced with an ice bath and the reaction was stirred for 1 h.

1

(

1

18{6}: mp 101ꢀ102 °C (hexanes/EtOAc); H NMR (500 MHz,

3

CDCl ) 6.92 (s, 1 H, HC(11)), 6.68 (s, 1 H, HC(11)), 4.17 (dd, J =

7.3, 11.7, 1 H, HC(4)), 3.47 (s, 1 H, OH), 2.65 (dd, J = 7.6, 10.1, 1 H,

HC(4)), 2.64 (dd, J = 10.4, 11.6, 1 H, HC(5)), 2.48 (s, 3 H, H C(8, 13, or

14), 2.41 (s, 3 H, H C(8, 13, or 14)), 2.22 (s, 3 H, H C(8, 13, or 14)),

3

1.98ꢀ1.92 (m, 1 H, HC(7)), 1.84ꢀ1.79 (m, 1 H, HC(7a)), 1.78ꢀ1.66

(m, 3 H, HC(7), HC(6), HC(5a)), 1.55ꢀ1.48 (m, 1 H, HC(6)), 1.09 (d,

The reaction was quenched by the dropwise addition of satd aq NH Cl

J = 6.7, 3 H, H

3

C(15)), 1.08 (s, 3 H, H

3

C(8, 13, or 14)), 0.8ꢀ2.5 (br, 3 H,

4

9

2

13

solution (15 mL) and further diluted with Et

mixture was poured into a 125 mL separatory funnel containing satd aq

NH Cl solution (10 mL) and Et O (30 mL). The layers were separated,

and the aqueous extract was washed with Et O (2 ꢁ 25 mL). The

2

O (10 mL). The biphasic

3 3

(H B) ); C NMR (126 MHz, CDCl ) 179.1 (C(2)), 139.4 (C(9)),

138.3 (C(10 or 12)), 136.6 (C(10 or 12)), 133.9 (C(10 or 12)), 131.9

(C(11)), 130.8 (C(11)), 82.3 (C(1)), 75.3 (C(7b)), 60.2 (C(5)), 58.9

(C(7a)), 50.9 (C(4)), 41.7 (C(5a)), 29.9 (C(6)), 27.1 (C(7)), 23.8

(C(8, 13 or 14)), 22.3 (C(8, 13, or 14)), 21.2 (C(8, 13, or 14)), 20.7

(C(8, 13, or 14)), 17.7 (C(15)); IR (NaCl plate, film) 3404 (br), 3000

(s), 2959 (s), 2922 (s), 2870 (s), 1693 (s), 1610 (m), 1560 (w), 1460 (s),

1407 (s), 1379 (m), 1351 (m), 1280 (w), 1232 (m), 1120 (m), 1072 (m),

1035 (m), 978 (w), 851 (m); MS (EI, 70 eV) 313 (10), 295 (84), 285 (4),

280 (6), 270 (3), 252 (5), 147 (100), 138 (10), 124 (21), 119 (29), 110

(90); mol formula C H NO (313.43); HRMS C H NO

4

2

organic extracts were washed with H

1 ꢁ 25 mL), and the combined organic extracts were dried over

Na SO , filtered (cotton plug), and concentrated by rotary evaporation

2

O (1 ꢁ 25 mL) and brine

(

2

4

(

15 mmHg, 20ꢀ25 °C). The resulting pale-yellow oil was purified by

silica gel column chromatography (1.8 cm ꢁ 6 cm, gradient elution,

hexanes/TBME, 17:3, 5:1, 3:1, 3:1, 3:2, 25 mL each) to afford 18{4} (43

1

mg, 33%) as a white solid. Data for 18{4}: H NMR (500 MHz, CDCl )

3

20 27

2

20 27

2

4

2

.30 (dd, J = 8.6, 12.2, 1 H, HC(4)), 2.63 (dd, J = 7.3, 12.2, 1 H, HC(4)),

.55 (s, 1 H, HC(OH)), 2.38 (d, J =8.0, 1 H, HC(7a)), 2.07 (dd, J = 5.3,

(313.2042) calcd 313.2042, found 313.2045; TLC R 0.32 (hexanes/

f

EtOAc, 3:1) [I , CAM].

