Catalytic, Enantioselective, Intramolecular Sulfenofunctionalization of Alkenes with Phenols

Article

pubs.acs.org/joc

Catalytic, Enantioselective, Intramolecular Sulfenofunctionalization

of Alkenes with Phenols

Scott E. Denmark* and David J. P. Kornﬁlt

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

S

* Supporting Information

ABSTRACT: The catalytic, enantioselective, cyclization of phenols with electrophilic sulfenophthalimides onto isolated or

conjugated alkenes aﬀords 2,3-disubstituted benzopyrans and benzoxepins. The reaction is catalyzed by a BINAM-based

phosphoramide Lewis base catalyst which assists in the highly enantioselective formation of a thiiranium ion intermediate. The

inﬂuence of nucleophile electron density, alkene substitution pattern, tether length and Lewis base functional groups on the rate,

enantio- and site-selectivity for the cyclization is investigated. The reaction is not aﬀected by the presence of substituents on the

phenol ring. In contrast, substitutions around the alkene strongly aﬀect the reaction outcome. Sequential lengthening of the

tether results in decreased reactivity, which necessitated increased temperatures for reaction to occur. Sterically bulky aryl groups

on the sulfenyl moiety prevented erosion of enantiomeric composition at these elevated temperatures. Alcohols and carboxylic

acids preferentially captured thiiranium ions in competition with phenolic hydroxyl groups. An improved method for the selective

C(2) allylation of phenols is also described.

difunctionalized chromans that are accessible through pathway

(a).⁵Selected examples of enantioselective reactions that set

the stereocenter at C(2) through pathway (a) include: (1)

asymmetric allylic substitution, which proceeds through capture

of an allylpalladium intermediate by the phenol oxygen

(Scheme 2a);⁶(2) tandem oxidative functionalization/Heck

coupling, wherein a similar allylpalladium intermediate is

generated from an isolated alkene (Scheme 2b);⁷(3) oxidative

functionalization of a skipped diene precursor to form an

allylpalladium complex (Scheme 2c);⁸and (4) Lewis base

catalyzed cyclofunctionalization of the γ-position of an ynoate

with a phenolic hydroxyl group (Scheme 2d).⁹

Despite the impressive selectivities and obvious utility of

these methods, they are not applicable to the synthesis of

chromans bearing an additional substituent at the 3-position.

Hence, direct synthesis of enantioenriched, anti-2,3-difunction-

alized chromans is usually accomplished diastereoselectively

from an acyclic epoxide or diol precursor (Scheme 3a).¹⁰The

difunctionalization of γ-substituted 2-allylphenol to aﬀord

chromans with stereocenters at both the 2- and 3-positions

remains rare.¹¹A recent example involves the stereocontrolled

generation and intramolecular opening of a seleniranium ion by

a phenolic hydroxyl group to form a 2-seleno-3-arylchroman

(Scheme 3b).¹²The reaction proceeds via a DYKAT

mechanism, since the seleniranium ion in question is not

INTRODUCTION

■

One of the simplest benzofused heterocycles, the chroman

core, is a privileged scaﬀold for bioactive compounds, with

representatives displaying antioxidant, antitumor, antibacterial,

and other therapeutic properties (Chart 1).¹Its presence in

pharmaceutically relevant targets has led to a variety of methods

for its enantioselective construction.

The importance of the chroman motif in organic molecules is

reﬂected in the myriad protocols that have been developed for

its synthesis. Five major disconnections have been identiﬁed,

each of which leads to a diﬀerent starting material (Scheme 1).

These are (a) cyclization of a 2-allylphenol, (b) C−H

functionalization of an O-alkyl phenol, (c) intermolecular

double Michael reaction of a phenol with an unsaturated

carbonyl derivative, (d) intramolecular Friedel−Crafts-type

cyclization of an O-allylphenol, and (e) direct functionalization

of the 2-position of the parent chroman.

To date, for each of the pathways (a)−(c), enantioselective

versions have been developed, whereas pathway (d) is

diastereoselective due to the presence of a stereogenic center.

The enantioselective variant of pathway (b) has been

accomplished under Rh(II) catalysis,²pathway (c) is amenable

to enantioselective catalysis by secondary amines,³whereas

pathway (d) can be accessed either as a two-step epoxidation−

ring-opening process or in a single step using transition-metal

catalysis.⁴Each of these pathways have distinct substitution

requirements that are reﬂected in the products they aﬀord.

Notably, none of these approaches directly lead to the 2- or 2,3-

Received: February 6, 2017

A

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

Chart 1. Chroman-Containing Therapeutic Agents

Scheme 1. Retrosynthetic Disconnections for Construction of Chromans

stable under the reaction conditions. Only aryl substituents

have been incorporated at the 2-position. The product could be

functionalized further by formation of a C−C bond through the

intermediacy of a C-centered radical with poor diastereose-

lectivity. The stereocenter adjacent to the oxygen atom

remained unaﬀected.

As part of our ongoing program on enantioselective

sulfenofunctionalization of alkenes, we were interested in

extending the scope of this transformation to the synthesis of

chromans. Sulfenofunctionalization of alkenes is a longstanding

strategy for the eﬃcient introduction of sulfur moieties into

organic molecules.¹³The introduction of diverse, electrophilic

sulfenylating reagents allows for tuning of both reactivity and

selectivity in the desired application. Furthermore, the

intermediate thiiranium ion is a strong electrophile that

which undergoes facile reaction with nucleophiles resulting in

a net anti-sulfenofunctionalization of alkenes (eq 1).¹⁴

aﬀord a conﬁgurationally unstable carbocation, (2) reversible

transfer of the sulfenium ion to the nucleophile, thereby

creating an achiral sulfenylating agent, and (3) oleﬁn-to-oleﬁn

transfer of the thiiranium ion.^14,15Subsequent investigations

identiﬁed that both carbocation formation and oleﬁn-to-oleﬁn

transfer could be suppressed by lowering the reaction

temperature to −20 °C. The conﬁgurational stability of the

thiiranium ions is maintained in the presence of hard

nucleophiles, but not in the presence of weak Lewis bases,

demonstrating that Lewis bases intimately interact with

thiiranium ions. Intramolecular capture of thiiranium ions by

C-, N-, and O-nucleophiles have all been successful. The

reactions of tethered alcohols led to the formation of 2,3-

difunctionalized pyrans in high yields and with good

enantioselectivities (Scheme 4a).¹⁶Lactones were accessible

from the intramolecular cyclization of carboxylic acids.¹⁶Tosyl-

protected amines eﬃciently captured thiiranium ions to aﬀord

pyrrolidine and piperidine derivatives (Scheme 4b).¹⁷Friedel−

Crafts-type capture by electron-rich arenes allowed access to

substituted tetralins (Scheme 4c).¹⁸Finally, silyl enol ethers

produce enantioenriched α-sulfenylated ketones in high

selectivity without the intermediacy of thiiranium ions (Scheme

4d).¹⁹The catalyst exhibited complete control over the

absolute conﬁguration of the product irrespective of

nucleophile. Comparable enantioselectivities with the same

absolute conﬁguration were observed for products from all

nucleophiles tested so far. The broad nucleophile scope and

high stereochemical control observed in both the formation and

capture of thiiranium ions are hallmarks of this Lewis base

catalyzed sulfenofunctionalization reaction.

In recent years, the extension of the concept of Lewis base

activation of Lewis acids developed in these laboratories has

enabled the enantioselective sulfenofunctionalization of alkenes

and enol ethers.^13aEarly studies identiﬁed three key

racemization processes that could lead to the erosion of

enantioenriched thiiranium ions: (1) C−S bond scission to

The enantioselective oleﬁn sulfenylation reaction relies on

the principle of Lewis base activation of Lewis acids extensively

B

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

Scheme 2. Enantioselective Construction of Chromans

developed for main group chemistry in these laboratories.²⁰

Comprehensive kinetic and spectroscopic studies have led to

the formulation of a detailed catalytic cycle (Figure 1).²¹The

cycle begins with protonation of N-phenylthiophthalimide (1a)

by methanesulfonic acid acting as a cocatalyst. Displacement of

the S-aryl group by chiral Lewis base catalyst 2 leads to the

formation of active complex i. The rate of formation of i is not

turnover-limiting, and i serves as the resting state of the catalyst.

The subsequent reaction of complex i with an alkene leads to

enantioenriched thiiranium ion ii. The formation of ii is both

the turnover-limiting and the stereodetermining step. Inter- or

intramolecular nucleophilic capture of the thiiranium ion and

subsequent deprotonation aﬀords the ﬁnal product iv as well as

regenerates 2 to continue the catalytic cycle.

The intermediates involved in the catalytic cycle have been

investigated both spectroscopically and crystallographically.

Active complex i, a competent intermediate in stoichiometric

reactions, has been independently synthesized and extensively

characterized.²¹Investigation of the reaction mixture by ³¹P

NMR spectroscopy revealed that 2 is rapidly and completely

converted to i under the reaction conditions. Interestingly, the

equimolar amounts of phthalimide and methanesulfonate

formed as a byproduct of the initial reaction of 2 with 1 play

important roles in the reaction by acting as buﬀers for the

remaining methanesulfonic acid and suppressing the back-

ground reaction.²²

Initial-rate kinetic experiments determined that the reaction

is ﬁrst-order in catalyst and substrate but zero-order in

electrophile 1a. These results are consistent with turnover-

limiting thiiranium ion formation.²¹Two further pieces of

evidence corroborate the formation of ii as the slow step: (1)

complex i is the resting state of the catalyst, and (2) the rate of

the reaction is sensitive to changes in alkene electron density.

Electron-poor alkenes react slower than electron-rich alkenes.

Formation of the thiiranium ion ii is also stereodetermining.

Product enantioselectivity was insensitive to identity (C-, N-,

O-nucleophiles), electron density, or steric bulk of the

nucleophile but was very sensitive to alkene substitution

pattern. Studies investigating the conﬁgurational stability of

thiiranium ions in sulfenoetheriﬁcation reactions also conﬁrmed

that no racemization occurs at the reaction temperature.

The transition state from i to ii was investigated computa-

tionally to formulate a stereochemical model that explains the

observed enantioselectivity. Distortion/interaction analysis²³

identiﬁed the degree of distortion experienced by complex i to

achieve optimal bond overlap with the alkene substrate to be

the primary contributor to the energy diﬀerence between the

competing diastereomeric transition states.²¹Computational

C

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

Scheme 3. Synthesis of 2,3-Disubstituted Chromans

calculated; 8.9 kcal/mol, experimental) and the entropy term

(ΔS^⧧= 12 kcal/mol, calculated; 13.3 kcal/mol, experimental)

contribute substantially to the overall reaction barrier (ΔG^⧧=

23.9 kcal/mol, calculated; 22.3 kcal/mol, experimental at −20

°C). The stereochemical model identiﬁed a van der Waals

interaction between the arylsulfenyl moiety and the binaphthyl

backbone. Gratifyingly, increased steric bulk at this position in

the electrophile leads to improved selectivity overall, which

substantiated predictions made by the current model. The

absolute conﬁgurations of the sulfenofunctionalized products

wholly depend on the absolute conﬁguration of the catalyst and

are in agreement with predictions made by the calculated

model, which supported thiiranium formation as the

enantiodetermining step.

Scheme 4. Lewis Base Catalyzed Sulfenofunctionalization of

Alkenes

Nucleophilic attack at either carbon of thiiranium ion ii is

possible, raising the issue of site-selectivity.²⁴In electronically

biased thiiranium ions, capture follows Markovnikov selectiv-

ity.^16,25If the thiiranium ion is sterically and electronically

unbiased, mixtures of constitutional isomers are obtained.²⁶

Thiiranium ions whose constituent carbon atoms are sterically

diﬀerentiated also aﬀord mixtures of isomers.¹⁶Site-selectivity

is further inﬂuenced by the nature of the nucleophile, as

alcohols, carboxylic acids and protected amines display

dissimilar levels of Markovnikov selectivity. It is worth noting

that reversibility of capture has been demonstrated under the

reaction conditions for electronically activated thiiranium ions,

such as those derived from isolated trans-alkenes. Products of

thiiranium ion capture by both tethered alcohols and tethered

tosylamides are capable of undergoing exo/endo isomerization.

The product distribution of sulfenofunctionalization therefore

represents a composite of intrinsic kinetic selectivity and a

substrate-dependent thermodynamic equilibration process.

In extension of these ﬁndings, we sought to apply the

expertise gained in the study of enantioselective oxysulfenyla-

and experimental determination of the activation energy

revealed that both the enthalpy (ΔH^⧧= 11.9 kcal/mol,

D

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

Figure 1. Catalytic cycle for the enantioselective, intramolecular sulfenylation.

Table 1. Optimization of the C-Cinnamylation of Phenol

a

b

entry

solvent

base

temp (°C)

ratio 3/3′/3″

yield of 3 + 3″ (%)

1

2

3

4

5

6

7

8

Et₂O

Na

rt

4.5:0.3:1

1:2.5:trace

4:0:1

65

12

37

0

THF

Na

CH₂Cl₂

CCl₄

Na

rt

Li

rt

c

K

rt

2:0.4:1

7:trace:1

9.5:0.4:1

9:1.5:1

29

66

68

74

NaH

40

80

benzene

a

b

c

1

Determined by H NMR spectroscopic analysis of the crude reaction mixtures. Yield of isolated product. Complex mixture.

tion of alkenes to the synthesis of difunctionalized chroman

derivatives. If both the 2- and the 3-positions of the newly

formed ring system are functionalized, the ﬁnal products can be

transformed into a variety of useful compounds, thus

constituting a general and selective method for the synthesis

of chroman derivatives.

substrates, C(2) allylic phenols, presented a number of special

challenges, the solutions to which merit a brief prelude.

Although many methods exist for the site-selective O-

allylation of phenols, a corresponding general method for the

C-selective allylation of phenols has not been described.²⁷

Because phenols are ambident nucleophiles (at O, C(2), and

C(4)) and allylating agents are ambident electrophiles (leading

to products of S_N2 or S_N2′ displacement), the formation of a

single product represents challenges of site-selectivity and

chemoselectivity. Moreover, the products of C-allylation are

also competent nucleophiles that can lead to over-allylation. In

a single study on the cinnamylation of phenols using sodium as

the base, the C- vs O-selectivity was shown to be highly solvent

dependent; C-selectivity dominated in diethyl ether, whereas

O-selectivity dominated in dioxane.²⁸Unfortunately, the site-

selectivity at the cinnamyl moiety was not reported.

RESULTS

■

1. Synthesis of C(2) Allylic Phenols. A prerequisite for

this investigation was the availability of the C(2)-functionalized

phenols. A number of diﬀerent routes were employed that

varied according to the number of intervening methylene

groups and the substituents on the terminus of the alkene.

Most of those routes are derived from literature precedent and

are detailed in the Supporting Information without need for

additional comment. However, one important class of

E

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

Table 2. Allylation of Substituted Phenols with Cinnamyl Chloride

a

b

Yield of isolated, puriﬁed product. Toluene was used as the reaction solvent.

Following from this precedent, the selectivity of cinnamyla-

alkylated under the same conditions in 47 and 62% yield,

respectively (entries 2 and 3). Cinnamylation of 4-hydrox-

yanisole aﬀorded 3e in 49% yield (entry 4). Modestly electron-

poor phenols such as 2f−h were also reactive under these

conditions, albeit with extended reaction times (entries 5−7).

