Enantioselective Heck Arylation of Acyclic Alkenol Aryl Ethers: Synthetic Applications and DFT Investigation of the Stereoselectivity

Article abstract of DOI:10.1002/adsc.201901471

Herein we report the enantioselective Heck-Matsuda arylation of acyclic E and Z-alkenyl aryl ethers. The reactions were carried out under mild conditions affording the enantioenriched benzyl ethers in a regioselective manner, moderate to good yields (up to 73%), and in good to excellent enantiomeric ratios (up to 97:3). The enantioselective Heck-Matsuda arylation has shown a broad scope (25 examples), and some key Heck-Matsuda adducts were further converted into more complex and valuable scaffolds including their synthetic application in the synthesis of (R)-Fluoxetine, (R)-Atomoxetine, and in the synthesis of an enantioenriched benzo[c]chromene. Finally, in silico mechanistic investigations into the reaction's enantioselectivity were performed using density functional theory. (Figure presented.).

Full text of DOI:10.1002/adsc.201901471

Title: Enantioselective Heck Arylation of Acyclic Alkenol Aryl
Ethers: Synthetic Applications and DFT Investigation of the
Stereoselectivity
Authors: Carlos Roque Correia, Ellen Polo, Martí Wang, Ricardo
Angnes, and Ataualpa Braga
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Adv. Synth. Catal. 10.1002/adsc.201901471
Link to VoR: http://dx.doi.org/10.1002/adsc.201901471

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
Very Important Publication
FULL PAPER
DOI: 10.1002/adsc.201((will be filled in by the editorial staff))
Enantioselective Heck Arylation of Acyclic Alkenol Aryl
Ethers: Synthetic Applications and DFT Investigation of the
Stereoselectivity
Ellen Christine Polo,ǂ,a Martí Fernández Wang,ǂ,a Ricardo Almir Angnes,b Ataualpa A.
C. Braga,b and Carlos Roque Duarte Correia*a
a
Departamento de Química Orgânica, Instituto de Química, Universidade Estadual de Campinas, Rua Josué de Castro, s/n, 13083-970,
Campinas, São Paulo (Brazil). E-mail: croque@unicamp.br
Departamento de Química Fundamental, Instituto de Química, Universidade de São Paulo, Avenida Lineu Prestes, 748, 05508-000,
b
São Paulo, São Paulo (Brazil)
These authors contributed equally to this work.
ǂ
Received: ((will be filled in by the editorial staff))
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201######.((Please
delete if not appropriate))
Abstract. Herein we report the enantioselective Heck- synthesis of (R)-Fluoxetine, (R)-Atomoxetine, and in the
Matsuda arylation of acyclic E and Z-alkenyl aryl ethers. The synthesis of an enantioenriched benzo[c]chromene. Finally,
reactions were carried out under mild conditions affording in silico mechanistic investigations into the reaction’s
the enantioenriched benzyl ethers in a regioselective manner, enantioselectivity were performed using density functional
moderate to good yields (up to 73%), and in good to theory.
excellent enantiomeric ratios (up to 97:3). The
enantioselective Heck-Matsuda arylation has shown a broad
scope (25 examples), and some key Heck-Matsuda adducts
were further converted into more complex and valuable
scaffolds including their synthetic application in the
Keywords: Asymmetric catalysis; Heck reaction; N,N-
ligand; Homogeneous catalysis; Density functional
calculations
arylations on these systems rely basically on the use
of BPMO ligands both with aryl triflates and
halides,[33,34] aryl boronic acids and N,N ligands,[35] or,
to a lesser extent, diazonium salts and N,N-
ligands.[36,37] Recently, an efficient palladium-
catalyzed enantioselective Heck alkenylation of
acyclic aryl enol ethers using alkenyl triflates was
reported by Sigman and coworkers (Scheme 1a).[38]
Despite holding considerable synthetic potential, to
the best of our knowledge, the enantioselective Heck
arylation of similar substrates has been attempted
only recently, leading to modest results.[39] Herein,
we report our investigations in this direction using
arenediazonium salts (Heck-Matsuda reaction)[40–42]
and chiral N,N-ligands. Critical to the success of this
endeavor was the use of methanol both as a reaction
solvent and protecting agent of the β-aryloxy
aldehydes Heck products.
