Unusual Kinetic Profiles for Lewis Base-Catalyzed Sulfenocyclization of ortho -Geranylphenols in Hexafluoroisopropyl Alcohol

Article abstract of DOI:10.1055/s-0039-1690111

The kinetic behavior of the Lewis base-catalyzed sulfenocyclization of polyenes in hexafluoroisopropyl alcohol (HFIP) was explored. The rate of reaction is not dependent on the electronic properties of the terminal nucleophile, suggesting that this captur

Full text of DOI:10.1055/s-0039-1690111

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2019, 30, A–F
letter
A
en
K. A. Robb et al.
Letter
Synlett
Unusual Kinetic Profiles for Lewis Base-Catalyzed Sulfenocycliza-
tion of ortho-Geranylphenols in Hexafluoroisopropyl Alcohol
Kevin A. Robb
0
0
0
0-0
0
0
3-
4
3
3
1-0
6
4
2
F
F
sulfenylating agent
HO
Soumitra V. Athavale
i-Pr
O
Scott E. Denmark*
0
0
0
0-
0
0
0
2-1
0
9
9-9
7
6
5
S
R
R
LB* catalyst
H
i-Pr
Roger Adams Laboratory, Department of Chemistry,
University of Illinois, Urbana, IL 61801, USA
sdenmark@illinois.edu
HFIP, 25 °C
R = OMe, H, Cl, CN
Rate unaffected by R
Unusual kinetic behavior!
~0.5 order in substrate
~0.5 order in sulf. agent
1st order in catalyst
Published as part of the Cluster Organosulfur and
Organoselenium Compounds in Catalysis
Me
O
i-Pr
N
N
Se
P
N
S
N(i-Pr)2
Me
i-Pr
O
Received: 11.05.2019
Accepted after revision: 17.06.2019
Published online: 08.07.2019
spective co-workers require specifically engineered sub-
strates that are not amenable to easy postcyclization func-
tionalization.
DOI: 10.1055/s-0039-1690111; Art ID: st-2019-w0261-l
A recent disclosure from this laboratory reported a Lew-
is base-catalyzed enantioselective sulfenocyclization of
polyenes, enabled by the use of 1,1,1,3,3,3-hexafluoroiso-
propyl alcohol (HFIP) as the reaction solvent (Scheme 1).11
Electron-rich homogeranylarenes, as well as electronically
diverse ortho-geranylphenols 1, are competent substrates
for this transformation, affording tricyclic products 2 in
good yields and high enantioselectivities with an opera-
tionally simple protocol. The resulting thioethers can be
subjected to a range of transformations to install useful A-
ring functionality.
Abstract The kinetic behavior of the Lewis base-catalyzed sulfenocy-
clization of polyenes in hexafluoroisopropyl alcohol (HFIP) was ex-
plored. The rate of reaction is not dependent on the electronic proper-
ties of the terminal nucleophile, suggesting that this capture step is not
rate limiting. Additionally, fractional orders were observed for two of
the reaction components. This intriguing profile appears unique to the
polyene sulfenocyclization reaction and is not merely due to solvent ef-
fects.
Key words polyenes, cyclization, sulfenocyclization, kinetics, Lewis
base catalysis, hexafluoroisopropyl alcohol
Polyene cyclizations constitute key biosynthetic path-
ways that rapidly generate structurally complex metabo-
lites from linear olefin precursors. These fascinating trans-
formations, which can construct multiple rings and stereo-
genic centers from simple achiral starting materials in a
single step, have inspired the development of nonenzymat-
ic approaches for the synthesis of polycyclic terpenoids.