2

4

288

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

The Journal of Organic Chemistry

FEATURED ARTICLE

gel column chromatography (1.8 cm ꢁ 8 cm column, gradient elution,

hexanes/TBME, 19:1, 9:1, 3:1, 3:2, 25 mL each) to afford 27 mg (93%) of

1

6{4} as a white solid. Data for 16{4}: H NMR (500 MHz, CDCl

3

) 3.80

(d, J = 13.9, 1 H, HC(2)), 3.63 (dd, J =8.5, 12.0, 1 H, HC(4)), 3.23 (d, J =

1

3.8, 1 H, HC(2)), 3.23ꢀ3.18 (m, 1 H, HC(4)), 2.48ꢀ2.40 (m,

H, HC(7a)), 2.41 (dd, J = 6.8, 9.2, 1 H, HC(5)), 2.09ꢀ1.99 (m, 1 H,

Preparation of (1R,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-

-[1]naphthyl-5-methyl-7b-methyl-2H-cyclopenta[gh]pyrrolizin-2-

HC(7)), 1.91ꢀ1.84 (m, 3 H, H C(6), HC(5a)), 1.72ꢀ1.63 (m, 1 H,

2

1

HC(7)), 1.62 (br, s, 1 H, HO), 1.46 (s, 3 H, H

C(9)), 0.92 (s, 3 H, H

C(11)), 0.8ꢀ2.5 (br, 3 H, (H

(126 MHz, CDCl ) 89.7 (C(1 or 7b)), 83.0 (C(1 or 7b)), 75.7 (C(2)),

3

C(8)), 0.99 (d, J = 6.6, 3 H,

9

3

13

one (18{7}). To a 25 mL, two-necked, round-bottomed flask equipped

with a nitrogen inlet adapter, a rubber septum, and a magnetic stir bar

was added 17 (100 mg, 0.517 mmol) followed by THF (2.7 mL). The

flask was immersed in an ice bath, and 1-naphthylmagnesium bromide

H

3

B) ); C NMR

3

74.6 (C(4)), 59.6 (C(5a)), 56.5 (C(7a)), 37.5 (C(10)), 35.0 (C(5)), 32.1

(C(6)), 25.5 (C(10)), 25.4 (C(7)), 24.7 (C(8)), 18.1 (C(11)); IR

(0.4 M, 2.0 mL, 1.5 equiv) was added dropwise via syringe. After being

(CDCl film) 3503 (s), 2961 (s), 2872 (s), 2376 (s), 2327 (s), 2271

3

stirred for 10 min, the cooling bath was removed, and the reaction was

stirred at room temperature for 20 min. The biphasic mixture was

(s), 1465 (m), 1400 (w), 1378 (m), 1335 (w), 1267 (w), 1174 (m), 1140

(m), 1091 (w), 1057 (w), 1002 (w), 980 (w), 958 (w), 940 (w), 910 (s),

882 (w), 841 (w), 733 (s), 649 (w); MS (ESI, Q-tof) 422.2 (10), 344.2

poured into a 125 mL separatory funnel containing satd aq NH Cl

4

solution (10 mL) and Et

and the aqueous extract was washed with Et

The organic extracts were washed with H O (1 ꢁ 25 mL) and brine

2

O (30 mL). The layers were separated,

(25), 318.2(40), 308.2(100);molformulaC15

27NO: (237.2093) calcd 237.2093, found 237.2091; TLC R

(hexanes/TBME, 3:1) [I , CAM].

H

30BNO(251.22);HRMS

2

O (2 ꢁ 25 mL).