Introduction
The palladium-catalyzed enantioselective Heck
reactions are among the most versatile synthetic
methods for the construction of stereogenic centers in
organic compounds. In particular, Heck arylations
have been instrumental in the synthesis of valuable
novel materials, including new medicines and fine
chemicals, constituting the activity of intense
research interest worldwide.[1–5] Recent developments
in the enantioselective Heck arylations have focused
mostly on the use of chiral bisphosphine monoxides
(BPMO),[6] P,N-ligands, N,N-ligands,[7] and phase-
transfer chiral counterions,[8,9] although the popular
bidentate phosphines are still commonplace.[10,11]
Additionally, the introduction of selectivity-inducing
strategies, such as the use of redox relay,[12–25] and
directing groups have also greatly improved and
extended the scope of this reaction.[26–31] However, in
spite of its tremendous success involving
nonactivated and cyclic electron-rich alkenes, the
effective arylation of acyclic electron-rich alkenes
has been lagging behind significantly. Among the
cyclic activated alkenes that can be employed in
enantioselective Heck reactions, it includes
tetrahydropyrans,
dihydrofurans,
endocyclic
enecarbamates, and enelactams.[32,33] Enantioselective
1
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
afforded the desired Heck product in 57% yield,
but with the e.r. dropping drastically to a nearly
racemic product (entry 6, Table 1). (S)-PyOx
ligands L4 and L5 afforded (S)-3a in only modest
yields and e.r. (entries 7 and 8, Table 1).
As shown in Table 1, the regioselectivity of the
arylation was highly dependent on the reaction
conditions. While changes on the base provided
only minor differences in its regioselectivity,
different ligands occasionally led to opposite
regioselectivities.
Table 1. Evaluation of Heck arylation of (E)-1aa
Scheme 1. Enantioselective Heck reactions of enol ethers
and analogs.
Results and Discussion
We started our investigation with the Heck-
Matsuda arylation of (E)-alkenyl phenyl ether 1a
using the 4-chlorophenyldiazonium salt 2a (a
readily available and very stable aryldiazonium
salt) and our previously reported reaction
conditions for similar Heck-Matsuda reactions
(Table 1),[15] i.e., Pd(TFA)2 as the palladium
source in combination with the pyrazine-
bisoxazoline ligand, (S)-PyraBox L1, basic zinc
carbonate as base and methanol as a solvent at
40°C (entry 1, Table 1). From the outset of our
studies, we envisioned the potential β-
elimination of the expected β-aryloxy aldehyde as
a critical issue, and we, therefore, reasoned that
trapping the aldehyde in situ with methanol could
Entry Ligand Base
(S)-3ab e.r.c
r.r.d
1:1.3
f
1
2
3
4
5
6
7
L1
L1
L1
L1
L2
L3
L4
L5
38
21
24
91:9
CaCO3
ZnCO3
89:11 1:1.5
92:8 1:2
90:10 1:1.4
2:1
44:56 2:1
86:14 1:1.5
62:38 1.2:1
DTBMP 36
f
46 (25e) 94:6
57
34
53
f
f
f
8
be
a
convenient way to circumvent the
a)
problem.25
These initial
Reaction conditions: (E)-1a (0.10 mmol), 2a (0.20
aforementioned
mmol), Pd(TFA)2 (5 mol%), ligand (6 mol% for L1; 12
mol% for L2−L5), base (0.01 mmol for
[ZnCO3]2[Zn(OH)2]3, 0.11 mmol for CaCO3 and DTBMP,
conditions gave the desired regioisomeric Heck
product (S)-3a in a modest 38% yield in a good
enantiomeric ratio (e.r.) of 91:9. As expected, the
aldehyde functionality formed after the redox
relay process was smoothly converted into the
corresponding dimethyl acetal under the reaction
conditions. Due to the polar nature of the Heck-
Matsuda arylation and the electronics of the
alkenyl phenyl ether, the regioselectivity of the
b)
and 0.05 mmol for ZnCO3), MeOH (1.0 mL). (%) Yield
1
determined by H NMR of the crude mixture using 3,5-
c)
bis(trifluoromethyl)bromobenzene as internal standard.