The highly electron poor 4-CF₃-substituted phenol 2i required

reﬂuxing toluene to eﬀect complete conversion (entry 8). 2-

Naphthol was alkylated at the 1-position selectively in 79%

yield (entry 9). The reaction was not limited to cinnamyl

chlorides: (E)-5-phenyl-1-pent-3-enyl chloride also reacted with

phenol to aﬀord 3l in 51% yield (entry 9). Finally, isoprenyl

chloride could also be used in a similar fashion to aﬀord the

desired trisubstituted alkene 3m in 65% yield (entry 11).²⁹

2. Sulfenoetheriﬁcation Reactions. In the course of

previous investigations from these laboratories, the dependence

of rate-, enantio-, and site-selectivity of sulfenoetheriﬁcation on

the substitution pattern and electron density of the alkene

component was studied.¹⁶However, changes in the reaction

outcome as a function of structural variations at the nucleophile

were not determined. Furthermore, the tether length was kept

constant across the diﬀerent substrates. The primary goals of

this study were to evaluate the inﬂuence of (1) the steric and

electronic properties of the nucleophile, in isolation and in

competition, (2) the tether length, and (3) the presence of

tion of phenol was examined with respect to the counterion and

the solvent (Table 1). Repeating the previously reported

conditions did in fact aﬀord a predominance (15:1) of C- vs O-

alkylation products 3 and 3′ in reasonable yield, but

unfortunately, the product of S_N2′ reaction at C(2) (3″) was

formed in signiﬁcant quantities as well (entry 1). Changing the

solvent to THF aﬀorded a lower yield and a predominance of

isomer 3′, whereas dichloromethane gave results similar to

those of diethyl ether (entries 2 and 3). Changing the

counterion by using lithium or potassium metal in dichloro-

methane solvent provided no improvement (entries 4 and 5).

In all of these cases, the formation of the phenolate was

observed to be slow in dichloromethane, prompting a change to

sodium hydride as the base. Under these conditions, the

phenolate formed almost instantaneously (as judged by

vigorous evolution of hydrogen). Warming the mixture to 40

°C led to a favorable increase in the formation of 3 with respect

to the other isomers as did a switch to carbon tetrachloride or

benzene and heating at 80 °C (entries 6−8).

The generality of this process was then evaluated by

cinnamylation of a number of 2- and 4-substituted phenols

under the optimized conditions. Thus, phenol itself aﬀorded 3a

in 71% yield (Table 2, entry 1). Both o- and p-cresol could be

F

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

other Lewis basic functional groups for the sulfenoetheriﬁcation

of alkenes with phenolic hydroxyl groups as nucleophiles

2.1. Optimization of the Sulfenoetheriﬁcation Reaction.

The use of a phenolic hydroxyl group as the nucleophile

aﬀorded a unique opportunity to systematically vary both the

steric and the electronic properties of the nucleophile. (E)-

Cinnamylphenol (3a)¹²was selected as a representative

substrate for initial reaction optimization. The reaction

temperature was chosen as −20 °C to avoid any potential

enantiomeric erosion of the thiiranium ion.^14bPrevious studies

showed that only 1.0 equiv of 1a was necessary for the

reaction.¹⁶The initial rate experiments revealed that excess acid

was detrimental to the reaction rate; thus, 0.75 equiv was

chosen as the starting point for optimization. Solvent and

concentration were unchanged from the previous conditions for

sulfenoetheriﬁcation reactions. For catalyst 2, the exocyclic

amine substituent strongly aﬀects the enantioselectivity of the

sulfenofunctionalization process. Thus, catalyst 2a (used in the

ﬁrst study) aﬀorded the desired product with poor selectivity

(Table 3, entry 1). Catalyst 2b formed product 4a in 46% yield

and with an er of 95.1:4.9 (entry 2). Changing to catalyst 2c

resulted in slightly lower enantioselectivity of 93.1:6.9 er (entry

3). Reducing the concentration of the reaction using catalyst 2b

allowed the product to be obtained in 95% yield (entry 4).

2.2. Sulfenocyclizations of Substituted (E)-2-Cinnamyl-

phenols: Inﬂuence of Phenol Ring Substituents. The

nucleophilicity of phenolic hydroxyl groups is inﬂuenced by

both steric and electronic properties of ring substituents.³⁰

Accordingly, a series of substituted (E)-2-cinnamylphenols was

prepared and evaluated (Table 4). The parent substrate, 3a,

which aﬀorded 4a in 84% yield, 94.9:5.1 er, and >30:1 endo-

selectivity was used as a benchmark (Table 4, entry 1). Methyl

substituents at the 4- and 6-positions on the ring did not alter

the reaction outcome meaningfully (entries 2 and 3). The

presence of an extended π-system resulted in no change in yield

and a very small decrease in er (entry 4). In all cases, high

(>30:1) constitutional selectivity was observed.

Next, the inﬂuence of heteroatom substituents on reaction

outcome was evaluated. Electron-donating groups had little

inﬂuence; for example, 3e bearing a 4-methoxy group cyclized

to form 4e in 84% yield and 94.4:5.6 selectivity (entry 5).

Bromophenol 3f cyclized to aﬀord 4f in 81% yield and 93.8:6.2

er (entry 6), albeit with slightly extended reaction time. The

more electronegative chloro- and ﬂuoro-substituted substrates

reacted in the same time frame, with 4g and 4h being produced

in 70 and 82% yields, respectively (entries 7 and 8). The

enantioselectivity of the reactions remained high with these

substrates.

Phenol 3i, bearing a highly electron-withdrawing 4-CF₃

group, was insuﬃciently reactive at −20 °C. Increasing the

reaction temperature to 22 °C caused signiﬁcant erosion of

enantioselectivity (er 70.7:29.3, SI). Sterically bulky electrophile

N-2,6-diisopropylphenylthiophthalimide (1c) prevents erosion

at elevated reaction temperatures by shielding the intermediate

thiiranium ion.²¹Cyclization of 3i at room temperature using

1c proceeded smoothly, and 4i was isolated with 89% yield and

95.2:4.8 er (entry 9).

a

Table 3. Optimization of the Sulfenocyclization Reaction

2.3. Sulfenocyclization of 2-Substituted Phenols: Inﬂuence

of Alkene Substituent and Tether Length. Changes in the

alkene substitution pattern have been documented to

dramatically alter selectivity for the sulfenofunctionalization

process.¹⁶(E)-Disubstituted alkenes are the most selective

substrates for sulfenoetheriﬁcation, whereas terminal alkenes

are only slightly less so. Trisubstituted, (Z)-, and 1,1-

disubstituted alkenes reacted with poor selectivity. Thus,

primarily (E)- and terminal alkenes were employed in this

study. Initially, the (E)-phenyl substituent was varied. 2-Furyl-

substituted benzopyran 4j was produced in 88% yield and

92.5:7.5 er, whereas 2-thienyl-substituted substrate 3k

produced the desired product in 86% yield and 93.9:6.1 er

(Table 5, entries 1 and 2). Changing the alkene substituent to

an aliphatic group as in substrate 3l led to a 74% combined

yield of a 1.5:1.0 mixture of isomers 4l and 5l with 96.6:3.4 er

for 4l (entry 3). Trisubstituted alkenes generally led to less

selective cyclizations with 1a.¹⁶To increase selectivity,

electrophile 1c, which has an improved selectivity proﬁle, was

tested with substrate 3m.²¹In this case, use of 1c led to the

formation of gem-disubstituted benzopyran 4m in 93% yield

and 95.4:4.6 er (entry 4).

b

c

entry

[3] (M)

acid (equiv)

catalyst

yield (%)

er

d

1

2

3

4

5

6

7

8

0.4

0.75

0.5

2a

2b

2c

2b

35

65.6:34.4

95.1:4.9

93.1:6.9

g

e

0.4

46

f

0.4

27

0.15

95

93

95.3:4.7

95.7:4.3

94.9:5.1

94.3:5.7

0.25

0.1

h

32

i

0.5

96

a

b

Reactions run on 0.1 mmol scale. Yield of isolated, puriﬁed product.

Determined by CSP-SFC. Isolated product contaminated with 2a.

Isolated product contaminated with 2b. Isolated product contami-

nated with 2c. Not determined. Incomplete conversion was

c

d

e

f

g

h

i

observed. EtSO₃H was used.

In all preceding sulfenofunctionalization studies, the loading

of the Brønsted acid greatly aﬀected the rate and selectivity of

the reaction.²¹In the sulfenoetheriﬁcation with alcohols,

maximum reactivity was reached at 0.6 equiv of methanesul-

fonic acid. Cyclization of substrate 3a using 0.75 equiv of acid

was complete within 24 h. Decreasing the amount of acid to 0.5

equiv or even 0.25 equiv did not decrease the yield in the same

time frame (entries 5 and 6). In all cases, high selectivity

(>30:1) for endo capture to form the chroman was observed.

Further decrease in the acid loading resulted in incomplete

conversion and reduced yield (entry 7). The use of

ethanesulfonic acid did not aﬀect the enantioselectivity and

was not pursued further (entry 8).

The site-selectivity of thiiranium capture during intra-

molecular sulfenocyclization is strongly inﬂuenced by the

relative rates of formation of diﬀerent size rings. The

preparation of substrates with varying tether lengths enabled

a systematic study of ring size eﬀects (Table 5). Substrate 3n,

which contains an (E)-2-styryl group at the end of a two-carbon

tether, aﬀorded the 7-endo cyclization product benzoxepane 4n

in 92% yield and 94.4:5.6 er (entry 5). Further extending the

G

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

a

Table 4. Sulfenocyclizations of Substituted (E)-2-Cinnamylphenols

a

b

c

d

Reactions run on 1.0 mmol scale. Yield of isolated, puriﬁed product. Determined by CSP-SFC. Electrophile 1c was used, Ar^iPr= 2,6-(i-Pr)-C₆H₃.

e

f

Reaction run at 22 °C. Determined after oxidation to the sulfone.

tether to three methylene groups in substrate 3o proved

problematic: under the optimized conditions using electrophile

1a, no desired product was observed. Gratifyingly, use of

electrophile 1c led to the surprising formation of exo cyclization

product 5o in 76% yield and 92.6:7.4 er (entry 6). Next, the

electronic bias imparted by the phenyl group was removed to

evaluate the site-selectivity of closure with a terminal alkene in

the reaction with electrophile 1c. Substrate 3p reacted via a 6-

exo-mode cyclization to aﬀord 5p with high site-selectivity, 91%

yield, and 97.2:2.8 er (entry 7). Extension of the tether length

by one more methylene group, as in substrate 3q, was

gratifyingly successful, as the cyclization proceeded in a 7-exo

mode to aﬀord product 5q, with 84% yield and 97.7:2.3 er

(entry 8).

The Lewis basic nature of the selenophosphoramide moiety

prompted an investigation into the compatibility of the reaction

conditions with other Lewis basic functional groups in the

substrate. Substrate 3r, containing a carboxylic ester group

three carbons removed from the reacting oleﬁn, aﬀorded 5r in

good yield albeit with somewhat diminished enantioselectivity

compared to 3l (cf. entry 9 and Table 5, entry 3). Replacement

of the ester by the corresponding ether in 3s restored the

enantioselectivity in comparable yield (entry 10). The relative

reactivity of other oxygen nucleophiles with respect to phenolic

hydroxyl groups was also tested. Both carboxylic acids and

alcohols outcompeted phenols for thiiranium capture. Substrate

3t, bearing a carboxylic acid moiety, preferentially aﬀorded

lactone 8t in 92% yield and 92.2:7.8 er (entry 11). In the

presence of a remote hydroxyl group, such as in substrate 3u,

H

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

a

Table 5. Sulfenocyclizations of 2-Substituted Phenols

a

b

c

d

Reactions run on 1.0 mmol scale. Yield of isolated, puriﬁed product. Ar^iPr= 2,6-(i-Pr)-C₆H₃. Determined by ¹H NMR analysis of crude reaction

e

f

g

h

mixtures. Determined by CSP-SFC. 0.25 equiv MsOH was used. Determined after oxidation to the sulfone. endo/exo ratio for alcohol capture.

I

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

Scheme 5. Manipulations of 3-Phenylthiochroman

saturated oxacycles 8u and 9u were formed as a 1.1:1 mixture

of constitutional isomers in 88% combined yield and with

96.9:3.1 er and 97.1:2.9 er respectively (entry 12).

2c established that up to 4 equiv of acid is needed before full

conversion to the sulfenylated species is observed.²²In the case

of cyclizations of phenolic hydroxyl groups, high conversion

could be achieved with as little as 2.5 equiv of MsOH with

respect to catalyst. In contrast to alcohols, the phenolic

hydroxyl group does not appear to act as a proton buﬀer.

The cyclization of 3a was complete within 24 h at −20 °C,

compared to 93% after 24 h at the same temperature for the

corresponding alcohol substrate.²¹Thus, the rate of phenol

cyclization is comparable to that of the alcohols. In the prior set

of experiments, a large excess (10 equiv with respect to

catalyst) of MsOH was employed, which led to rates slower

than υ_max. The decrease in rate at high MsOH concentrations

was ascribed to protonation of the substrate by excess acid.²¹In

the case of phenols, no substantial changes in rate were

observed as a result of increased acid concentration in the range

of 2.5−7.5 equiv of MsOH with respect to catalyst. The

substantially lower Brønsted basicity of phenols (pK_aof

2.4. Transformations of Sulfenofunctionalization Prod-

ucts. The sulfenyl moiety in the product thioether can act as a

locus for further transformations. Thus, a series of manipu-

lations to explore the reactivity of the thioether group were

carried out (Scheme 5). Reductive cleavage of 4a with nickel

boride formed 2-phenylchroman in 71% yield. Oxidation of 4a

with sodium metaperiodate led to sulfoxide 6 in 85% yield and

in a 2:1 diastereomer ratio. Both diastereomers underwent

thermal elimination in toluene, leading to the formation of 2-

phenylchromene in 92% yield. Attempts to engage the sulfoxide

in a Pummerer rearrangement were not successful. Instead,

vinyl thioether 7, which results from loss of a proton from the

thiocarbenium intermediate, was formed. In the presence of

triﬂuoroacetic anhydride and pyridine, 7 was obtained in 94%

yield. Unfortunately, 7 proved resistant to Ni-catalyzed cross

coupling with Grignard reagents.

31

+

PhOH₂, −6.5; pK_aEtOH₂, −2.2) implies that a much

smaller fraction of substrate is protonated even in the presence

of an excess of acid. Thus, similar reaction rates are observed

over a much broader range of acid stoichiometry.

DISCUSSION

■

The systematic variation of reaction conditions, phenol

substituents, tether length, and functional groups enabled a

more thorough understanding of the factors governing the

sulfenofunctionalization process. In this section, the inﬂuences

of each of these components on observable reaction outcomes

such as rate, enantioselectivity, and site-selectivity will be

discussed in turn.

2. Structural Eﬀects on Rate and Selectivity. 2.1

. Inﬂuence of the Nucleophile. The rate, enantio-, and site-

selectivity of sulfenofunctionalization of any alkene with a

pendant nucleophile is dependent on a multitude of structural

factors. Preceding studies showed a substantial impact of alkene

environment on all three of these observables.^13aHowever, no

such systematic investigation for the nucleophile was under-

taken. In an isolated example, the cyclization of a tertiary

alcohol was comparable with that of a primary alcohol (eq 2).¹⁶

A previous study regarding the rate of sulfenocyclization of a

number of protected amines did not identify speciﬁc reactivity

trends.¹⁷Thus, the cyclization of (E)-2-cinnamylphenols

provided an opportunity to understand how the aforemen-

tioned observables are impacted by the steric and electronic

properties of the nucleophile.