Polyene cyclizations have been used as strategy-level dis-
connections in the laboratory syntheses of numerous natu-
ral products.1 Early investigations of enantioselective vari-
ants of polyene cyclizations involved high catalyst loadings
and stoichiometric amounts of chiral Lewis or Brønsted ac-
ids.2 Examples of truly catalytic enantioselective cycliza-
tions have been reported only recently. In 2017, Samanta
and Yamamoto disclosed an enantioselective bromocycliza-
tion of polyenes with a BINOL-derived chiral Lewis basic
catalyst.3 Organometallic catalysts have also been employed
by the groups of Gagne (Pt),4 Toste (Au),5 Carreira (Ir),6 and
Snyder (Hg, stoichiometric).7 Halides, olefins, or vinyl
groups in the products can be used for further functional
elaboration. Highly selective organocatalytic methods in-
troduced by Jacobsen,8 MacMillan,9 and Zhao10 and their re-
sulf. agent 3 (1.01 equiv)
O
LB* cat. 4 (0.01 equiv)
HO
S
R
H
HFIP (0.1 M), 25 °C, 12 h
R
1
2
R = EDG or EWG, 8 examples
67–80% yield, up to 93:7 e.r.
EDG = electron-donating group; EWG = electron-withdrawing group
Scheme 1 Chiral Lewis base-catalyzed enantioselective sulfenocycliza-
tion of polyenes in HFIP
The mechanism for the Lewis base-catalyzed sulfeno-
functionalization of olefins has been extensively studied in
these laboratories, and the catalytic cycle shown in Scheme
2 has been proposed.12,13 Initially, acid-mediated transfer of
a sulfenyl group from 3 to 4 generates a cationic donor–ac-
ceptor complex i. This highly electrophilic complex reacts
with a nonactivated olefin 1 to generate an enantiomerical-
ly enriched thiiranium-ion intermediate ii. This thiiranium
© 2019. Thieme. All rights reserved. — Synlett 2019, 30, A–F
Georg Thieme Verlag KG, Rüdigerstraße 14, 70469 Stuttgart, Germany

B
K. A. Robb et al.
Letter
Synlett
ion is opened by nucleophilic attack, generating anti-func-
tionalized products. In the present case, the generation of
the thiiranium ion serves as the initiating event for a cat-
ionic polyene cascade cyclization that is ultimately termi-
nated by the pendent phenol nucleophile with formation of
the tricyclic product 2.
Secondly, it has a positive fractional order in mesic acid.
These data are consistent with a mechanism in which sulfe-
nyl-group transfer from the donor–acceptor complex i to
the olefin is the turnover-limiting step. Although this
mechanism is generally presumed to be operative, there
was reasonable suspicion that this might not be the case in
the aforementioned polyene cyclization for two reasons.
First, the reaction solvent is HFIP rather than DCM. This po-
lar, protic solvent can participate in hydrogen-bonding and
fluorophobic interactions with all of the reaction compo-
nents, which might affect the kinetic profile. Secondly, the
nature of the thiiranium-ion opening step is quite different
from that in previous systems because the transfer of elec-
tron density is propagated over the entire molecule as part
of a cationic cascade process. Over forty years ago, Johnson
observed a pronounced dependence of the rate of acid-me-
diated cyclization of epoxypolyenes on the electronic prop-
erties of the terminal arene nucleophile.14 Specifically, fast-
er reaction rates were observed for electron-rich arenes
than for electron-deficient ones ( < 0), even though these
motifs were located far from the site of cascade initiation, in
which case the electronic perturbation would be transmit-
ted through the alkene double bonds. It was hypothesized
that a similar phenomenon might be operating in the pres-
ent system (i.e. the rate-determining step has switched
from thiiranium generation to thiiranium opening), in
which case a rate dependence on the electronic character of
the terminal nucleophile would be expected.