Enantiomeric ratio (e.r.) determined by chiral supercritical
fluid chromatography (SFC) analysis of purified 3a.
d)
1
Regioisomeric ratio (r.r.) determined by H NMR of the
crude mixture. e) 0.30 mmol scale; isolated yield. DTBMP
Heck reaction was expected to be high in favour
f)
1
=
2,6-Di-tert-butyl-4-methylpyridine.
of the

arylation. However, H NMR of the
[ZnCO3]2[Zn(OH)2]3.
crude reaction mixture indicated formation of a
thermally unstable regioisomeric Heck product.
e were unable to isolate and fully characterise the
product obtained from arylation at the α-position
To extend our studies and to get a better
understanding of the arylation process, we prepared
the (Z)-alkenol 1a and submitted it to the same
reaction conditions as above (Table 2). Gratifyingly,
the Z-alkenol led not only to the Heck product with
the opposite sense of chirality (R)-3a but also with an
improved yield of 55% and an excellent e.r. of 95:5
using L1 as ligand and CaCO3 as base (entry 1, Table
2). Screening of bases showed that basic zinc
carbonate, zinc carbonate, and DTBMP had a worse
performance as compared to calcium carbonate
W
,
possibly due to its degradation during flash
chromatography. Nevertheless, its structure was
tentatively assigned by NMR analysis of the crude
reaction mixtures (see SI). Screening of bases using
CaCO3, ZnCO3 or DTBMP did not improve the
yields (entries 2−4, Table 1), or the enantiomeric
ratios. We then evaluated different chiral N,N
ligands. The (S)-Box L2 increased the yield of
the product (S)-3a to 46% and improved the e.r.
to 94:6 (entry 5, Table 1). (S)-QuinOx L3
regarding both yields and e.r.’s (entries 2
−
4, Table
2
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
2). Bisoxazoline ligand L2, which had the best
performance for (E)-1a proved unsuitable in this
case, as both yield and er dropped significantly (entry
5, Table 2). Quinoline-oxazoline ligand L3 furnished
(R)-3a in a good yield of 60%, but in a racemic form
(entry 6, Table 2). Ligands L4 and L5 provided lower
yields and e.r. of the Heck product (entries 7-8, Table
2). Pd(OAc)2 and Pd2(dba)3 gave similar results to
Pd(TFA)2 with slightly lower yields and e.r. (entries
9-10, Table 2). Finally, the olefin to aryldiazonium
salt ratio was changed from 1:2 to 2:1, furnishing (R)-
3a in good yield but with lower e.r. (entry 11, Table
2).
With established conditions in hand, we moved on
to investigate the reaction scope of the olefin (Z)-1a
as substrate with different arenediazonium salts, in
order to gain insight into the effect of the different
electronics on the outcome of the reaction (Table 3).
During this study, we noticed that the reaction time
was critical in order to convert the aldehyde, the
redox-relay Heck product, into the corresponding
dimethylacetal derivatives. Para-chloro, -bromo and -
fluoro arenediazonium salts afforded the desired
arylated products 3a, 3b, and 3c in moderate yields
and excellent enantiomeric ratios (95:5; 95:5; and
94:6
respectively).
Meta-
and
ortho-
From the regioselectivity point of view, alkenol
bromophenyldiazonium salts gave the corresponding
3d and 3e products in good yields and excellent e.r.
(95:5 and 94:6 respectively). Arenediazonium salts
containing a trifluoromethyl group in the para or
meta positions, as well as a nitro group in the ortho
position also proved efficient to produce the
corresponding arylated compounds 3f, 3g, and 3h in
good yields and excellent e.r. (95:5; 94:6; and 95:5
respectively). Electron-donating groups on the
arenediazonium salt were met with a decrease in the
yield as shown for 3i containing a 3,4,5-trimethoxy
substitution pattern, nevertheless the e.r. was still a
good 92:8. In contrast, a single methoxy group in the
ortho position of the diazonium salt furnished the
arylated product 3j in low yield and low e.r. (70:30).
Compared with the result obtained with 3h containing
an ortho-nitro group, we conclude that the low e.r.
observed for 3j is due to the different electronic
properties of the arenediazonium salt. The
electronically neutral benzenediazonium salt gave the
arylated product 3k in moderate yield and a good e.r.
of 92:8.
The regioselectivity observed on the crude reaction
mixture was always in favour of the β-arylated
product. Just slight differences were observed
depending on the electronics of the aryldiazonium
salt. Ironically, the fact that the α-arylated
dimethylacetal product seems to decompose on silica-
gel chromatography, eased the purification of the
desired products. Moreover, the scope shown in
Table 3 demonstrates some bias of the Heck reaction
towards electron-deficient aryldiazonium salts.
(Z)-1a favoured the desired regioisomer (R)-3a
.
The highest regioisomeric ratio was obtained only
with an excess of the olefin (entry 7, Table 2),
with a slight decrease in the enantiomeric ratio
(compare with entry 1, Table 2). A few tests
involving other solvents, such as dioxane,
acetonitrile and dimethylcarbonate, together with
a mixture of these solvents with 5% of MeOH,
led to a reaction mixture containing up to 30% of
the corresponding cinnamaldehyde derivative, the
product of β-aryloxy elimination (see SI for
details).