1. Optimization of the Sulfenofunctionalization

Reaction. 1.1. Catalyst. The optimization of the sulfenofunc-

tionalization reaction focused on two main variables, the

catalyst and the loading of the Brønsted acid. Exhaustive

catalyst optimization for the sulfenofunctionalization process

has been detailed elsewhere and was not repeated for this

reaction.¹⁶Of the three best catalysts for sulfenofunctionaliza-

tion, optimum selectivity was achieved with 2b for reasons

previously noted.²¹Conversion across all catalysts was

comparable.

1.2. Brønsted Acid. Previous kinetic studies established that

for an alcohol nucleophile the υ_maxwas reached with 0.6 equiv

of methanesulfonic acid (MsOH).²¹Further increases in the

Brønsted acid loading led to decreases in reaction rate. Two

major factors need to be considered for optimal MsOH

loading: (1) the amount of acid necessary to fully transform the

catalyst into active complex i and (2) the eﬀective acidity of the

reaction mixture as a result of solvation eﬀects in nonpolar

reaction media. In the proposed mechanistic cycle, the acid is

present only in cocatalytic amounts; therefore, only 1.0 equiv of

acid with respect to catalyst should be necessary. However,

titration studies with ethanesulfonic acid together with 1a and

2.1.1. Reaction Rate. The turnover-limiting step of the

sulfenofunctionalization reaction for alcohols is thiiranium ion

formation; hence, for these substrates thiiranium ion capture is

fast.²¹As mentioned previously, phenols are substantially

weaker nucleophiles than alcohols.³¹However, the rates of

J

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

cyclization of 3a, 3b, 3c, and 3e were comparable. Thus, the

turnover-limiting step does not change for electron-neutral or

-rich phenol nucleophiles. Naphthols are slightly stronger

nucleophiles than phenols, and in agreement with the

aforementioned turnover-limiting thiiranium ion formation,

the rate of cyclization of 3d was not aﬀected.³¹

reaction be run at 23 °C, leading to decreased product

enantioselectivity with electrophile 1a (eq 4).

Introduction of weakly electron-withdrawing substituents led

to slightly slower overall rates. The electron-withdrawing

property of the substituents on phenols is a composite eﬀect

of the inductive (σ_I) and resonance (σ_R) contributions of the

substituent to the electron density on the phenol.³²Chlorine

and bromine both strongly withdraw electron density

inductively (σ_IBr, 0.49; Cl, 0.43) but also donate electron

density through π-resonance (σ_RBr, −0.16; Cl, −0.16).³²

Fluorine is a strongly withdrawing substituent inductively, but

also a much better π-donor (σ_IF, 0.57; σ_RF, −0.33) such that

its overall eﬀect is comparable. In contrast, a triﬂuoromethyl

group is both inductively and mesomerically electron with-

drawing (σ_ICF₃, 0.46; σ_RCF₃, 0.09). The reaction of Br-, Cl-,

and F-substituted phenols (3f, 3g, and 3h, respectively) were all

only slightly slower at −20 °C compared to parent substrate 3a.

The reaction does not appear to be sensitive to either of these

parameters. However, for CF₃-bearing substrate 3i the reaction

needed to be performed at room temperature.

The substantial diﬀerence in rate between 3a and 3i (cf.

Table 4, entries 1 and 9) suggests that the turnover-limiting

step has changed from formation of the thiiranium ion to

capture for electronically deactivated nucleophiles. Although

the lifetime of the thiiranium ion intermediate derived from 3i

likely increases as a result of slow capture, the high chemical

yield implies that the thiiranium ion is stable when 1c is used as

the electrophile.

To attenuate the racemization, electrophile 1c, bearing a

bulky 2,6-diisopropylphenyl group, was used. The 2,6-

substituents impart both slightly higher intrinsic selectivity to

1c as well as increased stability to the resulting thiiranium ions

as was seen with alcohol 10 (eq 5).²¹The use of 1c in the

reaction of 3i resulted in a dramatic increase in selectivity (eq

4). Thus, conﬁgurational erosions of thiiranium ions can be

ameliorated through increased steric shielding of the sulfur

atom.

2.1.2. Enantioselectivity. Because thiiranium ions are

conﬁgurationally stable in dichloromethane at −20 °C,^14b

their enantiomeric composition should be retained throughout

the remaining reaction steps.²¹Notably, capture of thiiranium

ions derived from styrenes by C-, N-, and O-nucleophiles

results in the same absolute conﬁguration and comparable

levels of product enantioenrichment (eq 3).^16,17

2.1.3. Site-Selectivity. The site-selectivity for the sulfenoe-

theriﬁcation follows the established Markovnikov rule.²⁵

Reaction of (E)-2-cinnamylphenols can proceed through either

a 5-exo cyclization to aﬀord a benzofuran or a 6-endo cyclization

to aﬀord a chroman. Only chromans were observed as reaction

products in the cyclization of (E)-2-cinnamylphenols. Changes

in the electron density or steric bulk of the nucleophile had no

eﬀect on site-selectivity. Opening of the thiiranium ion through

a Friedel−Crafts-type process (i.e., C-aryl cyclization) was not

observed, demonstrating that the chemoselectivity of capture is

high, irrespective of arene electron density.^4c,18

2.2. Inﬂuence of Alkene Environment. The alkene environ-

ment represents the most important variable that can inﬂuence

both the rate and the selectivity of the sulfenocyclization

process.²¹This sensitivity has been evident in numerous

sulfenofunctionalizations.¹⁶⁻¹⁹In general, higher alkene elec-

tron density leads to increased reactivity. Enantioselectivity is

most sensitive to the alkene substitution pattern, whereas site-

selectivity is governed by the aforementioned Markovnikov

selectivity, albeit complicated by the potential for isomerization

of certain sets of constitutional isomers under the reaction

conditions.

2.2.1. Rate. The rate of cyclization was expected to follow a

well-deﬁned trend of alkene electron density. The reaction

times for substrates with disubstituted alkenes demonstrated

that the reactivity diﬀerence between heteroaryl, aryl and alkyl

substituents was not substantial for disubstituted alkenes

For the participation of phenolic hydroxyl groups, the

enantioselectivity of sulfenocyclization remained uniformly high

for electron-rich phenols. However, the change in mechanism

from turnover-limiting formation to turnover-limiting capture

raises the possibility of erosion of enantioselectivity as a

consequence of increased thiiranium ion lifetime and attendant

racemization. No decrease in enantioselectivity is observed for

substrates bearing halogen substituents, conﬁrming the overall

stereochemical stability of the thiiranium ion for slightly

extended lifetimes at −20 °C For the triﬂuoromethyl-

substituted phenol, the lack of reactivity required that the

K

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

(3a,k,l,s; 24 h at −20 °C). However, because both mono- and

trisubstituted alkenes required electrophile 1c for high

selectivity, direct comparison of the rate of cyclization for

terminal alkene 3p and trisubstituted alkene 3m with their

respective analogs 3s and 3l is not possible.

The strongly acidic nature of the reaction conditions raised

the possibility of cationic alkene cyclization or polymerization

as a competitive side reaction, especially for substrates with

electron-rich alkenes.³³For example, in the cyclization of a 4-

tolyl-substituted alkene in the Friedel−Crafts alkylation

process, a proton-initiated cyclization was observed (eq 6).²²

No evidence of acid-catalyzed polymerization or cyclization was

observed for the current set of styrenes.

lectivities have been observed previously for sulfenofunction-

alization with this class of substrates (eq 7).¹⁶Therefore, more

selective electrophile 1c was employed for the cyclization of

terminal alkenes 3p and 3q. Thus, the use of 1c in the

cyclization of 3p aﬀorded an enantioselectivity of 97.2:2.8. (cf.

Table 5, entries 5 and 7). Similarly, cyclization of 3q proceeded

with high enantioselectivity even at elevated temperature

(97.7:2.3).

Trisubstituted alkenes are challenging substrates for selective

sulfenofunctionalization. In the cyclization of pendant alcohols,

both (E)- and (Z)-trisubstituted alkenes aﬀord the correspond-

ing products with poor enantioselectivities (60:40 and 70:30

respectively, eq 8).¹⁶The higher intrinsic selectivity of

sulfenylating agent 1c was beneﬁcial to this class of substrates

as well, as benzopyran 4m was produced with 95.4:4.6 er.

Conjugated electron-rich heteroarenes are signiﬁcantly more

susceptible to polymerization. Thus, when MsOH was added to

a mixture of 2-furylalkene 3j, 2b, and 1a in dichloromethane at

−20 °C, rapid polymerization was observed and no identiﬁable

product was obtained. If the order of addition was changed to

introduce 3j last, the desired reaction pathway was restored.

The eﬀective acid concentration in solution clearly has a

substantial impact on the rate of polymerization for sensitive

alkenes.²²

2.2.2. Enantioselectivity. The enantiomeric composition of

the ﬁnal products in sulfenofunctionalizations is determined by

the enantioenrichment of their precursor thiiranium ions.^16,17

The enantioselectivity of thiiranium ion formation is, in turn,

determined by the transition-state complex consisting of alkene

and intermediate i. Computation of the energies of the

diastereomeric transition states revealed that catalyst distortion

is the most important contributor to the diﬀerence in

transition-state free energies.²¹Thus, changes in alkene

substitution pattern and consequently the degree of distortion

that the catalyst experiences to accommodate the substituents

are expected to substantially impact the stereoselectivity of the

process. Indeed, foregoing studies have demonstrated that the

enantiotopic faces of (Z)- and 1,1-disubstituted alkenes are

poorly diﬀerentiated by the active sulfenylating agent derived

from 2b.^16,21

The enantioselectivity of the sulfenoetheriﬁcations herein

was consistent among aryl-, heteroaryl-, and alkyl-substituted

(E)-alkenes (4a, 94.9:5.1; 4k, 93.9:6.1; 4l, 96.6:3 4k). The

nature of the oxygen nucleophile played only a minor role as

phenols, alcohols, and carboxylic acids cyclized with com-

parable enantioselectivities (4a, 94.9:5.1; 8u, 96.9:3.1; 8t,

92.2:7.8). The high enantioselectivities observed independent

of nucleophile parameters are in good agreement with the

current hypothesis of stereodetermining thiiranium formation.

Terminal alkenes are usually diﬃcult substrates for

enantioselective alkene functionalizations due to the absence

of steric diﬀerentiation at the terminus.³⁴Lower enantiose-

2.2.3. Site-Selectivity. The Markovnikov rule for site-

selectivity holds well for the cyclization of alkenes wherein

the nucleophile is three atoms removed (cf. 2.1.3). The

investigation of constitutional site-selectivity of the reaction

alcohols to thiiranium ions derived from biased alkenes

demonstrated that cyclization preferentially occurs at the

stabilized position. For unbiased alkenes, a mixture of isomers

is obtained, although an in situ isomerization process from the

5-exo isomer to the 6-endo isomer precluded analysis of

kinetically controlled selectivity.¹⁶In contrast, high 5-exo

selectivity is observed for carboxylic acid cyclizations. The

cyclization of phenolic hydroxyl groups proceeds similarly to

the alcohols, with high Markovnikov site-selectivity. The

inﬂuence of resonance stabilization (presence of a Ph

substituent, cf. Table 5, entry 5) or inductive stabilization

(disubstituted carbon atom of the thiiranium ion, Table 5, entry

4) is suﬃcient for high selectivity. Moreover, a mixture of

isomers is observed in the absence of electronic bias (Table 5,

entry 3). The presence of a phenolic hydroxyl group did not

otherwise impact selectivity, as the cyclization of 3u proceeded

to give a mixture of isomers, whereas carboxylic acid 3t

displayed very high 5-exo selectivity, conﬁrming previous trends

(Table 5, entries 11 and 12).¹⁶

For capture of an electronically unbiased thiiranium ion by a

phenolic hydroxyl group three atoms away, the Markovnikov

rule predicts poor selectivity. Instead, the controlling factor is

L

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

the relative rate of 5-exo to 6-endo cyclization. Indeed, in the

case of substrate 3l, a 1.5:1.0 ratio of 4l/5l from endo vs exo

cyclization, respectively, was obtained (Table 5, entry 3). Direct

comparison to the corresponding aliphatic alcohol, which

aﬀorded a 5:1 ratio of endo:exo isomers, is not warranted due to

the presence of C_sp2atoms in the tether.¹⁶Instead, comparisons

can be drawn to intramolecular phenolic opening of other

tethered three-membered electrophiles. The opening of a

disubstituted alkyl epoxide by a phenolic hydroxyl group

proceeds exclusively exo, whereas the iodine-mediated cycliza-

tion of 2-crotylphenol,^35awhich proceeds through an iodonium

intermediate, aﬀords solely the endo product^35b(eqs 9 and 10).

Scheme 6. Inﬂuence of Tether Length and Substituents on

Constitutional Site Selectivity

The intrinsic selectivity of phenol capture is therefore highly

dependent on the nature of the electrophilic moiety, wherein

increased charge and larger atoms in the three-membered ring

result in higher formation of the endo product. Thiiranium ions

are between these two extremes, and consequently, low

selectivity is observed.

The ratio 4l/5l is independent of conversion. In this

particular case, the reduced basicity of a benzopyran oxygen

compared to a pyran as well as the reduced MsOH loading

likely retard isomerization, leading to poor site-selectivity that

represents kinetic control.

2.3. Inﬂuence of Tether Length. On the basis of the

observed characteristics of the sulfenofunctionalization reac-

tion, changes in the tether length were primarily expected to

aﬀect the rate- and site-selectivity of the process. As the alkene

environment is constant among the various substrates, the

enantioselectivity was not expected to vary. Indeed, comparable

substrates displayed similar enantioselectivities irrespective of

tether length (3a and 3n, 3p and 3q).

2.3.1. Rate. The free energy barrier to cyclization for

medium-sized rings is highly dependent on ring size.³⁶

Consequently, the relative rates of endo and exo capture of

the thiiranium ions are dependent on the sizes of the rings

being formed. The rate of cyclization of substrate 3n, bearing

one more atom in the tether than 3a, is similar to the rate of 3a

(cf. Table 4, entry 1, and Table 5, entry 5, Scheme 6a).

Similarly, the rates of cyclization of dialkyl-substituted

substrates 3l and 3s are comparable (Table 5, entries 3 and

10, Scheme 6b,c). Overall, for 5-exo, 6-endo, 6-exo, and 7-endo

capture, thiiranium ion formation appears to be turnover-

limiting, and no eﬀect on rate as a function of tether length is

observed. In contrast, substantially diﬀerent rates are observed

for the cyclization of 3p and 3q (12 h vs 48 h) which have,

respectively, two and three methylene units in their tethers

(Table 5, entries 7 and 8, Scheme 6d). The cyclization of 3o,

also bearing a three-carbon tether, with 1c, required elevated

temperature compared to the shorter 3n (22 °C vs −20 °C,

Table 5, entries 5 and 6). Attempts to cyclize 3o with 1a for

comparison purposes failed, with primarily decomposition

products being formed (Scheme 6a). Thus, both 7-exo and 8-

endo cyclizations are generally disfavored in comparison to the

aforementioned modes. The overall relative rates as a function

of cyclization mode can then be expressed as 8-endo ≪ 7-exo <

7-endo ∼ 6-exo ∼ 6-endo ∼ 5-exo. In conjunction with these

rates, a structural feature for change of the rate-determining

step can be established. For 7-endo and more facile closures,

thiiranium formation remains turnover-limiting; however, for

less favored closures such as 7-exo and 8-endo, capture becomes

turnover-limiting. The importance of sterically shielding the

thiiranium ion reactions with turnover-limiting capture is clearly

illustrated by the failure of 1a to promote the cyclization of 3o.