CF3
3
F3C
O
O
OH
N
i-Pr
i-Pr
N
S
S
O
i-Pr
i-Pr
O
O
NH
F3C
OH
CF3
O
Me
N
Se
P
N(i-Pr)2
i-Pr
N
Me
Me
N
N
Se
S
P
4
N(i-Pr)2
Me
i-Pr
2
CF3
i
F3C
O
CF3
1
F3C
R
O
R
HO
SAryl
SAryl
H
HO
H
To test this hypothesis, the following experiments were
carried out (Scheme 3). A series of ortho-geranylated phe-
nols 1a–d bearing electronically diverse para-substituents
were subjected to the standard reaction conditions. Addi-
tionally, all substrates were ortho-fluorinated, so that reac-
tion conversion could be monitored in real time by 19F NMR
spectroscopy. A routine solvent-suppression protocol was
employed to decrease the intensity of the HFIP 19F reso-
nance ( = –77.9 ppm), which permitted accurate inte-
gration of the 19F resonances corresponding to 1 and 2
ii
Scheme 2 Proposed catalytic cycle
In previous mechanistic studies on Lewis base-cata-
lyzed, intramolecular oxysulfenylation,12 with dichloro-
methane (DCM) as the solvent, the following kinetic profile
was observed. First, the reaction is first order in both cata-
lyst and olefin, and zeroth order in the sulfenylating agent.
Experimental Design
sulf. agent 3
F
O
F
O
LB* cat. 4
HO
N
S
S
R
H
R
O
[1,2-difluorobenzene]
HFIP, 25 °C
sulf. agent 3
σpara
Me
1a
1b
1c
1d
R = OMe
R = H
–0.27
2a
R = OMe
R = H
N
N
Se
0
2b
2c
2d
P
N(i-Pr)2
R = Cl
+0.24
+0.70
R = Cl
Me
R = CN
R = CN
catalyst 4
Scheme 3 Substrates employed for kinetic studies of the sulfenocyclization reaction
© 2019. Thieme. All rights reserved. — Synlett 2019, 30, A–F

C
K. A. Robb et al.
Letter
Synlett
( = –143.0 to –141.0 ppm). Comparison of the reaction
rates across the series would provide valuable insight into
the turnover-limiting step.
ed in situ and subsequent trapping with geranyl bromide
afforded the desired alkylation products 7a–c in modest but
synthetically useful yields. Most importantly, the isolated
products were isomerically pure. Finally, the carbamate
was readily hydrolyzed under basic aqueous conditions to
afford phenols 1a–c.18
Synthesis of the geranylated fluorophenols proved to be
a formidable challenge. Initially, it was envisioned that 1a–d
might be obtained in one step from commercially available
fluorophenols 5 by using previously developed conditions
for C-selective phenol alkylation.15 Unfortunately, in the
case of ortho-fluorinated phenols, this alkylation protocol
afforded complex product mixtures and low yields of 1. Ad-
ditionally, chromatographic separation of pure 1 from the
various reaction byproducts proved difficult. Alternative
strategies were explored for the clean, selective alkylation
of 5 (Scheme 4). Kauch and Hoppe previously developed a
protocol for ortho-lithiation of ortho-fluorophenols that
employs an N-isopropyl carbamate as a directing group.16,17
The aryllithium can be trapped by diverse electrophiles in
high yields. Although allylic halides were not included in
the demonstrated scope, this route appeared to be promis-
ing for installation of a geranyl side chain. Preparation of
carbamates 6a–d from the corresponding phenols was triv-
ial. Directed lithiation of the N-silylated carbamate generat-
This synthetic sequence was unfortunately not appro-
priate for the preparation of 1d, as the nitrile was suscepti-
ble to nucleophilic addition of butyllithium during the di-
rected lithiation step. In the interest of retaining the same
general synthetic strategy, the substitution of lithium bases
with less-nucleophilic magnesium amide bases was investi-
gated (Scheme 5). Knochel has reported efficient methods
for directed ortho-magnesiation of electron-deficient
arenes, including those bearing fluorine atoms and nitrile
groups. First, the directing group ability of N,N,N′,N′-te-
tramethylphosphorodiamidate19 was investigated with
both monobasic and dibasic magnesium tetramethylpiperi-
dide reagents. Treatment of 8 with dibasic (tmp)2Mg·2LiCl
complex, followed by transmetalation and trapping with
geranyl bromide, resulted in a dialkylated arene as the only
isolable product. Encouragingly, the nitrile was untouched
i-PrHN
O
F
i-PrHN
O
F
F
HO
F
O
a
b
c
HO
O
R
R
R
R
5a
R = OMe
R = H
6a
6b
6c
6d
78%
7a
7b
7c
7d
39%
40%
17%
–
1a
1b
1c
1d
87%
5b
5c
5d
78%
82%
71%
88%
84%
–
R = Cl
R = CN
Scheme 4 Synthesis of geranylated fluorophenols. Reaction conditions: (a) i-PrNCO (1.1 equiv), DMAP (0.05 equiv), THF, 60 °C; (b) TMEDA (1.1 equiv),
TMSOTf (1.05 equiv), Et2O, 25 °C, then TMEDA (2.0 equiv), n-BuLi (2.0 equiv), –78 °C, then geranyl bromide (1.25 equiv), –78 °C; (c) aq NaOH (2.5
equiv), EtOH, 25 °C. All values are isolated yields after recrystallization (6), chromatography (7), or distillation (1).