Table 2. Evaluation of Heck arylation of (Z)-1aa
Entry Ligand Base
3ab
e.r.c
r.r.d
1
2
3
4
5
6
7
8
CaCO3
55 (49e) 95:5
3.8:1
2.6:1
1.4:1
2.3:1
L1
L1
L1
L1
L2
L3
L4
L5
L1
L1
L1
i
51
30
92:8
93:7
93:7
76:24 2.7:1
50:50 3.4:1
92:8
61:39 1.4:1
92:8 3.7:1
90:10 3.5:1
90:10 5.7:1
ZnCO3
DTBMP 34
CaCO3
CaCO3
CaCO3
CaCO3
CaCO3
CaCO3
CaCO3
27
60
28
29
49
54
55
1.5:1
9f
10g
11h
a)
Table 3. Enantioselective Heck reaction of (Z)-1a with
Reaction conditions: (Z)-1a (0.10 mmol), 2a (0.20
mmol), Pd(TFA)2 (5 mol%), ligand (6 mol% for L1; 12
different arenediazonium saltsa
mol%
for
L2−L5),
base
(0.01
mmol
for
[ZnCO3]2[Zn(OH)2]3, 0.11 mmol for CaCO3 and DTBMP,
b)
and 0.05 mmol for ZnCO3), MeOH (1 mL). (%) Yield
1
determined by H NMR of the crude mixture using 3,5-
c)
bis(trifluoromethyl)bromobenzene as internal standard.
Enantiomeric ratio (e.r.) determined by SFC analysis of
d)
1
purified 3a. Regioisomeric ratio (r.r.) determined by H
e)
NMR of the crude mixture. 0.30 mmol scale; isolated
yield. Pd(OAc)2 (5 mol%) was used as the palladium
source. Pd2(dba)3 (2.5 mol%) was used as the palladium
f)
g)
h)
source. (Z)-1a (0.20 mmol) and 2a (0.10 mmol) were
used. i) [ZnCO3]2[Zn(OH)2]3.
3
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
a) Reaction conditions: (Z)-1a (0.30 mmol), 2 (0.60 mmol),
Pd(TFA)2 (5 mol%), L1 (6 mol%), CaCO3 (0.33 mmol).
Isolated yields and enantiomeric ratios (e.r.) of the (R)-3,
an average of two experiments. Regioisomeric ratio (r.r.)
determined by 1H NMR of the crude reaction mixture.
a)Reaction conditions: (Z)-1a (0.30 mmol), 2 (0.60 mmol),
Pd(TFA)2 (5 mol%), L1 (6 mol%), CaCO3 (0.33 mmol).
Isolated yields and enantiomeric ratios (e.r.) of the (R)-3,
average of two experiments. Regioisomeric ratio (rr)
determined by 1H NMR of the crude reaction mixture.
Next, to investigate the influence of the electronics
of the enol ether in the arylation process we
investigated the reaction of several phenoxy-
substituted (Z)-olefins with o-bromo and m-
trifluoromethyl-substituted arenediazonium salts, as
these were the ones that performed best with olefin
(Z)-1a (Table 4). Olefins with substituents in the para
position of the phenoxy group afforded the arylated
products in good yields and high e.r., even when
different groups were used such as trifluoromethyl (3l
and 3r, Table 4), chlorine (3m and 3s) and methoxy
(3n and 3t). Strong EWG on the ortho position, such
as the nitro, led to a slight decrease in the yield and er
(3o and 3u). On the other hand, a weak electron-
donating group, such a methyl group, on the same
position gave the desired products in good yield and
high e.r. (3p and 3v). Finally, an olefin derived from
β-naphthol was also tested furnishing the arylated
compounds in good yield and excellent e.r. (3q and
3w).
Regarding the regioselectivity of the reaction, it
can be observed that o-bromo-benzenediazonium salt
2e led to better regioselectivities than m-CF3
benzenediazonium salt 2g. The worst r.r. was
obtained for 3o and 3u containing an o-nitro group
(3.5:1; and 2:1 respectively), whilst the best was 3l
(10:1) for the o-bromo-benzenediazonium salt (2e)
and 3s (6:1) for m-CF3 benzenediazonium salt (2g).
Olefins containing opposite electron-demanding
groups did not affect the r.r. as much as it would be
expected (compare 3l and 3n, and 3r and 3t). These
results clearly show slight higher regioselectivity for
olefins containing EWD groups in the aryloxy
substituent.