2.3.2. Site-Selectivity. Changing the tether length introduces

substantial bias into the cyclization due to the higher energy

associated with forming 7-membered and larger rings.^36,37The

cyclization of electronically unbiased alkenes with two-carbon

tethers (3s and 3r) shows that the intrinsic selectivity for 6-exo

over 7-endo highly favors the 6-exo product. However, 3n,

which bears a phenyl substituent, cyclized selectively to the

benzoxepane 4n. Thus, a kinetic preference exists for

cyclization following the Markovnikov rule. Surprisingly, the

cyclization of 3o did not follow the expected trend. Under

standard reaction conditions, decomposition was observed. The

increased kinetic barrier to either 8-endo or 7-exo cyclization

extends the lifetime of the thiiranium ion such that

decomposition processes intervene. The use of the electrophile

1c led to preferential 7-exo cyclization. In this case, the

increased stability imparted to the thiiranium ion is suﬃcient to

prevent decomposition, whereupon successful capture can take

M

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

place. The observed site-selectivity is dominated by the kinetic

preference for formation of a seven- vs eight-membered ring.

The subtle interplay between enthalpic and entropic con-

tributions is similar to that exhibited with epoxide substrates

wherein the cyclization of an electronically unbiased epoxide

aﬀords the 5-exo isomer vs the 6-endo isomer with high

selectivity whereas an electronically biased epoxide opens with

opposite selectivity (eq 11).³⁸

Interestingly, the chemoselectivity does not correlate with

proton aﬃnity, as the pK_aof a protonated carboxylic acid is

somewhat higher than that for a protonated phenol (pK_a,aq

:

PhCO₂H₂⁺−7.8 vs PhOH₂⁺−6.5).³¹The discrepancy suggests

that the kinetic preference for formation of a ﬁve-membered

ring via 5-exo closure dominates the reaction of 3t.¹⁶In

contrast, alcohol 3u formed 8u and 9u in almost equimolar

ratio, compared to previous results where a 5:1 isomer ratio

favoring the pyran was observed (Schemes 7a,b).¹⁶There

appears to be no intrinsic kinetic preference for 5-exo vs 6-endo

cyclization for alcohols with three carbon tethers, though the

product ratio for the aliphatic alcohol may not represent kinetic

control.

CONCLUSIONS

■

A highly enantioselective sulfenocyclization of alkenes with

tethered phenolic hydroxyl groups to aﬀord substituted

chromans has been developed. A systematic variation of the

nucleophile component and tether length allowed further

trends in sulfenofunctionalization to be identiﬁed. The reaction

was insensitive to changes in the steric properties of the

nucleophile. The nucleophile electron density only made a

diﬀerence for highly electron-deﬁcient phenols. The reaction

rate did not change for one- and two-methylene tethers, and

both benzopyrans and benzoxepanes were readily prepared,

whereas further increases in tether length resulted in slower

reactions. Enantioselectivity was unaﬀected by changes in the

nucleophile component or tether length. Nucleophilic capture

occurred at the more electronically biased location of the

thiiranium ion. In the absence of electronic bias, the intrinsic

site-selectivity of sulfenofunctionalization was low for one-

methylene tethers but high for two-methylene tethers.

Carboxylic acids and alcohols were more reactive toward

thiiranium ions than phenolic hydroxyl groups, and high

chemoselectivity was observed in competition experiments.

Substrates which displayed low reactivity at −20 °C were

amenable to sulfenofunctionalization with hindered electrophile

1c at higher temperatures. The increased thiiranium ion

stability as a result of shielding prevented erosion of

enantioselectivity that had previously plagued reactions at

such temperatures. The eﬀects and inﬂuences identiﬁed here

are predicted to be general for capture of thiiranium ions by

other nucleophiles. Expansion of both scope and selectivity to

other nucleophiles will be reported in due course.

2.4. Inﬂuence of Lewis Basic Functional Groups. The

sulfenofunctionalization process relies on a moderately Lewis

basic selenophosphoramide to promote the catalytic process.

Therefore, the sensitivity of the reaction to other Lewis basic

functional groups that may be present in the reaction was of

interest. In comparison with the reaction of 3a, no eﬀect on rate

was observed for any of the four Lewis basic groups tested

(alcohol, acid, ester, ether), although in these cases 0.5 equiv of

acid was used to counteract any potential buﬀering eﬀects.

The enantioselectivity of the reactions was consistently high

except for 4r and 8t, which showed slight erosion. The

interaction of the carbonyl group either directly with the

thiiranium ion during its formation or with the complex during

its capture may lead to this erosion. Given the small magnitude

of the change, the interaction appears to be weak. The site-

selectivity for ester 4r was slightly lower than for 3s, which also

suggests an interaction not present with the ether functional

group.

If two competent nucleophiles are present in the molecule,

such as in 3t and 3u, capture with either nucleophile is possible.

The chemoselectivity of the reaction is then dependent on, in

addition to the ring size as discussed, the relative rates of

cyclization for either nucleophile. Both a carboxylic acid and an

alcohol outcompeted the phenolic hydroxyl group for

thiiranium capture.³⁹Clearly, the lesser nucleophilicity of the

phenolic hydroxyl group with respect to other oxygen

nucleophiles is suﬃcient to disfavor aryl ether formation.³⁰

Scheme 7. Constitutional Site Selectivity for 4-Pentenols

N

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

tyl-1-amino)-3,5-dimethyl-3,5-dihydrodinaphtho[2,1-d:1′,2′-f ][1,3,2]-

diazaphosphepine 4-selenide (2c).¹⁸Methanesulfonic acid and

ethanesulfonic acid were dried as described in the literature.²²

Substrates 2-((E)-3-thiophene-ylprop-2-en-1-yl)phenol (3k),⁴⁰2-

(but-3-en-1-yl)phenol (3p)⁴¹and 2-(pent-4-en-1-yl)phenol (3q)⁴²

were prepared as described. Intermediates (E)-3-(furan-2-yl)-1-(2-

hydroxyphenyl)prop-2-en-1-one (S2)⁴³and diethyl (E)-2-(5-(2-

hydroxyphenyl)pent-2-en-1-yl)malonate (S3)⁴⁴were prepared as

described.

EXPERIMENTAL SECTION

■

General Experimental Methods. All reactions were performed in

oven- (160 °C) and/or ﬂame-dried glassware under an atmosphere of

dry argon unless otherwise noted. All reaction temperatures

correspond to internal temperatures measured with Teﬂon-coated

thermocouples. A ThermoNesLab CC-100 or a ThermoNesLab IBC-

4A Cryocool with an attached Cryotrol was used for reactions at

subambient temperatures.

¹H and ¹³C NMR spectra were recorded on Varian Unity (400

Preparation of 2-Substituted Phenols with One Methylene

Tether.

1

MHz, H; 101 MHz, ¹³C) or Inova (500 MHz, H; 126 MHz, ¹³C)

spectrometers. ³¹P and ¹⁹F NMR spectra were recorded on Inova (202

MHz) and Inova (470 MHz) spectrometers, respectively. Acquisition

times were 4.096 s for ¹H NMR, 1.024 s for ¹³C NMR, 0.655 s for ³¹

P

NMR, and 0.328 s for ¹⁹F NMR. Spectra are referenced to residual

1

chloroform (δ = 7.26 ppm, H; 77.0 ppm, ¹³C). Chemical shifts are

reported in parts per million; multiplicities are indicated by s (singlet),

d (doublet), t (triplet), q (quartet), p (pentet), h (hextet), sept

(septet), m (multiplet), and br (broad). Coupling constants, J, are

reported in hertz, integration is provided, and assignments are

indicated. Assignments were conﬁrmed through 2-D COSY and

HMQC experiments. Elemental analysis for was performed by the

University of Illinois Microanalysis Laboratory or Robertson Microlit

Laboratories. Mass spectrometry (MS) was performed by the

University of Illinois Mass Spectrometry Laboratory. Electron Impact

(EI) spectra were performed at 70 eV using methane as the carrier gas

on a Finnegan-MAT C5 spectrometer. Chemical Ionization (CI)

spectra were performed with methane reagent gas on a Micromass 70-

VSE spectrometer. Electrospray Ionization (ESI) spectra were

performed on a Micromass Q-Tof Ultima spectrometer. Data are

reported in the form of m/z (intensity relative to the base peak = 100).

Infrared spectra (IR) were recorded on a PerkinElmer FT-IR system,

and peaks are reported in cm⁻¹with indicated relative intensities: s

(strong, 0−33% T); m (medium, 34−66% T), w (weak, 67−100% %),

and br (broad). Melting points (mp) were determined on a Thomas-

Hoover capillary melting point apparatus in sealed tubes under

vacuum and are corrected.

Analytical thin-layer chromatography (TLC) was performed on

Merck silica gel 60 F254 or Merck silica gel 60 RP-18 F254s plates.

Silver-impregnated silica was prepared by ﬁrst dissolving 5 g of AgNO₃

in 30 mL of deionized water. The resulting homogeneous solution was

added to 50 g of SiO₂in a mortar. The clumpy silica paste was ground

with a pestle until free-ﬂowing. The mortar was then covered in

aluminum foil and placed in an oven (160 °C) for 3 h. The vessel was

then removed from the oven and allowed to cool in a desiccator. No

special precautions were necessary during the chromatography

process; however, prolonged (>6 h) exposure to visible light resulted

in the formation of black Ag nanoparticles and a loss of column

separation power. Visualization was accomplished with UV light and/

or ceric ammonium molybdate (CAM) solution. R_fvalues reported

were measured using a 10 × 2 cm TLC plate in a developing chamber

containing the solvent system described. Flash chromatography was

performed using Merck silica gel 60 230−400 mesh (60−63 μ, 60 Å

pore size). Analytical chiral stationary phase supercritical ﬂuid

chromatography (CSP-SFC) was performed on an Agilent 1100

HPLC equipped with an Aurora Systems A-5 supercritical CO₂

adapter for supercritical ﬂuid chromatography and a UV detector

(220 or 254 nm) using Daicel Chiralcel OD, OJ, OB or Chiralpak AD,

and AS columns as well as a Regis Whelk-O1 column. Normal-phase

HPLC was performed on an Agilent 1100 HPLC equipped with AD-

H, OJ-H, IB-3, naphthylleucine, and R,R-Beta-Gem columns.

Reversed-phase HPLC was performed on an Agilent 1100 HPLC

using a Chiralpak AD-RH or Chiralcel OJ-RH column.

General Procedure 1. To a 50 mL, oven-dried Schlenk ﬂask was

added washed, dry NaH (1.05 equiv) in a glovebox. The ﬂask was

transferred to a Schlenk line. Solvent (0.5 M) was added to aﬀord a

cloudy mixture. The mixture was placed in an ice bath and the

corresponding phenol (1 equiv) was added portion wise. During this

phase, substantial gas evolution was observed. The insoluble material

appeared to change in texture from an amorphous, cloudy particulate

matter to a ﬁnely distributed crystallized solid. The reaction was

removed from the ice bath and then allowed to stir at rt for 30 min.

Subsequently, cinnamyl chloride (1.1 equiv) was added dropwise via

syringe. After the addition was complete, the reaction was placed in an

oil bath and then heated to reﬂux for the speciﬁed amount of time.

After completion of the reaction as judged by TLC, the ﬂask was

removed from the heat source and allowed to cool to rt. The mixture

was then transferred to a separatory funnel, diluted with water (20

mL) and dichloromethane (20 mL), and then acidiﬁed to pH < 1 with

slow addition of a 6 M HCl solution. The aqueous layer was separated

and then back-extracted with a further 30 mL of dichloromethane. The

combined organic layers were then washed with brine, dried over

MgSO₄, and then ﬁltered through glass wool. The ﬁltrate was dried

under reduced pressure (∼3 mmHg, rotary evaporator). The residue

was then taken up in 10 mL of dichloromethane. Celite was added,

and the mixture was concentrated to aﬀord a white powder, which was

then subjected to ﬂash column chromatography. A second ﬂash

column chromatography operation was then performed using silica

impregnated with 10% AgNO₃(w/w).

Preparation of (E)-2-(3-Phenyl-2-propen-1-yl)phenol (3). Follow-

ing general procedure 1, NaH (252 mg, 10.5 mmol), CCl₄(20 mL),

phenol (0.94 g, 10 mmol), and cinnamyl chloride (1.53 mL, 11 mmol)

were combined in a 50 mL Schlenk ﬂask. The reaction was worked up

according to the general procedure. Silica gel ﬂash column

chromatography (3:1 hexanes/toluene, 30 mm diameter, 14 cm of

SiO₂) followed by a second silica gel ﬂash column chromatography

(9:1, hexanes/ethyl acetate, 30 mm diameter, 50 g of SiO₂(10%

AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 1.49 g of 3

(71%) as a pale oil that solidiﬁed upon standing. The spectroscopic

data matched those reported in the literature.⁴⁵Data for 3: H NMR

1

(500 MHz, CDCl₃) δ 7.39−7.35 (m, 2H), 7.33−7.28 (m, 2H), 7.25−

7.14 (m, 3H), 6.92 (td, J = 7.4, 1.2 Hz, 1H), 6.83 (dd, J = 7.9, 1.2 Hz,

1H), 6.52 (d, J = 15.9 Hz, 1H), 6.40 (dt, J = 15.9, 6.5 Hz, 1H), 3.59 (d,

J = 6.0 Hz, 1H); ¹³C NMR (126 MHz, CDCl₃) δ 154.2, 137.3, 131.8,

130.7, 128.8, 128.2, 128.1, 127.6, 126.5, 125.9, 121.3, 116.0, 34.4; MS

(EI, 70 eV, m/z) 210 (100, M⁺), 119 (33), 115 (38),104 (66), 91

(82), 69 (45).

Literature Preparations. Synthesis of Reagents. The following

compounds were prepared according to literature procedures: N-

(phenylthio)phthalimide (1a),¹⁶N-[(2,6-diisopropylphenyl)thio]-

phthalimde (1c),^{2 1}4-(N-azepano)-3,5-dimethyl-3,5-

dihydrodinaphtho[2,1-d:1′,2′][1,3,2]diazaphosphepine 4-selenide

(2a),¹⁶4-(diisopropylamino)-3,5-dimethyl-3,5-dihydrodinaphtho[2,1-

d:1′,2′-f ][1,3,2]diazaphosphepine 4-selenide (2b),¹⁷and 4-(diisobu-

O

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

Preparation of (E)-2-(3-Phenyl-2-propen-1-yl)-4-methylphenol

(3b). Following general procedure 1, NaH (252 mg, 10.5 mmol),

CCl₄(20 mL), p-cresol (1.08 g, 10 mmol), and cinnamyl chloride

(1.53 mL, 11 mmol) were combined in a 50 mL Schlenk ﬂask. The

reaction was worked up according to the general procedure. Silica gel

ﬂash column chromatography (3:1 hexanes/toluene, 30 mm diameter,

16 cm of SiO₂) followed by a second silica gel ﬂash column

chromatography (9:1, hexanes/ethyl acetate, 50 g of SiO₂(10%

AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 1.05 g of 3b

(47%) as a pale oil that solidiﬁed upon standing. The spectral data

matched those reported in the literature.⁴⁶Data for 3b: ¹H NMR (500

MHz, CDCl₃) δ 7.40−7.35 (m, 2H), 7.32 (td, J = 7.6, 1.1 Hz, 2H),

7.26−7.21 (m, 1H), 7.06 (t, J = 8.3 Hz, 1H), 6.84 (t, J = 7.5 Hz, 1H),

6.56 (d, J = 16.0 Hz, 0H), 6.41 (dtd, J = 15.8, 6.6, 1.0 Hz, 1H), 4.96 (s,

1H), 3.59 (dd, J = 6.7, 1.6 Hz, 2H), 2.28 (s, 3H); ¹³C NMR (126

MHz, CDCl₃) δ 149.7, 135.1, 129.3, 128.9, 128.2, 126.5, 126.2, 126.1,

125.2, 124.2, 124.1, 123.4, 113.6, 32.1, 18.5; MS (EI, 70 eV, m/z) 224

(100, M⁺), 209 (30), 133 (55), 115 (30), 104 (35), 91 (69), 77 (22).