NMe2
NMe2
F
Me2N
P
O F
HO
Me2N
P
O F
a
b
c
O
O
84%
53%
57%
CN
CN
CN
F
8
9
1d
HO
CN
t-BuO
O
t-BuO
O
5d
F
F
d
b
O
O
94%
27%
CN
CN
11
10
Scheme 5 Synthesis of nitrile 1d. Reaction conditions: (a) ClP(O)(NMe2)2 (1.2 equiv), Et3N (1.2 equiv), DMAP (0.1 equiv), THF, 25 °C; (b) tmpMgCl·LiCl
(1.1 equiv), THF, 0 °C, then ZnCl2 (1.2 equiv), THF, –40 °C, then CuCN·2LiCl (0.5 equiv), geranyl bromide (1.5 equiv), THF, –40 °C to 25 °C; (c) HCO2H–
EtOH–H2O (1:9:1) (0.9 M), MW, 140 °C; (d) Boc2O (1.5 equiv), Et3N (2.2 equiv), DMAP (0.05 equiv), DCM, 25 °C.
© 2019. Thieme. All rights reserved. — Synlett 2019, 30, A–F

D
K. A. Robb et al.
Letter
Synlett
under these reaction conditions. To prevent overmetalation,
the monobasic (tmp)MgCl·LiCl complex was substituted for
the dibasic reagent. Gratifyingly, this led to the formation of
the desired ortho-alkylation product 9 in good yield. The
phosphorodiamidate moiety is crucial for directing magne-
siation to the desired position. When compound 10 was
treated with (tmp)MgCl·LiCl under identical conditions,
magnesiation occurred at the most acidic position, leading
to the undesired isomer 11, even though tert-butyl carbon-
ate is known to be an effective directing group for magne-
sium amides in other aromatic systems.20 Removal of the
directing group was accomplished by microwave-assisted
acidic hydrolysis to afford phenol 1d.
With all of the desired substrates 1a–d in hand, the ki-
netics experiments outlined in Scheme 3 were carried out.
To obtain the order in each reaction component, the load-
ings of catalyst (S)-4, sulfenylating agent 3, and substrate 1
were varied from run to run, and the data were processed
according to the variable time normalization analysis (VT-
NA) method described by Burés.21–23 The VTNA semiquanti-
tative data treatment permits the user to extract more in-
formation from fewer experiments, compared with classi-
cal methods, at the cost of slightly diminished accuracy (e.g.
the treatment can easily differentiate between reaction or-
ders of 2.0, 1.0, and 0.5, but not, perhaps, between 1.1, 1.0,
and 0.9). As an example, for the conversion of 1c into 2c, the
time-normalized rate plots from four different experiments
(run at variable concentrations of each reactant, Exp1
through Exp4) only overlaid when the exponent terms
within the time integral are equal to 0.5, 0.5 and 1.0 (Figure
1). Nearly identical behavior was likewise observed for all
other substrates 1 (Table 1).