15

C H

f

0.27

2

(

1 ꢁ 25 mL), and the combined organic extracts were dried over

MgSO , filtered (cotton plug), and concentrated by rotary evaporation

4

(

15 mmHg, 20ꢀ25 °C). The resulting pale-yellow oil was purified by

silica gel column chromatography (1.8 cm ꢁ8 cm, gradient elution,

hexanes/EtOAc, 9:1, 4:1, 7:3, 1:1, 25 mL each) to afford 18{7} (157 mg,

9

5%) as a white solid. Data for 18{7}: mp 150ꢀ151 °C (hexanes/

1

EtOAc); H NMR (500 MHz, CDCl

3

) 8.43 (d, J = 8.4, 1 H, HC(11)),

7

.87 (d, J = 7.9, 1 H, HC(14)), 7.78 (d, J = 7.9, 1 H, HC(16)), 7.56ꢀ7.48

Preparation of (1R,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-1-mesityl-

-methyl-7b-methyl-2H-cyclopenta[gh]pyrrolizine Borane (16{6}). To a

(

4

7

m, 2 H, HC(12), HC(13)), 7.34ꢀ7.28 (m, 2 H, HC(17), HC(18)),

5

3

.27 (dd, J = 6.8, 11.5, 1 H, HC(4)), 3.44 (br, s, 1 H, OH), 2.96 (dd, J =

.6, 10.1, 1 H, HC(7a)), 2.69 (dd, J = 10.0, 11.5, 1 H, HC(4)), 2.21ꢀ2.14

25 mL round-bottomed flask equipped with a nitrogen inlet adapter, a

rubber septum, a reflux condenser, and a magnetic stir bar were added

(

m, 1 H, HC(7)), 2.00ꢀ1.92 (m, 1 H, HC(7)), 1.83ꢀ1.73 (m, 3 H,

sequentially 18{6} (50 mg, 0.160 mmol), THF (200 μL), and BH

3

HC(6), HC(5a), HC(5)), 1.60ꢀ1.55 (m, 1 H, HC(6)), 1.11 (d, J = 6.4, 3

93

13

THF complex (3.2 mL, 1.0 M solution, 20 equiv). The reaction flask was

immersed in an oil bath and heated to reflux (70 °C). After being stirred

for 12 h at reflux, the solution was allowed to reach room temperature

and then was quenched with methanol (5 mL) and concentrated

by rotary evaporation (15 mmHg, 20ꢀ25 °C). The resulting colorless

oil was purified by silica gel column chromatography (1.8 cm ꢁ 8 cm

column, gradient elution, hexanes/TBME, 19:1, 9:1, 4:1, 3:1,25 mL

H, H

3

C(19)), 0.87 (s, 3 H, H

3

C(8)), 0.8ꢀ2.5 (br, 3 H, (H

3

B) ); C

NMR (126 MHz, CDCl ) 177.2 (C(2)), 138.9 (C(9)), 135.0 (C(15)),

3

131.1 (C(10)), 129.3 (C(16)), 129.1 (C(14)), 126.6 (C(11)), 126.0

(C(12 or 13)), 125.7 (C(12 or 13)), 124.7 (C(17 or 18)), 124.7 (C(17 or

1

5

8)), 84.1 (C(1 or 7b)), 75.9 (C(1 or 7b)), 58.9 (C(5a)), 57.9 (C(7a)),

1.2 (C(4)), 41.9 (C(5)), 30.0 (C(6)), 27.0 (C(7)), 24.4 (C(8)), 17.7

(C(19)); IR (NaCl plate, film) 3386 (s), 3050 (w), 3007 (m), 2959 (s),

each) to afford 36 mg (72%) of 16{6} as a white solid. Data for 16{6}:

2

1

871 (s), 1693 (s), 1596 (w), 1509 (m), 1461 (s), 1411 (s), 1352 (m),

376 (m), 1330 (m), 1280 (m), 1230 (s), 1213 (m), 1174 (w), 1131

1

mp 131ꢀ132 °C (MTBE/hexanes); H NMR (500 MHz, CDCl

3

) 6.81

(

s, 2 H, HC(11)), 4.09 (d, J = 14.1, 1 H, HC(9)), 4.02 (d, J = 14.1, 1 H,

(s), 1074 (s), 1030 (w), 967 (w), 915 (w), 862 (w), 802 (s); MS (EI, 70

eV) 321 (19), 293 (6), 183 (9), 165 (13), 155 (68), 127 (46), 110

HC(9)), 3.72 (dd, J = 8.2, 12.1, 1 H, HC(4)), 3.29 (dd, J = 9.9, 12.0, 1 H,

HC(4)), 3.10 (dd, J = 6.0, 8.5, 1 H, HC(7a)), 2.62ꢀ2.52 (m, 1 H,

(100); mol formula C21

2 2

H23NO (321.41); HRMS C21H23NO

(321.1729) calcd 321.1729, found 321.1732; TLC R 0.28 (hexanes/

HC(5)), 2.52 (s, 6 H, H

3

C(13), H

3

C(13)), 2.40ꢀ2.31 (m, 1 H,

f

EtOAc, 3:1) [I

2

, CAM].

HC(7)), 2.22 (s, 3 H, H

.99ꢀ1.93 (m, 3 H, H C(6), HC(5a)), 1.51 (s, 3 H, H

d, J = 6.7, 3 H, H

C(14)), 0.8ꢀ2.5 (br, 3 H, (H

MHz, CDCl ) 139.1 (C(12)), 136.9 (C(9)), 136.4 (C(10)), 132.5

3

C(14)), 2.06ꢀ1.98 (m, 1 H, HC(7)),

1

2

3

C(8)), 1.04

9

3

13

(

3

B) ); C NMR (126

3

(

C(11)), 88.4 (C(7b)), 82.4 (C(1)), 79.9 (C(2)), 75.1 (C(4)), 62.5

C(7a)), 60.6 (C(5a)), 35.4 (C(5)), 31.8 (C(6)), 27.0 (C(7)), 25.6

C(8)), 25.1 (C(13 or 14)), 20.4 (C(13 or 14)), 18.2 (C(15)); IR (thin

film) 3508 (s), 2383 (s), 2319 (s), 2274 (s), 1606 (m), 1378 (s), 1309

w), 1287 (w), 1187 (s), 1128 (w), 1071 (m), 1034 (m), 943 (m), 860

(w), 734 (w); MS (ESI, Q-tof) 314.2 (12), 300.2 (30), 296.2 (100),

268.2 (25); mol formula C20 32BNO (313.29); HRMS C20

(300.2327) calcd 300.2327, found 300.2332; TLC R 0.33 (hexanes/

TBME, 9:1) [I , CAM].

4

. Borane Reduction of Hydroxy Amides 19.

Preparation of (1R,3S,5S,5aS,7aS,7bR)-Octahydro-1-hydroxy-1-tert-

(

butyl-5-methyl-7b-methyl-2H-cyclopenta[gh]pyrrolizine Borane

(

3

16{4}). To a 25 mL, single-necked, round-bottomed flask equippedwith

H

H30NO

a nitrogen inlet adapter, a rubber septum, a reflux condenser, and a

magnetic stir bar were added sequentially 18{4} (29 mg, 0.115 mmol),

f

2

3

THF (200 μL), and BH THF complex (3.3 mL, 1.0 M solution,

3

2

0 equiv). The reaction flask was immersed in an oil bath and heated to

reflux (70 °C). After being stirred for 12 h at reflux, the solution was

allowed to reach room temperature and then quenched by dropwise

addition of methanol (10 mL) and concentrated by rotary evaporation

(15 mmHg, 20ꢀ25 °C). The resulting colorless oil was purified by silica

4

289

dx.doi.org/10.1021/jo2005445 |J. Org. Chem. 2011, 76, 4260–4336

Article Doi

A systematic investigation of quaternary ammonium ions as asymmetric phase-transfer catalysts. Synthesis of catalyst libraries and evaluation of catalyst activity

DOI: 10.1021/jo2005445

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jo2005445

Full text of DOI:10.1021/jo2005445

Products guided by the article

R&D Labs maybe for 1304686-93-8

Relevant to this article

Hot Product