Preparation of (E)-2-(3-Phenyl-2-propen-1-yl)-4-bromophenol

(3f). Following general procedure 1, NaH (252 mg, 10.5 mmol),

CCl₄(20 mL), 4-bromophenol (1.73 g, 10 mmol), and cinnamyl

chloride (1.53 mL, 11 mmol) were combined in a 50 mL Schlenk ﬂask.

The reaction was worked up according to the general procedure. Silica

gel ﬂash column chromatography (3:1 hexanes/toluene, 30 mm

diameter, 16 cm of SiO₂) followed by a second silica gel ﬂash column

chromatography (9:1, hexanes/ethyl acetate, 30 mm diameter, 50 g of

SiO₂(10% AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 1.62

g of 3f (56%) as a pale oil that solidiﬁed upon standing. The

spectroscopic data matched those reported in the literature.⁴⁹Data for

3f: ¹H NMR (500 MHz, CDCl₃) δ 7.42−7.19 (m, 5H), 6.74 (ddd, J =

9.9, 8.5, 1.3 Hz, 1H), 6.53 (dd, J = 15.9, 1.7 Hz, 1H), 6.36 (dt, J = 15.9,

6.6 Hz, 0H), 4.99 (d, J = 1.3 Hz, 0H), 3.55 (dd, J = 6.6, 1.6 Hz, 2H);

¹³C NMR (126 MHz, CDCl₃) δ 153.4, 150.1, 137.1, 133.2, 132.4,

130.8, 128.8, 128.3, 127.8, 127.1, 126.5, 117.7, 113.2, 34.1; MS (EI, 70

eV, m/z) 290 (71, M⁺+ 2), 288 (76, M⁺), 209 (93), 131 (24), 118

(39), 115 (45), 104 (100), 91 (82), 77 (41).

Preparation of (E)-2-(3-Phenyl-2-propen-1-yl)-4-chlorophenol

(3g). Following general procedure 1, NaH (252 mg, 10.5 mmol),

CCl₄(20 mL), 4-chlorophenol (1.28 g, 10 mmol), and cinnamyl

chloride (1.53 mL, 11 mmol) were combined in a 50 mL Schlenk ﬂask.

The reaction was worked up according to the general procedure. Silica

gel ﬂash column chromatography (3:1 hexanes/toluene, 30 mm

diameter, 16 cm of SiO₂) followed by a second silica gel ﬂash column

chromatography (9:1, hexanes/ethyl acetate, 50 g of SiO₂(10%

AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 1.49 g of 3g

(61%) as a yellow oil that solidiﬁed upon standing. The spectroscopic

data matched those reported in the literature.⁵⁰Data for 3g: ¹H NMR

(500 MHz, CDCl₃) δ 7.43−7.07 (m, 5H), 6.77 (d, J = 8.5 Hz, 1H),

6.57−6.49 (m, 1H), 6.36 (dt, J = 15.9, 6.6 Hz, 1H), 4.97 (s, 1H),

3.60−3.47 (m, 2H); ¹³C NMR (126 MHz, CDCl₃) δ 159.4, 152.8,

137.1, 132.4, 130.3, 128.8, 127.8, 127.7, 127.1, 126.5, 117.2,34.1; MS

(EI, 70 eV, m/z) 246 (33, M⁺+2), 244 (96, M⁺), 209 (75), 165 (19),

153 (42), 115 (42), 105 (60), 104 (100), 91 (87), 81 (40), 77 (59), 69

(81).

Preparation of (E)-2-(3-Phenyl-2-propen-1-yl)-6-methylphenol

(3c). Following general procedure 1, NaH (252 mg, 10.5 mmol),

CCl₄(20 mL), o-cresol (1.08 g, 10 mmol), and cinnamyl chloride

(1.53 mL, 11 mmol) were combined in a 50 mL Schlenk ﬂask. The

reaction was worked up according to the general procedure. Silica gel

ﬂash column chromatography (3:1 hexanes/toluene, 30 mm diameter,

14 cm of SiO₂) followed by a second silica gel ﬂash column

chromatography (9:1, hexanes/ethyl acetate, 30 mm diameter 50 g of

SiO₂(10% AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 1.38

g of 3c (62%) as a pale oil that solidiﬁed upon standing. The spectral

data matched those reported in the literature.⁴⁷Data for 3c: ¹H NMR

(500 MHz, CDCl₃) δ 7.41−7.35 (m, 2H), 7.34−7.30 (m, 2H), 7.23

(ddd, J = 12.8, 6.5, 1.7 Hz, 1H), 7.06 (t, J = 8.7 Hz, 1H), 6.90−6.80

(m, 1H), 6.56 (d, J = 16.0 Hz, 0H), 6.41 (ddd, J = 15.9, 8.0, 5.6 Hz,

1H), 3.59 (dd, J = 6.6, 1.7 Hz, 2H); ¹³C NMR (126 MHz, CDCl₃) δ

152.5, 136.9, 131.6, 129.4, 128.5, 128.1, 127.9, 127.4, 126.2, 124.9,

124.0, 120.4, 34.6, 15.9; MS (EI, 70 eV, m/z) 224 (100, M⁺), 209

(33), 168 (26), 141 (28), 120 (29), 115 (52), 105 (47), 91 (71), 77

(76), 69 (41).

Preparation of (E)-2-(3-Phenyl-2-propen-1-yl)-4-ﬂuorophenol

(3h). Following general procedure 1, NaH (252 mg, 10.5 mmol),

CCl₄(20 mL), 4-ﬂuorophenol (1.12 g, 10 mmol), and cinnamyl

chloride (1.53 mL, 11 mmol) were combined in a 50 mL Schlenk ﬂask.

The reaction was worked up according to the general procedure. Silica

gel ﬂash column chromatography (3:1 hexanes/toluene, 30 mm

diameter, 16 cm of SiO₂) followed by a second silica gel ﬂash column

chromatography (9:1, hexanes/ethyl acetate, 30 mm diameter, 50 g of

SiO₂(10% AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 1.73

g of 3h (76%) as a yellow oil that solidiﬁed upon standing.⁵¹Data for

3h: bp 100 °C (ABT, 3 × 10⁻⁵mm Hg); ¹H NMR (500 MHz,

CDCl₃) δ 7.40−7.32 (m, 2H, HC(12)), 7.29 (ddd, J = 7.9, 6.2, 1.3 Hz,

2H, HC(11)), 7.24−7.18 (m, 1H, HC(13)), 6.88 (dd, J = 9.0, 3.0 Hz,

1H, HC(5)), 6.82 (td, J = 8.3, 3.1 Hz, 1H, HC(3)), 6.74 (dd, J = 8.7,

4.7 Hz, 1H, HC(2)), 6.49 (dt, J = 15.7, 1.5 Hz, 1H, HC(9)), 6.33 (dt, J

= 15.9, 6.6 Hz, 1H, HC(8)), 3.52 (dd, J = 6.6, 1.5 Hz, 2H, HC(7));

¹³C NMR (126 MHz, CDCl₃) δ 157.2 (d, J = 232 Hz, C4), 149.8

Preparation of (E)-2-(3-Phenyl-2-propen-1-yl)-4-methoxyphenol

(3e). Following general procedure 1, NaH (252 mg, 10.5 mmol), CCl₄

(20 mL), 4-methoxyphenol (1.24 g, 10 mmol), and cinnamyl chloride

(1.53 mL, 11 mmol) were combined in a 50 mL Schlenk ﬂask. The

reaction was worked up according to the general procedure. Silica gel

ﬂash column chromatography (3:1, hexanes/toluene, 30 mm diameter,

14 cm of SiO₂) followed by a second silica gel ﬂash column

chromatography (9:1 hexanes/ethyl acetate, 30 mm diameter, 50 g of

SiO₂(10% AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 1.17

g of 3e (49%) as a pale oil that solidiﬁed upon standing. The spectral

data matched those reported in the literature.⁴⁸Data for 3e: ¹H NMR

(500 MHz, CDCl₃) δ 7.39−7.14 (m, 5H), 6.81−6.65 (m, 3H), 6.48

(d, 1H, J = 15.9 Hz), 6.35 (dt, J = 15.9, 6.5 Hz, 1H), 3.74 (s, 3H), 3.52

(d, J = 5.7 Hz, 1H); ¹³C NMR (126 MHz, CDCl₃) δ 153.8, 147.9,

137.0, 131.5, 128.5, 127.7, 127.3, 126.8, 126.2, 116.4, 116.0, 112.6,

55.7, 34.3; MS (EI, 70 eV, m/z) 240 (100, M⁺), 149 (66), 136 (99),

121 (25), 115 (31), 108 (44), 91 (79), 77 (21).

(C1), 136.9 (C10), 132.0 (C9), 128.6 (C12), 127.5 (C13), 127.3 (d, J

= 7 Hz, C6), 126.9 (C9), 126.2 (C11), 116.6 (d, J = 23 Hz, C5)),

116.5 (d, J = 8.5 Hz, C2), 113.9 (d, J = 25 Hz, C3), 34.0 (C7); ¹⁹F

NMR (470 MHz, CDCl₃) δ −123.9 (m); IR (ATR, cm⁻¹) 3429 (br),

3027 (w), 1619 (w), 1598 (w), 1494 (s), 1438 (s), 1327 (w), 1254

(w), 1176 (s), 1141 (m), 1090 (w), 1028 (w), 958 (m), 928 (w), 872

P

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

(w), 809 (m), 748 (s), 731 (m), 716 (w), 692 (s); MS (EI, 70 eV, m/

z) 228 (100, M⁺), 137 (46), 115 (21), 109 (21), 104 (85), 91 (57)

HRMS calcd for C₁₅H₁₃OF 228.0950, found 228.0952, error 0.7; TLC

R_f0.30 (4:1, hexanes/ethyl acetate) [UV,CAM].

Preparation of (E)-2-(5-Phenylpent-2-en-1-yl)phenol (3l). Follow-

ing general procedure 1, NaH (126 mg, 5.25 mmol), toluene (10 mL),

phenol (470 mg, 5 mmol), and (E)-(5-chloro-3-penten-1-yl)benzene⁵³

(990 mg, 5.5 mmol) were combined in a 50 mL Schlenk ﬂask. The

reaction was worked up according to the general procedure. Silica gel

ﬂash column chromatography (3:1, hexanes/toluene, 30 mm diameter,

16 cm of SiO₂) followed by a second silica gel ﬂash column

chromatography (9:1, hexanes/ethyl acetate, 30 mm diameter SiO₂

(10% AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 940 mg

of 3l (79%) as a clear oil. Data for 3l: bp 160 °C (ABT, 0.05 mmHg);

¹H NMR (500 MHz, CDCl₃) δ 7.30−7.26 (m, 2H, HC(aryl)), 7.21−

7.10 (m, 5H, HC(aryl)), 7.05 (dd, J = 7.5, 1.7 Hz, 1H, HC(5)), 6.87

(td, J = 7.4, 1.2 Hz, 1H, HC(4)), 6.82 (dd, J = 8.0, 1.2 Hz, 1H,

HC(2)), 5.64 (ddd, J = 5.5, 3.9, 1.7 Hz, 2H, HC(8), HC(9)), 5.01 (s,

1H, OH), 3.35 (dd, J = 4.5, 1.8 Hz, 2H, HC(7)), 2.70 (dd, J = 8.8, 6.7

Hz, 2H, HC(11)), 2.42−2.32 (m, 2H, HC(10)); ¹³C NMR (126

MHz, CDCl₃) δ 154.2 (C1), 141.7 (C12), 131.9 (C9), 130.2 (C5),

128.4 (C13), 128.3 (C8), 128.2 (C14), 127.8 (C15), 125.8 (C3, C6),

120.8 (C4), 115.8 (C2), 35.7 (C11), 34.2 (C10), 34.0 (C7); IR (ATR,

cm⁻¹) 3531 (br), 3083 (w), 3062 (w), 3026 (w), 2922 (w), 2852 (w),

1602 (w), 1591 (w), 1585 (w), 1489 (m), 1453 (s), 1435 (w), 1328

(w), 1254 (m), 1212 (m), 1169 (m), 1152 (w), 1097 (m), 1038 (w),

1030 (w), 970 (m), 936 (w), 916 (w), 842 (w), 798 (w), 749 (s), 697

(s); MS (EI, 70 eV, m/z) 238 (95, M⁺), 147 (100), 131 (21), 107

(42), 91 (62); TLC R_f0.27 (4:1, hexanes/ethyl acetate) [UV,CAM].

Anal. Calcd for C₁₇H₁₈O (238.33): C, 85.67; H, 7.61. Found: C, 85.31;

H, 7.63.

Preparation of (E)-2-(3-Phenyl-2-propen-1-yl)-4-

(triﬂuoromethyl)phenol (3i). Following general procedure 1, NaH

(126 mg, 5.25 mmol), toluene (10 mL), 4-(triﬂuoromethyl)phenol

(810 mg, 5 mmol), and cinnamyl chloride (715 μL, 5.5 mmol) were

combined in a 50 mL Schlenk ﬂask. The reaction was worked up

according to the general procedure. Silica gel ﬂash column

chromatography (3:1 hexanes/toluene, 30 mm diameter, 16 cm of

SiO₂) followed by a second silica gel ﬂash column chromatography

(9:1, hexanes/ethyl acetate, 30 mm diameter, 50 g of SiO₂(10%

AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 885 mg of 3i

(64%) as a clear oil that solidiﬁed upon standing. Data for 3i: bp 120

°C (ABT, 3.2 × 10⁻⁵mm Hg); ¹H NMR (500 MHz, CDCl₃) δ 7.47−

7.28 (m, 5H, HC(aryl)), 7.28−7.21 (m, 2H, HC(11)), 6.88 (d, J = 8.2

Hz, 1H, HC(2)), 6.54 (dt, J = 15.8, 1.6 Hz, 1H, HC(9)), 6.36 (dt, J =

15.9, 6.6 Hz, 1H, HC(8)), 5.26 (s, 1H, OH), 3.61 (d, J = 6.6 Hz, 2H,

HC(7)); ¹³C NMR (126 MHz, CDCl₃) δ 156.9 (C1), 136.9 (C10),

132.5 (C9), 128.9 (C-aryl), 127.9 (q, J = 3.8 Hz, C5), 127.8 (C-aryl),

126.9, (C8), 126.50 (C-aryl), 125.5 (q, J = 3.9 Hz, C3), 116.0 (C2),

34.3 (C7); ¹⁹F NMR (470 MHz, CDCl₃) δ −61.56 (s); IR (ATR,

cm⁻¹) 3498 (br), 3027 (w), 2923 (w), 1614 (w), 1496 (w), 1485 (w),

1448 (w), 1355 (s), 1335 (s), 1250 (w), 1201 (s), 1152 (s), 1114 (s),

1028 (w), 967 (m), 928 (w), 889 (w), 805 (w), 743 (m), 692 (s); MS

(EI, 70 eV, m/z) 278 (4, M⁺), 260 (83), 223 (30),184 (23), 168 (60),

156 (35), 149 (40), 141 (60), 128 (51), 115 (74), 104 (40), 91 (69),

77 (100); HRMS calcd for C₁₆H₁₃OF₃278.0919, found 278.0912,

error −2.4; TLC R_f0.33 (4:1, hexanes/ethyl acetate) [UV,CAM].