Figure 1 VTNA kinetic analysis of the sulfenocyclization reaction of 1c
in HFIP by analysis of four different experiments (exp 1 through exp 4).
Values in the graph legend denote concentrations in molarity units.
lier kinetic studies performed in this laboratory for related
systems12 was seen. For the sulfenocyclization of 1 to 2, the
reaction was observed to be first order in catalyst (S)-4 and
fractional order in both the substrate 1 (~0.5 order) and the
sulfenylating agent 3 (also ~0.5 order). Similar fractional or-
ders were obtained for all substrates, regardless of the elec-
tronic nature of the phenol (Table 1; see also the Supple-
mentary Information). A catalyst order of 1.0 is consistent
with sulfenyl-group transfer (thiiranium ion formation) as
the rate-determining step, and is also consistent with previ-
ous mechanistic studies. The fractional orders observed for
both 1 and 3 were unexpected, and are more difficult to ex-
plain. In particular, the presence of any nonzero order for
sulfenylating agent 3 is puzzling, because the catalyst (S)-4
is presumed to be saturated at all times (i.e., donor–accep-
tor complex i is presumed to be the resting state of the cata-
lyst prior to the rate-determining step). The results indicate
that, at least for the present system, the concentration of 3
does influence the rate of reaction, although the nature and
origin of this influence remains unclear and is a topic of ac-
tive study. The observation of a fractional order for sub-
strate 1 was also surprising, as an order of 1.0 is expected
for a rate-determining step that involves sulfenyl-group
transfer between complex i and one molecule of 1. Interest-
ingly, similar fractional orders were obtained in the arene-
terminated cyclization of the nonphenolic substrate 12a to
13a (first order in (S)-4, 0.7 order in 3, and 0.5 order in 12a)
(Scheme 6). This observation rules out the possibility that
The results of these experiments were quite surprising;
a drastically different kinetic profile as compared with ear-
Table 1 Results of VTNA Kinetic Analyses for the Reaction of Sub-
strates 1a–d
Substrate
Rate equation
kobs
1a (R = OMe)
1b (R = H)
1c (R = Cl)
1d (R = CN)
k
k
k
k
obs[(S)-4]1[1a]0.6[3]0.5
obs[(S)-4]1[1a]0.5[3]0.4
obs[(S)-4]1[1a]0.5[3]0.5
obs[(S)-4]1[1a]0.5[3]0.6
0.062 ± 0.004
0.051 ± 0.003
0.055 ± 0.001
0.075 ± 0.001
sulf. agent 3
Reaction Kinetics by VTNA
OMe
OMe
F
LB* cat. (S)-4
1st order in catalyst (S)-4
F
0.5 order in olefin 12a
[1,2-difluorobenzene]
HFIP, 25 °C
ArS
0.7 order in sulf. agent 3
H
12a
13a
Scheme 6 VTNA kinetic analysis of the sulfenylation reaction of 12a in HFIP
© 2019. Thieme. All rights reserved. — Synlett 2019, 30, A–F

E
K. A. Robb et al.
Letter
Synlett
the fractional order in 1 might arise because of the necessi-
ty for dissociation from a phenolic hydrogen-bonded dimer.24
To ascertain whether this intriguing kinetic profile is an
innate property of the polyene sulfenocyclization reaction
or whether it is caused by carrying out the reaction in HFIP,
the previously studied oxysulfenylation reaction12 was per-
formed in HFIP (Scheme 7). The results of this experiment
(first order in catalyst 4, first order in alkene 14, and zeroth
order in sulfenylating agent 3) matched those obtained pre-
viously when the reaction was carried out in CH2Cl2 with
mesic acid. This outcome suggests that the polyene sulfeno-
cyclization indeed displays a unique kinetic behavior, which
clearly warrants further and more-detailed investigation.25
leophilic terminator strongly suggests that the rate-deter-
mining step is not the nucleophilic capture of a thiiranium
ion, consistent with previous mechanistic proposals.