Preparation of 2-(3-Methylbut-2-en-1-yl)phenol (3m). Following

general procedure 1, NaH (126 mg, 5.25 mmol), CCl₄(10 mL),

phenol (470 mg, 5 mmol), and prenyl chloride (605 μL, 5.5 mmol)

were combined in a 50 mL Schlenk ﬂask. The reaction was worked up

according to the general procedure. Silica gel ﬂash column

chromatography (3:1, hexanes/toluene, 30 mm diameter, 16 cm of

SiO₂) followed by a second silica gel ﬂash column chromatography

(9:1, hexanes/ethyl acetate, 30 mm diameter, 50 g of SiO₂(10%

AgNO₃w/w)) aﬀorded a pale oil. Distillation aﬀorded 525 mg (65%)

of 3m as a clear oil. The spectroscopic data matched those reported in

the literature.⁵⁴Data for 3m: ¹H NMR (500 MHz, CDCl₃) δ 7.14 (t, J

= 7.4 Hz, 1H), 6.89 (td, J = 7.5, 1.2 Hz, 1H), 6.83 (dt, J = 7.6, 1.3 Hz,

1H), 5.35 (dddd, J = 7.3, 5.8, 2.9, 1.5 Hz, 1H), 5.09 (s, 1H), 3.38 (d, J

= 7.2 Hz, 2H), 1.80 (dd, J = 5.1, 1.4 Hz, 6H); ¹³C NMR (126 MHz,

CDCl₃) δ 154.1, 129.8, 127.3, 121.8, 120.6, 115.5, 29.4, 25.7, 17.7; MS

(EI, 70 eV, m/z) 162 (21, M⁺), 145 (39), 133 (38), 115 (33), 107

(100), 91 (77), 77 (47).

Preparation of (E)-1-(3-Phenyl-2-propen-1-yl)naphthalen-2-ol

(3d). Following general procedure 1, NaH (252 mg, 10.5 mmol),

CCl₄(20 mL), 2-naphthol (1.44g, 10 mmol), and cinnamyl chloride

(1.53 mL, 11 mmol) were combined in a 50 mL Schlenk ﬂask. The

reaction was worked up according to the general procedure. Silica gel

ﬂash column chromatography (3:1 hexanes/toluene, 30 mm diameter,

16 cm of SiO₂) followed by a second silica gel ﬂash column

chromatography (9:1, hexanes/ethyl acetate, 30 mm diameter, 50 g of

SiO₂(10% AgNO₃w/w)) aﬀorded 2.05 g (79%) of 3d as a pale pink

solid. The spectroscopic data matched those reported in the

literature.⁵²Data for 3d: mp 62−64 °C (hexanes); H NMR (500

1

MHz, CDCl₃) δ 8.00 (d, J = 8.6 Hz, 1H), 7.82 (d, J = 8.1 Hz, 1H),

7.72 (d, J = 8.8 Hz, 1H), 7.52 (ddt, J = 8.2, 6.9, 1.3 Hz, 1H), 7.39−7.35

(m, 1H), 7.34−7.29 (m, 2H), 7.30−7.24 (m, 2H), 7.22−7.18 (m, 1H),

7.16−7.09 (m, 2H), 6.47 (p, J = 1.9 Hz, 2H), 4.01 (dd, J = 3.1, 1.4 Hz,

2H); ¹³C NMR (126 MHz, CDCl₃) δ 151.1, 137.2, 133.3, 130.9,

129.4, 128.6, 128.4, 128.3, 127.6, 127.1, 126.6, 126.1, 126.0, 123.2,

123.1, 117.9, 117.1, 28.43; MS (EI, 70 eV, m/z) 260 (100, M⁺), 169

(59), 156 (36), 128 (38), 117 (25), 115 (24), 104 (44), 91 (38).

Preparation of (E)-2-(3-(Furan-2-yl)allyl)phenol (3j). To a 5 mL

Schlenk ﬂask equipped with a stir bar were added, under argon, (E)-3-

(furan-2-yl)-1-(2-hydroxyphenyl)prop-2-en-1-one S2 (749 mg, 3.5

mmol), THF (3.5 mL), and triethylamine (535 μL, 3.85 mmol, 1.1

equiv). To this solution was added with vigorous stirring ethyl

Q

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

chloroformate (370 μL, 3.85 mmol, 1.1 equiv). The color of the

solution changed from dark brown to light gold with concomitant

precipitation of solid triethylammonium chloride. The mixture was

stirred for a further 30 min, and the reaction was assayed for

completion by TLC. To a separate 50 mL ﬂask, under argon and

equipped with a stir bar, were added ethanol (15 mL) and cerium

chloride heptahydrate (1.56 g, 4.2 mmol, 1.2 equiv), and the resulting

clear solution was stirred for 20 min. The solution containing the ethyl

carbonate of the starting material was then ﬁltered into the second

ﬂask through glass wool. THF (2 × 5 mL) was used to wash the ﬂask

and ﬁlter cake. To the resulting light yellow solution was added

NaBH₄(160 mg, 4.2 mmol, 1.2 equiv), and the solution was stirred for

a further 1 h. Water (1 mL) was added to quench the reaction, and the

solution was then stirred for a further 15 min. The resulting

heterogeneous mixture was transferred to a 125 mL separatory funnel

and diluted with water (15 mL) and CH₂Cl₂(30 mL). The layers were

separated, and the aqueous layer was extracted with CH₂Cl₂(2 × 30

mL). The organic layers were combined, dried over MgSO₄, and

concentrated by rotary evaporation (30 °C, 3 mmHg). The crude

material thus obtained was redissolved in 10 mL of CH₂Cl₂and

adsorbed onto Celite. Puriﬁcation by silica gel ﬂash column

chromatography (12:1 hexanes/ethyl acetate, 20 mm diameter, 18

cm of SiO₂) followed by bulb-to-bulb diﬀusion pump distillation

aﬀorded 112 mg (16%) of 3j as a pale yellow oil. Data for 3j: bp 70 °C

(ABT, 2 × 10⁻⁴mm Hg); ¹H NMR (500 MHz, CDCl₃) δ 7.33 (d, J =

1.8 Hz, 1H, HC(11)), 7.17 (m, 2H, HC(3), HC(5)), 6.93 (t, J = 7.5

Hz, 1H, HC(4)), 6.83 (d, J = 7.9 Hz, 1H, HC(6)), 6.41−6.31 (m, 2H,

HC(8), HC(12)), 6.28 (d, J = 15.9 Hz, 1H, HC(9)), 6.18 (d, J = 3.3

Hz, 1H, HC(13)), 4.87 (s, 1H, OH), 3.55 (d, J = 6.3 Hz, 2H, HC(7));

¹³C NMR (126 MHz, CDCl₃) δ 154.1 (C1), 141.9 (C2), 130.8 (C3),

chromatography operations (1, 9:1 hexanes/ethyl acetate, 30 mm

diameter, 15 cm of SiO₂; 2, 50:1 toluene/ethyl acetate, 30 mm

diameter, 14 cm of SiO₂) and then distilled under vacuum to aﬀord

330 mg (15%) of 3n as a clear oil. The spectroscopic data match those

reported in the literature.⁵⁵Data for 3n: ¹H NMR: (400 MHz, CDCl₃)

δ 7.38 (d, J = 6.9 Hz, 2H, HC(12)), 7.33 (t, J = 7.7 Hz, 2H, HC(13)),

7.24 (t, J = 7.2 Hz, 1H, HC(14)), 7.20 (dd, J = 7.5, 1.6 Hz, 1H,

HC(3)), 7.14 (td, J = 7.7, 1.7 Hz, 1H, HC(5)), 6.93 (td, J = 7.4, 1.1

Hz, 1H, HC(4)), 6.81 (dd, J = 8.0, 1.2 Hz, 1H, HC(6)), 6.48 (dt, J =

15.9, 1.4 Hz, 1H, HC(10)), 6.33 (dt, J = 15.8, 6.9 Hz, 1H, HC(9)),

4.69 (s, 1H, OH), 2.84 (dd, J = 8.8, 6.6 Hz, 2H, HC(7)), 2.62−2.54

(app q, J = 7.5 Hz, 2H, HC(8)); ¹³C NMR (126 MHz, CDCl₃) δ

153.7, 137.9, 130.7, 130.6, 130.4, 128.7, 127.5, 127.2, 126.2, 121.1,

115.5, 115.1, 33.5, 30.4; MS (EI, 70 eV, m/z) 224 (59, M⁺), 117

(100), 115 (26), 107 (97), 91 (20), 77 (20).

Preparation of (E)-2-(5-Phenylpent-4-en-1-yl)phenol (3o). Di-

chloromethane was degassed by being purged with argon for 30 min.

To a 10 mL ﬂask was added Grubbs-I-indenylidene catalyst (46 mg,

0.05 mmol, 0.033 equiv) in a glovebox. The ﬂask was transferred to a

Schlenk line, and degassed dichloromethane (3 mL), 3q (241 mg, 1.5

mmol) and styrene (890 μL, 7.5 mmol, 5 equiv) were added in order.

The solution was stirred for 24 h, whereupon a second portion of

catalyst (23 mg, 0.025 mmol, 0.016 equiv) was added. The solution

was then stirred for a further 24 h. The solution was then transferred

to a 100 mL RB ﬂask, and the volatiles were removed by rotary

evaporation (30 °C. Three mm Hg). The material was redissolved in

10 mL of CH₂Cl₂and adsorbed onto Celite. Puriﬁcation by silica gel

ﬂash column chromatography (12:1 hexanes/ethyl acetate, 20 mm

diameter, 16 cm of SiO₂) followed by distillation aﬀorded 150 mg

(42%) of 3o as a clear oil. Data for 3o: bp 150 °C (ABT, 0.05 mmHg);

¹H NMR (400 MHz, CDCl₃) δ 7.35 (d, J = 8.1 Hz, 2H, HC(13)), 7.30

(t, J = 7.6 Hz, 2H, HC(14)), 7.20 (t, J = 7.2 Hz, 1H, HC(15)), 7.14

(d, J = 7.4 Hz, 1H, HC(3), 7.09 (t, J = 7.7 Hz, 1H, HC(5)), 6.88 (t, J =

7.4 Hz, 1H, HC(4)), 6.76 (d, J = 8.0 Hz, 1H, HC(6)), 6.42 (d, J = 16.1

Hz, 1H, HC(11), 6.26 (dt, J = 15.9, 6.7 Hz, 1H, HC(10), 4.65 (s, 1H,

OH), 2.68 (t, J = 7.7 Hz, 1H, HC(7), 2.29 (q, J = 6.9 Hz, 1H, HC(9)),

1.82 (p, J = 7.4 Hz, 1H, HC(8)); ¹³C NMR (126 MHz, CDCl₃) δ

153.8 (C1), 138.1 (C12), 130.9 (C3), 130.6 (C11), 130.5 (C10),

128.8 (C14), 128.6 (C2), 127.4 (C5), 127.2 (C15), 126.3 (C13),

121.1 (C4), 115.6 (C6), 33.0 (C9), 29.7 (C7), 29.6 (C8); IR (ATR,

cm⁻¹) 3401 (br), 3059 (w), 3025 (w), 2926 (w), 2857 (w), 1650 (w),

1592.1 (w), 1490 (m), 1454 (s), 1327 (m), 1234 (m), 1169 (w), 1104

(w), 1068 (w), 1042 (w), 1028 (w), 963 (s), 933 (w), 911 (w), 842

(w), 747 (s), 691 (s); MS (EI, 70 eV, m/z) 238 (100, M⁺), 147 (45),

131 (34), 117 (42), 107 (74), 91 (43), 77 (25); TLC R_f0.44 (4:1

hexanes/ethyl acetate) [UV,CAM]. Anal. Calcd for C₁₆H₁₆O (224.34):

C, 85.67; H, 7.61. Found: C, 85.27; H, 7.43.

128.2 (C5), 127.0 (C8), 121.3 (C9), 120.2 (C4), 115.9 (C6), 111.4,

107.3 (C13), 33.9 (C7); IR (ATR, cm⁻¹) 3409 (br), 1593 (w), 1489

(m), 1455 (s), 1232 (m), 1095 (w), 1012 (m), 964 (m), 753 (s); MS

(EI, 70 eV, m/z) 200 (100, M⁺), 171 (16), 131 (19), 107 (16), 94

(54), 81 (24); HRMS calcd for C₁₃H₁₂O₂200.0837, found 200.0842,

error 2.3; TLC R_f0.36 (4:1, hexanes/ethyl acetate) [UV,CAM].

Preparation of 2-Substituted Phenols with Two or More

Methylene Tethers.

Preparation of (E)-2-(4-Phenylbut-3-en-1-yl)phenol⁵⁵(3n). To a 50

mL Schlenk ﬂask under argon was added BuLi (2.23 M in hexanes, 9.5

mL, 21 mmol, 2.1 equiv) followed by a further 4 mL of hexanes.

Tetramethylethylenediamine (TMEDA, 2.44 g, 21 mmol, 2.1 equiv)

was added via syringe, and the ﬂask was placed in a −78 °C bath (dry

ice/i-PrOH). The internal temperature was monitored until it dropped

below −40 °C. Cresol (1.08 g, 10 mmol) was dissolved in hexanes (6

mL) and added to the cold mixture as a solution. To this cold mixture

was then added solid KO-t-Bu (2.36 g, 21 mmol, 2.1 equiv). The

formation of a yellow suspension was observed. The ﬂask was then

removed from the −78 °C bath and placed in a −20 °C bath (i-PrOH,

IBC-4A Cryocool), and the solution within was allowed to stir for 30

min. The ﬂask was then removed and allowed to warm to rt over 15

min. THF (10 mL) was added. The ﬂask was then returned to the

aforementioned −78 °C bath, and the internal temperature was

monitored until it was below −60 °C and then allowed to equilibrate

for a further 20 min. Cinnamyl chloride (1.98 g, 13 mmol, 1.3 equiv)

was dissolved in THF (2 mL) and added to the mixture. The solution

was then allowed to warm to rt and stirred for 1 h. The reaction was

quenched by the addition of aq satd NH₄Cl (10 mL). The mixture was

transferred to a 250 mL separatory funnel and the pH adjusted to <1

with 6 M HCl. The layers were separated, and the aqueous layer was

extracted with ether (2 × 30 mL). The organic layers were combined,

dried over MgSO₄, ﬁltered, and concentrated (30 °C, 3 mmHg). The

product was puriﬁed by two successive silica gel ﬂash column

Preparation of Ethyl (E)-7-(2-Hydroxyphenyl)hept-4-enoate (3r).

To a 50 mL RB ﬂask equipped with a stir bar were added, under argon,

LiI (1.87 g, 14 mmol, 4 equiv) and NaCN (189 mg, 3.85 mmol, 1.1

equiv). To this was added DMSO (10 mL), and the solution was

R

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

stirred for 15 min. The diester S3 (1.12 g, 3.5 mmol) was then

dissolved in DMSO (10 mL) and added to the stirring solution. The

solution was heated to 160 °C (internal temperature, oil bath) for 3 h.

Consumption of starting material was monitored by TLC. The ﬂask

was then removed from the heat source and allowed to cool to rt. The

solution was transferred to a 125 mL separatory funnel and diluted

with water (50 mL) and ethyl acetate (50 mL). Use of other solvent

ratios occasionally resulted in persistent emulsions. The layers were

separated, and the aqueous layer was extracted with ethyl acetate (50

mL). The organic layers were washed thoroughly (5 × 30 mL water)

and then washed with brine (15 mL), dried over MgSO₄, ﬁltered, and

concentrated by rotary evaporation (30 °C, 3 mmHg). The material

was then redissolved in 10 mL of diethyl ether and adsorbed onto

Celite. Puriﬁcation by silica gel ﬂash chromatography (7:1 hexanes/

ethyl acetate, 20 mm diameter, 18 cm of SiO₂) followed by bulb-to-

bulb vacuum distillation aﬀorded 476 mg (57%) of 3r as a clear oil.

(C10), 130.6 (C5), 128.8 (C9), 128.0 (C2), 127.4 (C3), 121.1

(C4),115.6 (C6), 34.1 (C11), 32.9 (C7), 30.3 (C8), 27.8 (C12); IR

(ATR, cm⁻¹) 3184 (br), 3042 (w), 2934 (w), 2908 (w), 1703 (s),

1613 (w), 1591 (m), 1503 (w), 1455 (s), 1441 (m), 1425 (m), 1409

(m), 1373 (m), 1280 (m), 1262 (w), 1236 (s), 1192 (s), 1109 (w),

1041 (w), 990 (w), 974 (s), 930 (w), 909 (w), 845 (w), 821 (m), 747

(s), 688 (w); MS (EI, 70 eV, m/z) 220 (16, M⁺), 137 (16), 107 (100);

TLC R_f0.06 (4:1, hexanes/ethyl acetate) [UV,CAM]. Anal. Calcd for

C₁₃H₁₆O₃(220.27): C, 70.89; H, 7.32. Found: C, 71.12; H, 7.24.

1

Data for 3r: bp 130 °C (ABT, 0.05 mmHg); H NMR (500 MHz,

CDCl₃) δ 7.15−7.05 (m, 2H, HC(3), HC(5)), 6.87 (t, J = 7.4 Hz, 1H,

HC(4)), 6.78 (d, J = 7.9 Hz, 1H, HC(6)), 5.57 (dt, J = 15.3, 6.7 Hz,

1H, HC(9)), 5.47 (dt, J = 15.3, 6.0 Hz, 1H, HC(10)), 5.24 (s, 1H,

OH), 4.16 (q, J = 7.1 Hz, 2H, HC(14)), 2.68 (dd, J = 8.7, 6.7 Hz, 2H,

HC(11)), 2.42−2.26 (m, 6H, HC(7), HC(8), HC(10)), 1.28 (t, J =

7.1 Hz, 3H, HC(15)); ¹³C NMR (126 MHz, CDCl₃) δ 173.8 (C13),

153.9 (C1), 131.2 (C9), 130.5 (C3), 129.1 (C10), 128.2 (C2), 127.4

(C5), 120.9 (C4), 115.6 (C6), 60.7 (C14), 34.5 (C7/C8/C10), 32.9

(C7/C8/C10), 30.4(C11), 28.1 (C7/C8/C10), 14.5 (C15); IR (ATR,

cm⁻¹) 3416 (br), 2981 (w), 2926 (w), 2854 (w), 1706 (s), 1608 (w),

1593 (w), 1504 (w), 1490 (w), 1455 (s), 1372 (m), 1344 (w), 1299

(w), 1232 (s), 1176 (s), 1150 (s), 1101 (m), 1033 (m), 968 (m), 850

(m), 751 (s); MS (EI, 70 eV, m/z) 248 (29, M⁺), 107 (100); TLC R_f

0.31 (4:1, hexanes/ethyl acetate) [UV,CAM]. Anal. Calcd for

C₁₅H₂₀O₃(248.32): C, 72.55; H, 8.12. Found: C, 72.43; H, 8.14.

Preparation of (E)-2-(7-Hydroxyhept-3-en-1-yl)phenol (3u). To a

50 mL Schlenk ﬂask under argon were added lithium aluminum

hydride (126 mg, 3.3 mmol, 1.5 equiv) and THF (4 mL). The ﬂask

was placed in an ice bath for 10 min. Ester 3r (520 mg, 2.2 mmol) was

dissolved in THF (4 mL) and added dropwise to the cold solution.

The solution was then allowed to stir at 0 °C for 2 h. The ﬂask was

then once again placed in an ice bath for 15 min. The reaction was

quenched by the addition of water (2 mL, dropwise) with substantial

gas evolution and the formation of copious amounts of solids. After

addition of water was complete, a further 5 mL of water was added,

followed by dropwise addition of 6 M HCl to pH < 2. The resulting

biphasic mixture was transferred to a 125 mL separatory funnel and

then shaken well. The layers were separated, and the aqueous layer was

extracted with diethyl ether (3 × 20 mL). The organic layers were

combined, washed with brine (15 mL), dried over MgSO₄, ﬁltered,

and concentrated by rotary evaporation (30 °C, 3 mmHg). The

resulting material was dissolved in 10 mL of ether and adsorbed onto

Celite. Puriﬁcation by silica gel ﬂash column chromatography (3:1

hexanes/ethyl acetate, 20 mm diameter, 17 cm of SiO₂) followed by

bulb-to-bulb distillation aﬀorded 416 mg (91%) of 3u as a clear oil.

1

Data for 3u: bp 140 °C (ABT, 0.05 mmHg); H NMR (500 MHz,

CDCl₃) δ 7.17−7.07 (m, 2H, HC(3), HC(5)), 6.90 (t, J = 8.0 Hz, 1H,

HC(4)), 6.79 (d, J = 8.2 Hz, 1H, HC(6)), 5.56 (dt, J = 15.1, 6.5 Hz,

1H, HC(9)), 5.48 (dt, J = 15.6, 6.5 Hz, 1H, HC(10)), 3.65 (t, J = 6.5

Hz, 2H, HC(13)), 2.71 (dd, J = 8.3, 6.8 Hz, 2H, HC(7)), 2.35 (q, J =

7.4 Hz, 2H, HC(8)), 2.12 (q, J = 6.8 Hz, 2H, HC(11)), 1.65 (p, J = 6.8

Hz, 2H, HC(12)); ¹³C NMR (126 MHz, CDCl₃) δ 154.2 (C1), 130.8

(C10), 130.5 (C5), 130.3 (C9), 128.5 (C2), 127.3 (C3), 120.7 (C4),

115.7 (C6), 62.6 (C13), 33.1 (C8), 32.2 (C12), 30.3 (C7), 29.0

(C11); IR (ATR, cm⁻¹) 3307 (br), 3033 (w), 2932 (w), 2851 (w),

1607 (w), 1592 (w), 1504 (w), 1489 (w), 1455 (m), 1354 (w), 1238

(m), 1178 (w), 1153 (w), 1094 (w), 1042 (m), 1015 (w), 968 (w),

929 (w), 847 (w), 750 (s); MS (EI, 70 eV, m/z) 206 (20, M⁺), 120

(14), 107 (100); TLC R_f0.1 (4:1 hexanes/ethyl acetate) [UV,CAM].

Anal. Calcd for C₁₃H₁₈O₂(206.29): C, 75.69; H, 8.80. Found: C,

75.48; H, 8.56.

Preparation of (E)-7-(2-Hydroxyphenyl)hept-4-enoic Acid (3t). To

a 10 mL Schlenk ﬂask under argon were added ester 3r (447 mg, 1.8

mmol) and THF (10 mL). Water was added via syringe (2.5 mL)

followed by solid LiOH·H₂O (272 mg, 6.48 mmol, 3.6 equiv). The

solution was allowed to stir 16 h at rt. The mixture was transferred to a

60 mL separatory funnel and diluted with ether (30 mL) and water

(20 mL). The biphasic mixture was acidiﬁed with 1 M HCl to pH < 2,

whereupon a white precipitate formed, which disappeared upon

thorough shaking. The layers were separated, and the aqueous layer

was extracted with ether (2 × 15 mL). The organic layers were

combined, washed with brine (15 mL), dried over MgSO₄, ﬁltered,

and concentrated by rotary evaporation (30 °C, 3 mmHg). The

material was redissolved in 10 mL of diethyl ether and then was

adsorbed onto Celite. Puriﬁcation by silica gel ﬂash column

chromatography (4:1 hexanes/ethyl acetate, 30 mm diameter, 16 cm

of SiO₂) followed by recrystallization from hexanes (5 mL) aﬀorded

326 mg (83%) of 3t as white spindles. Data for 3t: mp 108−110 °C

(hexanes); ¹H NMR (500 MHz, CDCl₃) δ 7.17−7.08 (m, 2H, HC(3),

HC(5)), 6.91 (t, J = 7.4 Hz, 1H, HC(4)), 6.80 (d, J = 7.9 Hz, 1H,

HC(2)), 5.60 (dt, J = 15.6, 6.5 Hz, 1H, HC(9)), 5.50 (dt, J = 15.4, 6.3

Hz, 1H, HC(10)), 2.70 (dd, J = 8.5, 6.6 Hz, 2H, HC(8)), 2.45 (t, J =

7.1 Hz, 2H, HC(11)), 2.35 (app p, J = 6.8 Hz, 4H, HC(7), HC(12));

¹³C NMR (126 MHz, CDCl₃) δ 179.0 (C13), 153.6 (C1), 131.4

S

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

1

Preparation of (E)-2-(7-Methoxyhept-3-en-1-yl)phenol (3s). This

compound was prepared in two ways. Method A: To a 10 mL Schlenk

ﬂask in a glovebox was added NaH (27 mg, 1.1 mmol, 2.1 equiv). The

ﬂask was transferred to the Schlenk line, and THF (1.5 mL) was

added. To this was added the alcohol 3u (103 mg, 0.5 mmol) in THF

(0.5 mL). The solution was stirred for 15 min at rt, and MeI (34 μL,

0.55 mmol, 1.1 equiv) was added dropwise. The solution was then

allowed to stir at rt for 16 h. The reaction was quenched by the

addition of water (1 mL) and 1 M HCl (0.5 mL). The mixture was

transferred to a 60 mL separatory funnel and diluted with ether (10

mL) and water (10 mL) and the pH adjusted to <2 with 3 M HCl.

The layers were separated, and the aqueous layer was extracted with

ether (2 × 10 mL). The organic layers were combined, washed with

brine (15 mL), dried over MgSO₄, and concentrated by rotary

evaporation (30 °C, 3 mmHg). Puriﬁcation by silica gel ﬂash column

chromatography (12:1 hexanes/ethyl acetate, 20 mm diameter, 16 cm

of SiO₂) aﬀorded 63 mg (57%) of 3s as a clear oil. Method B: To a 10

mL Schlenk ﬂask were added S4 (368 mg, 1.3 mmol) and THF (15

mL). The ﬂask was cooled to −76 °C (internal temperature, dry ice/i-

PrOH). BuLi (2.3 M, 620 μL, 1.43 mmol, 1.1 equiv) was added via

syringe, and the solution was allowed to stir for 1 h. Triisopropyl

borate (489 mg,2.6 mmol, 2.0 equiv) was added via syringe, and the

solution was allowed to warm to rt. After the solution was stirred for 6

h at rt, the ﬂask was placed in an ice bath. In a separate 25 mL RB

ﬂask, a basic hydrogen peroxide solution was prepared by combining

20 mL of 30% H₂O₂with 2 g of NaOH. A portion of this solution (2

mL) was then added dropwise to the ﬂask containing the borate

(strong exotherm) followed by the remaining 8 mL of basic peroxide,

and the solution was allowed to stir for 5 h at rt. The excess peroxide

solution was quenched with satd aq Na₂S₂O₃. After the allotted time

had passed, the mixture was transferred to a 125 mL separatory funnel,

and satd aq Na₂S₂O₃was added until no more peroxide was evident

(Quantiﬁx test strip). The mixture was then diluted with ether (15

mL) and water (15 mL) and acidiﬁed with 1 M HCl to pH < 2. The

layers were separated, and the aqueous layer was extracted with ether

(3 × 10 mL). The organic layers were combined, dried over MgSO₄,

ﬁltered, and concentrated by rotary evaporation (30 °C, 3 mmHg).

Puriﬁcation by silica gel ﬂash column chromatography (12:1 hexanes/

ethyl acetate, 20 mm diameter, 16 cm of SiO₂) followed by bulb-to-

bulb vacuum distillation aﬀorded 213 mg (76%) of 3s as a clear oil

mixture. After the reaction was complete as judged by TLC and H

NMR spectroscopy, Et₃N was added directly to the cold reaction

mixture. The ﬂask was then allowed to warm to rt, whereupon the

white crystals slowly dissolved to aﬀord a homogeneous solution. This

solution was then either (1) poured into a separatory funnel

containing 1 M NaOH solution, shaken well, extracted with

dichloromethane, dried, and concentrated or (2) directly concentrated.

1

Subsequently, H NMR spectra of the crude residue were recorded.

Products were puriﬁed by silica ﬂash chromatography, and analytically

pure samples were obtained by recrystallization or distillation as

speciﬁed.

Preparation of (2S,3R)-2-Phenyl-3-(phenylthio)chromane (4a).

Following general procedure 2, 3a (210 mg, 1.0 mmol) was weighed

into a dried 10 mL Schlenk ﬂask. Subsequently, CH₂Cl₂(7 mL),

electrophile 1a (255 mg, 1.0 mmol, 1.0 equiv) and catalyst (S)-2b (52

mg, 0.1 mmol, 0.1 equiv) were added. The ﬂask was placed in an i-

PrOH bath and cooled to −20 °C (probe). After equilibration (ca. 20

min), MsOH (17 μL, 0.25 mmol, 0.25 equiv) was added directly via

syringe. The solution was allowed to stir for 24 h at constant

temperature during which time phthalimide precipitated. Upon

1

consumption of the starting material (TLC, H NMR), the reaction

was quenched with triethylamine (300 μL) and allowed to warm to rt,

whereupon the white solid dissolved. The solution was transferred to a

60 mL separatory funnel and then was diluted with CH₂Cl₂(10 mL)

and 1 M NaOH (15 mL). The phases were separated, and the aqueous

layer was extracted with 15 mL of CH₂Cl₂. The organic phases were

combined, dried over MgSO₄, ﬁltered, and concentrated by rotary

evaporation (30 °C, 3 mmHg). Puriﬁcation by silica gel ﬂash column

chromatography (40:1 hexanes/ethyl acetate, 20 mm diameter, 16 cm

of SiO₂) followed by recrystallization from hexanes (3 mL) aﬀorded, in

two crops, 268 mg (84%) of 4a as white needles. Data for 4a: mp

1

(76%). Data for 3s: bp 100 °C (ABT, 0.05 mmHg); H NMR: (500

MHz, CDCl₃) δ 7.15−7.05 (m, 2H, HC(3), HC(5)), 6.88 (t, J = 7.6

Hz, 1H, HC(4)), 6.77 (d, J = 8.0 Hz, 1H, HC(6)), 5.62−5.38 (m, 2H,

HC(9), HC(10)), 3.45−3.32 (m, 5H, HC(13), HC(14)), 2.68 (dd, J =

8.6, 6.7 Hz, 2H, HC(7)), 2.32 (app q, J = 7.8 Hz, 2H, HC(8)), 2.11−

2.01 (app q, J = 7.7 Hz, 2H, HC(11)), 1.64 (app p, J = 7.2, 6.8 Hz, 2H,

HC(12)); ¹³C NMR (126 MHz, CDCl₃) δ 153.8 (C1), 130.7 (C9),

130.5 (C5), 130.3 (C10), 128.2 (C2), 127.4 (C3), 120.9 (C4), 115.6

(C6), 72.4 (C13), 58.7 (C14), 33.0 (C8), 30.5 (C7), 29.5 (C11), 29.3

(C12); IR (ATR, cm⁻¹) 3308 (br), 2930 (w), 2852 (w), 1607 (w),

1593 (w), 1504 (w), 1489 (w), 1455 (s), 1353 (w), 1234 (m), 1179

(m), 1099 (m), 1042 (w), 968 (m), 931 (w), 847 (w), 750 (s); MS

(EI, 70 eV, m/z) 220 (17, M⁺), 149 (15), 107 (100), 81 (35); TLC R_f

0.33 (4:1 hexanes/ethyl acetate) [UV,CAM]. Anal. Calcd for

C₁₄H₂₀O₂(220.31): C, 76.33; H, 9.15. Found: C, 76.10; H, 8.91.