sulf. agent 3
Reaction Kinetics by VTNA
F
F
LB* cat. (S)-4
1st order in catalyst (S)-4
1st order in olefin 14
O
HO
Figure 2 Reaction progression for the sulfenocyclization of 1a–d and 12a
[1,2-difluorobenzene]
HFIP, 25 °C
ArS
15
0th order in sulf. agent 3
To summarize, the kinetic behavior of Lewis base-cata-
lyzed polyene sulfenocyclization reactions in HFIP26
demonstrates the following salient features: (1) the reac-
tion rate shows a fractional-order (~0.5) concentration de-
pendence on both reactants and a first-order concentration
dependence on the catalyst, (2) this kinetic profile is unique
to the polyene substrate and is observed for both carbon-
and oxygen nucleophiles, and (3) the reaction rate is essen-
tially invariant across the substituted phenols tested. The
fractional orders indicate a more complex mechanism than
previously postulated (Scheme 2). One possibility includes
a preequilibrium substrate–aggregate dissociation step. An-
other possibility is a mechanistic sequence in which both
the reactants are involved in elementary steps that promote
the reaction as well as in separate elementary steps that in-
hibit the overall transformation. Although the kinetic and
mechanistic picture is clearly incomplete, these experi-
ments represent an important and necessary step in under-
standing Lewis base-catalyzed transformations, and might
provide further insight into optimization of sulfenocycliza-
tions to include more-diverse polyene substrates. A satis-
factory explanation that justifies the unusual kinetic profile
remains elusive and is the subject of ongoing mechanistic
studies.
14
Scheme 7 VTNA kinetic analysis of the sulfenylation reaction of 14 in HFIP
As to the relative rates of reaction of 1a through 1d, the
value of kobs seemed immune to changes in the electronic
properties of the terminating phenol. From a qualitative as-
sessment of the raw concentration–time data (Figure 2),
one might conclude that the reaction of electron-deficient
1d (R = CN) is marginally slower compared with those of
1a–c. These differences are quite small, however, especially
when compared with the over sixfold rate difference be-
tween electron-rich and electron-deficient substrates origi-
nally reported by Johnson.14 Furthermore, qualitatively, the
rate of C-capture (substrate 12a) appears essentially identi-
cal to the rate of O-capture (substrates 1a–d). However, it is
important to note that a Hammett plot of the rate data from
Johnson’s study indicates two distinct mechanistic regimes.
For the electronically deficient terminating arenes, the rate
was strongly influenced by the electronic character of the
substrate ( = –1.4), whereas for electron-rich terminating
arenes, this dependence was much weaker ( = –0.2). This
implies a potential change in the rate-determining step
from capture (for electron-deficient terminators) to initia-
tion (for electron-rich terminators). In the present case, one
cannot exclude the possibility that all of the phenols 1a–d
are sufficiently electron-rich for all four cyclizations to op-
erate in the latter mechanistic regime. Alternatively, one
also cannot exclude the possibility that 1a–d are too similar
(i.e. the para-substituent exerts little influence on the over-
all electronic character compared with the other three sub-
stituents, which are preserved across the series), which
would result in similar rates of reaction in either mechanis-
tic regime. However, the fact that a comparable reaction
rate was measured for 12a containing a markedly less-nuc-
Funding Information
We are grateful to the National Institutes of Health (GM R35 127010)
for generous financial support.
N
ati
o
n
a
lInstituteof
G
e
n
eralM
e
d
i
c
a
lS
c
i
e
n
c
es(G
M
R
3
5
1
2
7
0
1
0)
Acknowledgment
We thank Dr. Lingyang Zhu (UIUC) for assistance in planning and exe-
cution of the kinetic experiments. We also thank the UIUC SCS sup-
port facilities (microanalysis, mass spectrometry, and NMR
spectroscopy) for their assistance.
© 2019. Thieme. All rights reserved. — Synlett 2019, 30, A–F

F
K. A. Robb et al.
Letter
Synlett
Supporting Information
(24) An additional homogeranylarene substrate 12b (see Supple-
mentary Information) was prepared, but did not convert cleanly
into 13b under the standard reaction conditions, indicating that
the arene of 12b is a poor nucleophile compared with 12a. This
disparity is rationalized by the large difference between the
meta (+0.12) and para (–0.27) Hammett constants for the
methoxy group.
Supporting information for this article is available online at
https://doi.org/10.1055/s-0039-1690111.
S
u
p
p
orit
n
gInformati
o
n
S
u
p
p
orti
n
gInformati
o
n
References and Notes
(25) In addition to nucleophilic ring-opening reactions at carbon,
thiiranium ions are also susceptible to nucleophilic attack at
sulfur. The relative rates of these reactions are dependent on the
substitution patterns of the thiiranium ion; see: Lucchini, V.;
Modena, G.; Pasi, M.; Pasquato, L. J. Org. Chem. 1997, 62, 7018; It
is noted that differences in reactivity probably exist between
thiiranium ions resulting from the trisubstituted alkene 1 (or
12) and the disubstituted alkene 14, although it is unclear how
these differences would impact the overall mechanistic picture.
(26) Hexahydroxanthenes 2a–d; General Procedure
A 50-mL round-bottomed flask equipped with a stirrer bar was
charged with sulfenylating agent 3 (1.01 mmol, 1.01 equiv),
HFIP (10 mL), and substrate 1 (1.0 mmol). Catalyst (S)-4 (0.01
mmol, 0.01 equiv) was added and the mixture was stirred at
25 °C for 12 h. Some white precipitates and/or a color change
were typically observed at longer reaction times. Upon comple-
tion of the reaction [TLC; hexanes–CH2Cl2 (80:20)], the mixture
was diluted with CH2Cl2 (5 mL) and volatile components were
removed by rotary evaporation (30 °C, 15 mm Hg). The crude
product was purified by chromatography [silica gel, hexanes–
CH2Cl2 (gradient elution)] to give a white solid. The product was
triturated in boiling MeOH or EtOH (~1.5 mL) and the mother
liquor was decanted to afford 2 in >99% purity (quantitative 1H
NMR analysis).
(1) Yoder, R. A.; Johnston, J. N. Chem. Rev. 2005, 105, 4730.
(2) Ungarean, C. N.; Southgate, E. H.; Sarlah, D. Org. Biomol. Chem.
2016, 14, 5454.
(3) Samanta, R. C.; Yamamoto, H. J. Am. Chem. Soc. 2017, 139, 1460.
(4) Cochrane, N. A.; Nguyen, H.; Gagne, M. R. J. Am. Chem. Soc. 2013,
135, 628.
(5) Sethofer, S. G.; Mayer, T.; Toste, F. D. J. Am. Chem. Soc. 2010, 132,
8276.
(6) Schafroth, M. A.; Sarlah, D.; Krautwald, S.; Carreira, E. M. J. Am.
Chem. Soc. 2012, 134, 20276.
(7) Snyder, S. A.; Treitler, D. S.; Schall, A. Tetrahedron 2010, 66,
4796.
(8) Knowles, R. R.; Lin, S.; Jacobsen, E. N. J. Am. Chem. Soc. 2010, 132,
5030.
(9) Rendler, S.; MacMillan, D. W. C. J. Am. Chem. Soc. 2010, 132,
5027.
(10) Fan, L.; Han, C.; Li, X.; Yao, J.; Wang, Z.; Yao, C.; Chen, W.; Wang,
T.; Zhao, J. Angew. Chem. Int. Ed. 2018, 57, 2115.
(11) Tao, Z.; Robb, K. A.; Zhao, K.; Denmark, S. E. J. Am. Chem. Soc.