Sulfenocyclizations of Substituted (E)-2-Cinnamylphenols.

General Procedure 2. To a 10 mL, oven-dried ﬂask equipped with a

magnetic stir bar under an argon atmosphere were added substrate and

CH₂Cl₂. The catalyst and the electrophile were added as solids and

allowed to dissolve to obtain a clear or pale yellow solution. The

reaction vessel was placed in an i-PrOH bath kept at constant

temperature by means of a Neslab IBC-4A Cryocool with probe. The

reaction mixture was cooled to the appropriate reaction temperature,

and the internal temperature was checked. After the temperature

stabilized to 2 °C, MsOH was added to the stirring reaction mixture

via syringe. (Note: It is important to not let the acid touch the walls of

the reaction vessel as it may immediately freeze.) After addition of the

acid, rapid formation of a yellow color was observed. The reaction

mixture was then stirred for the appropriate time. Over the course of

the reaction, white crystals of phthalimide precipitate out of the

1

121−122 °C (hexanes); H NMR (500 MHz, CDCl₃) δ 7.41−7.32

(m, 5 H, HC(aryl)), 7.30 (m, 2H, HC(aryl)), 7.26 (m, 3 H,

HC(aryl)), 7.20 (t, J = 7.8 Hz, 1H, HC(7)), 7.08 (d, J = 7.5 Hz, 1H,

HC(5)), 6.97 (d, J = 8 Hz, 1H, HC(8)), 6.94 (t, J = 8 Hz, 1H,

HC(6)), 5.09 (d, J = 7.8 Hz, 1H, HC(2)), 3.81 (ddd, J = 9.5, 7.7, 5.2

Hz, 1H, HC(3)), 3.15 (dd, J = 16.6, 5.1 Hz, 1H, HC(4)), 2.99 (dd, J =

16.6, 9.1 Hz, 1H, HC(4)); ¹³C NMR (126 MHz, CDCl₃) δ 154.3

(C9), 139.4 (C15), 133.7 (C11), 133.3 (C17), 129.5 (C5), 129.1

(C13), 128.6 (C12), 128.6 (C14), 128.1 (C7), 127.7 (C18), 127.3

(C16), 121.0 (C6), 120.7 (C10), 116.8 (C8), 81.2 (C2), 47.2 (C3),

31.8 (C4); IR (ATR, cm⁻¹) 2912 (w), 1583 (w), 1487 (m), 1449 (w),

1439 (w), 1427 (w), 1306 (w), 1277 (w), 1222 (m), 1205 (w), 1193

(w), 1150 (w), 1111 (w), 1086 (w), 1069 (w), 1024 (w), 1000 (m),

969 (w), 904 (w), 851 (w), 831 (w), 797 (w), 757 (s), 750 (s), 727

(w), 716 (m), 700 (s); MS (EI, 70 eV, m/z) 318 (36, M⁺), 208 (40),

200 (100), 199 (98), 119 (19); TLC R_f0.48 (1:1, hexanes/CH₂Cl₂)

[UV,CAM]; [α]_D²³= −40.0 (c = 0.87, CHCl₃); CD, (−), Cotton sign,

230−280 nm; CSP-SFC, (2R,3S)-4a t_min13.4 min, (5.1%), (2S,3R)-4a

t_maj18.5 min, (94.9%) (Chiralpak AD, 220 nm, 200 bar, 40 °C, 95:5,

sCO₂/MeOH, 2 mL/min). Anal. Calcd for C₂₁H₁₈OS (318.43): C,

79.21; H, 5.70. Found: C, 78.91; H, 5.62.

T

DOI: 10.1021/acs.joc.7b00295

J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry

Article

reaction was quenched with triethylamine (300 μL), and the mixture

was allowed to warm to rt, whereupon the white solid dissolved. The

solution was transferred to a 60 mL separatory funnel and diluted with

CH₂Cl₂(10 mL) and 1 M NaOH (15 mL). The phases were

separated, and the aqueous layer was extracted with 15 mL of CH₂Cl₂.

The organic phases were combined, dried over MgSO₄, ﬁltered, and

concentrated by rotary evaporation (30 °C, 3 mmHg). Puriﬁcation by

silica gel ﬂash column chromatography (60:1 hexanes/ethyl acetate

then 40:1 hexanes/ethyl acetate, 30 mm diameter, 16 cm of SiO₂)

aﬀorded a white solid. Recrystallization from hexanes (3 mL) aﬀorded,

in two crops, 259 mg (78%) of 4c as white needles. Data for 4c: mp

Preparation of (2S,3R)-6-Methyl-2-phenyl-3-(phenylthio)-

chromane (4b). Following general procedure 2, 3b (224 mg, 1.0

mmol) was weighed into a dried 10 mL Schlenk ﬂask. Subsequently,

CH₂Cl₂(7 mL), electrophile 1a (255 mg, 1.0 mmol, 1.0 equiv), and

catalyst (S)-2b (52 mg, 0.1 mmol, 0.1 equiv) were added. The ﬂask

was placed in an i-PrOH bath and cooled to −20 °C (probe). After

equilibration (ca. 20 min), MsOH (17 μL, 0.25 mmol, 0.25 equiv) was

added directly via syringe. The solution was allowed to stir for 24 h at

constant temperature during which time phthalimide precipitated.

Upon consumption of the starting material (TLC, ¹H NMR), the

reaction was quenched with triethylamine (300 μL), and the mixture

was then allowed to warm to rt, whereupon the white solid dissolved.

The solution was transferred to a 60 mL separatory funnel and then

was diluted with CH₂Cl₂(10 mL) and 1 M NaOH (15 mL). The

phases were separated, and the aqueous layer was extracted with 15

mL of CH₂Cl₂. The organic phases were combined, dried over MgSO₄,

ﬁltered, and concentrated by rotary evaporation (30 °C, 3 mmHg).

Puriﬁcation by silica gel ﬂash column chromatography (4:1 hexanes/

CHCl₃to 3:1 hexanes/CHCl₃, 20 mm diameter, 20 cm of SiO₂)

followed by recrystallization from hexanes (3 mL) aﬀorded, in two

crops, 272 mg (82%) of 4b as white needles. Data for 4b: mp 114−

115 °C (hexanes); ¹H NMR (500 MHz, CDCl₃) δ 7.40−7.29 (m, 7H,

HC(aryl)), 7.27−7.23 (m, 3H, HC(aryl))), 7.00 (d, J = 8.5 Hz, 1H,

HC(7)), 6.88, (s, 1H, HC(5)), 6.87 (d, J = 8.5 Hz, 1H, HC(8)), 5.06

(d, J = 7.7 Hz, 1H, HC(2)), 3.80 (ddd, J = 8.8, 7.6, 5.2 Hz, 1H,

HC(3)), 3.09 (dd, J = 16.6, 5.2 Hz, 1H, HC(4)), 2.95 (dd, J = 16.6, 8.9

Hz, 1H, HC(4)), 2.31 (s, 3H, HC(19)); ¹³C NMR (126 MHz,

CDCl₃) δ 152.1 (C9), 139.6 (C15), 133.8 (C11), 133.2 (C17), 130.2

(C6), 129.9 (C5), 129.1 (C13), 128.7 (C7), 128.6 (C12), 128.5

(C14), 127.7 (C18), 127.3 (C16), 120.3 (C10), 116.5 (C8), 81.1

(C2), 47.3 (C3), 31.7 (C4), 20.8 (C19); IR (ATR, cm⁻¹) 2923 (w),

1588 (w), 1497 (m), 1474 (w), 1457 (w), 1436 (w), 1301 (w), 1239

(m), 1227 (s), 1148 (w), 1127 (w), 1027 (w), 988 (m), 937 (w), 909

(w), 880 (w), 852 (w), 831 (m), 818 (w), 798 (w), 745 (s), 729 (w),

691 (s); MS (EI, 70 eV, m/z) 332 (60, M⁺), 200 (88), 199 (80),133

(60),91 (100); TLC R_f0.67 (1:1, hexanes/CH₂Cl₂) [UV,CAM];

[α]_D²³= −32.9 (c = 0.88 in CHCl₃); CD, (−), Cotton sign, 230−280

nm; CSP-SFC, (2R,3S)-4b t_min10.2 min (5.0%), (2S,3R)-4b t_maj14.3

min (95.0%) (Chiralpak AD, 220 nm, 200 bar, 40 °C, 85:15 sCO₂/

MeOH, 2 mL/min). Anal. Calcd for C₂₂H₂₀OS (332.46): C, 79.48; H,

6.06. Found: C, 79.15; H, 5.97.

1

89−90 °C (hexanes); H NMR (500 MHz, CDCl₃) δ 7.38−7.30 (m,

5H, HC(aryl)), 7.28 (d, J = 2.1 Hz, 2H, HC(aryl)), 7.25−7.19 (m, 3H,

HC(aryl)), 7.05 (d, J = 7.5 Hz, 1H, HC(5)), 6.88 (dd, J = 7.7, 1.9 Hz,

1H, HC(7)), 6.81 (t, J = 7.4 Hz, 1H, HC(6)), 5.11 (d, J = 7.6 Hz, 1H,

HC(2)), 3.75 (ddd, J = 8.9, 7.5, 5.0 Hz, 1H, HC(3)), 3.08 (dd, J =

16.6, 5.1 Hz, 1H, HC(4)), 2.95 (dd, J = 16.5, 8.8 Hz, 1H, HC(4)),

2.23 (s, 3H, HC(19)); ¹³C NMR (101 MHz, CDCl₃) δ 152.4 (C9),

139.9 (C15), 133.9 (C11), 133.0 (C17), 129.2 (C5), 129.2 (C16),

128.6 (C13), 128.4 (C18), 127.6 (C14), 127.1 (C7, C16), 126.0

(C10), 120.4 (C6), 120.1 (C8), 81.0 (C2),47.3 (C3), 31.8 (C4), 16.4

(C19); IR (ATR, cm⁻¹) 1594 (w), 1467 (m), 1432 (w), 1379 (w),

1304 (w), 1259 (w), 1239 (w), 1204 (s), 1101 (w), 1072 (w), 1026

(w), 985 (m), 959 (w), 924 (w), 910 (w), 828 (w), 798 (w), 757 (s),

744 (s), 728 (m), 699 (s); MS (EI, 70 eV, m/z) 332 (43, M⁺), 222

(48), 200 (100), 199 (60), 149 (20),133 (41); TLC R_f0.77 (1:1,

23

hexanes/CH₂Cl₂) [UV,CAM]; [α]_D= −4.5 (c = 0.88 in CHCl₃);

CD, (−), Cotton sign, 230−280 nm; CSP-SFC, (2R,3S)-4c t_min10.5

(4.0%), (2S,3R)-4c t_maj11.8 (96.0%) (Chiralpak AD, 220 nm, 200 bar,

40 °C, 95:5, sCO₂/MeOH, 2 mL/min). Anal. Calcd for C₂₂H₂₀OS

(332.46): C, 79.48; H, 6.06. Found: C, 79.54; H, 6.19.

Preparation of (2R,3S)-3-Phenyl-2-(phenylthio)-2,3-dihydro-1H-

benzo[f ]chromene (4d). Following general procedure 2, 3d (260 mg,

1.0 mmol) was weighed into a dried 10 mL Schlenk ﬂask.

Subsequently, CH₂Cl₂(7 mL), electrophile 1a (255 mg, 1.0 mmol,

1.0 equiv), and catalyst (S)-2b (52 mg, 0.1 mmol, 0.1 equiv) were

added. The ﬂask was placed in an i-PrOH bath and cooled to −20 °C

(probe). After equilibration (ca. 20 min), MsOH (17 μL, 0.25 mmol,

0.25 equiv) was added directly via syringe. The solution was allowed to

stir for 24 h at constant temperature during which time phthalimide

1

precipitated. Upon consumption of the starting material (TLC, H

NMR), the reaction was quenched with triethylamine (300 μL) and

then was allowed to warm to rt, whereupon the white solid dissolved.

The solution was transferred to a 60 mL separatory funnel and diluted

with CH₂Cl₂(10 mL) and 1 M NaOH (15 mL). The phases were

separated, and the aqueous layer was extracted with 15 mL of CH₂Cl₂.

The organic phases were combined, dried over MgSO₄, ﬁltered, and

concentrated by rotary evaporation (30 °C, 3 mmHg). Puriﬁcation by

silica gel ﬂash column chromatography (60:1 hexanes/ethyl acetate

then 40:1 hexanes/ethyl acetate, 20 mm diameter, 16 cm of SiO₂)

followed by recrystallization (hexanes, 3 mL, ether, 0.2 mL) aﬀorded,

in two crops, 290 mg (79%) of 4d as pale pink prisms. Data for 4d: mp

136−138 °C (hexanes/ether); ¹H NMR (500 MHz, CDCl₃) δ 7.83 (d,

J = 7.8 Hz, 1H, HC(5)), 7.77 (d, J = 8.5, 1H, HC(8)), 7.72 (d, J = 8.9

Hz, 1H, HC(9)), 7.53 (dd, J = 8.3, 7.0 Hz, 1H, HC(7)), 7.46−7.38

(m, 4H, HC(aryl), HC(6)), 7.38−7.27 (m, 4H, HC(aryl)), 7.27−7.22

(m, 3H HC(aryl)), 7.19 (d, 1H, 8.9 Hz, HC(10)), 5.15 (d, J = 8.0 Hz,

1H, HC(2)), 3.95 (ddd, J = 8.9, 8.0, 5.6, 1H, HC(3)), 3.49 (dd, J =

16.8, 5.6 Hz, 1H, HC(4)), 3.26 (dd, J = 16.8, 8.9 Hz, 1H, HC(4)); ¹³C

NMR (101 MHz, CDCl₃) δ 152.0 (C11), 139.1 (C19), 133.9 (C14),

Preparation of (2S,3R)-8-Methyl-2-phenyl-3-(phenylthio)-

chromane (4c). Following general procedure 2, 3c (224 mg, 1.0