2018, 140, 3569.
(12) Denmark, S. E.; Hartmann, E.; Kornfilt, D. J. P.; Wang, H. Nat.
Chem. 2014, 6, 1056.
(13) Hartmann, E.; Denmark, S. E. Helv. Chim. Acta 2017, 100,
e1700158.
(14) Bartlett, P. A.; Brauman, J. I.; Johnson, W. S.; Volkmann, R. A.
J. Am. Chem. Soc. 1973, 95, 7502.
(2R,4aR,9aR)-2-[(2,6-Diisopropylphenyl)thio]-5-fluoro-7-
methoxy-1,1,4a-trimethyl-2,3,4,4a,9,9a-hexahydro-1H-xan-
thene (2a)
(15) Denmark, S. E.; Kornfilt, D. J. P. J. Org. Chem. 2017, 82, 3192.
(16) Kauch, M.; Hoppe, D. Can. J. Chem. 2001, 79, 1736.
(17) Kauch, M.; Hoppe, D. Synthesis 2006, 1578.
(18) An additional phenol 1e (R = Me) was prepared from 5e (R =
Me) in an analogous fashion. Unfortunately, it was discovered
that the 19F NMR resonances of 1e and product 2e overlapped
when HFIP was used as the reaction solvent. As a result, rate
data could not be obtained for this substrate by using this
experimental setup.
(19) Rohbogner, C. J.; Clososki, G. C.; Knochel, P. Angew. Chem. Int. Ed.
2008, 47, 1503.
(20) Lin, W.; Baron, O.; Knochel, P. Org. Lett. 2006, 8, 5673.
(21) Burés, J. Angew. Chem. Int. Ed. 2016, 55, 2028.
(22) Burés, J. Angew. Chem. Int. Ed. 2016, 55, 16084.
(23) Nielsen, C. D.-T.; Burés, J. Chem. Sci. 2019, 10, 348.
White solid; yield: 355.5 mg (75%); 1H NMR (500 MHz, CDCl3):
 = 7.33 (t, J = 7.7 Hz, 1 H), 7.18 [d, J = 7.7 Hz, 2 H, HC(19)], 6.52
[dd, J = 12.3, 2.7 Hz, 1 H, HC(11)], 6.41 [br s, 1 H, HC(9)], 3.96
(hept, J = 6.7 Hz, 2 H), 3.73 (s, 3 H), 2.77 (dd, J = 16.7, 5.3 Hz, 1
H), 2.75–2.70 (m, 1 H), 2.69 (dd, J = 12.1, 3.9 Hz, 1 H), 1.97 (dt,
J = 12.7, 2.9 Hz, 1 H), 1.76 (dd, J = 12.6, 5.3 Hz, 1 H), 1.74–1.64
(m, 1 H), 1.61 (dq, J = 14.0, 3.7 Hz, 1 H), 1.48 (td, J = 13.3, 3.6 Hz,
1 H), 1.41 (s, 3 H), 1.26 (d, J = 6.8 Hz, 6 H), 1.23 (s, 3 H), 1.20 (d,
J = 6.9 Hz, 6 H), 1.08 (s, 3 H). 13C NMR (126 MHz, CDCl3):  =
154.1, 152.5 (d, JC–F = 10.0 Hz), 152.0 (d, JC–F = 244.4 Hz), 135.3
(d, JC–F = 11.4 Hz), 130.3, 129.2, 124.9 (d, JC–F = 3.2 Hz), 123.9,
109.0 (d, JC–F = 3.0 Hz), 101.3 (d, JC–F = 21.8 Hz), 77.0, 60.9, 55.9,
49.6, 39.9, 38.7, 31.5, 28.9, 26.7, 25.0, 24.1, 23.9 (d, JC–F = 2.7 Hz),
19.7, 16.6.
© 2019. Thieme. All rights reserved. — Synlett 2019, 30, A–F