To further extend the method, we evaluated olefins
containing a secondary allylic alcohol moiety such as
(E)-4 and (Z)-4 (Scheme 2). Surprisingly, the reaction
of olefin 4 with arenediazonium salt 2e under the
conditions used in Table 4 failed. However, a simple
change of base from CaCO3 to ZnCO3 has allowed
the synthesis of arylated methyl ketones 5a and 5b,
albeit in low yield and in moderate to excellent e.r.
(Scheme 2). The enantiomer obtained was highly
dependent on the geometry of the double bond of the
olefins 4.
Table 4. Enantioselective Heck reaction of olefin (Z)-1
derivatives and arenediazonium saltsa
4
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
Scheme 4. Synthetic application for the Heck arylated
product (R)-3e.
Next, the dimethyl acetals were converted into the
synthetically valuable methyl esters using
mCPBA/BF3·OEt2 (Table 5).[47] For the specific case
of our Heck products 3, this was a challenging
transformation due to the liability of β–aryloxy esters
under Lewis acid conditions. Oxidation of compound
3e with mCPBA led to the corresponding ester 7a in a
modest 25% yield. However, for substrates bearing
an electron-deficient aryloxy group, there was a
significant increase in yields. Therefore, esters 7b, 7c,
7d, and 7e bearing a para-trifluoromethyl group,
para-chloro, and an ortho-nitro group, good yields of
the respective esters were obtained. As expected, the
yield dropped significantly for ester 7f containing a
para-methoxy group in the aryloxy ring. Ester 7g,
containing an ortho-methyl group in the aryloxy ring
was obtained only in trace amounts, together with
other by-products. Preparative TLC of the crude
reaction mixture allowed us to obtain an analytical
sample of ester 7g for structural characterization. The
e.r. of esters 7b and 7f were compared to those of the
dimethyl acetal substrates and confirmed that this
oxidation method did not affect the stereogenic centre
formed in the Heck reaction.
Scheme 2. Enantioselective Heck reactions of (E)-4 and
(Z)-4.
The results obtained above allowed us to envision
the synthetic potential of the method toward complex
frameworks and pharmacologically active molecules
of interest, as summarised in Scheme 3.
Table 5. Oxidation of dimethyl acetals (R)-3 to the
corresponding methyl esters (R)-7.
Scheme 3. Synthetic application for the Heck arylated
product (R)-3.
First, we subjected compound (R)-3e to the method
described by Fagnou and collaborators for the
intramolecular cyclization by C−H activation.[43] As
shown in Scheme 4, the direct intramolecular
arylation of (R)-3e led to compound (R)-6 containing
the 6H-benzo[c]chromene core in excellent yield
without any signs of erosion of the enantiomeric
ratio.[44,45] It is worth mentioning that this core
structure is present in a number of biologically active
compounds, such as the cannabinols.[46]
The lower yields of methyl esters 7 from the
dimethyl acetals possessing electron-rich β-aryloxy
groups is not totally surprising. These groups could
easily compete with the dimethyl acetals for the BF3
Lewis acid during the reaction, leading to β-aryloxy
elimination and other by-products. The complexity
and sensibility of this transformation are clearly
demonstrated by the isolation and structure
elucidation of its main side product. This was the
case for the oxidation of dimethyl acetals 3e and 3p.
The major side-products of the oxidation of these
dimethyl acetals were identified as their 3,4-dihydro-
5
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
2H-1,5-benzodioxepin-2-one derivatives (R)-7aʼ and
(R)-7gʼ (Figure 1). We postulate that compounds (R)-
7aʼ and (R)-7gʼ are formed from a Friedel-
Crafts/Baeyer-Villiger sequence as illustrated in the
rationale presented in the SI.
literature allowed us to assign the absolute
configuration of our Heck products as indicated (see
SI for details).[49] We extended the configuration
determined by this method to all the compounds
synthesized in this study.
Figure 1. The main side-products formed by the oxidation
of acetals 3e and 3p into their respective esters using
mCPBA/BF3·OEt2.
Scheme 6. Synthesis of (R)-fluoxetine and (R)-
atomoxetine and stereochemical assignments. Reaction
conditions: a) Amberlyst®-15, acetone. b) NaH2PO4,
NaClO2, amylene, tBuOH/H2O. c) MeNH2·HCl, EDC·HCl,
HOBt, DIPEA, DCM. d) BH3·Me2S, THF, then HCl,
reflux and then NaOH. For more information see SI.
As stated previously, the free aldehydes are
susceptible to β-aryloxy elimination (Table 1).
Nonetheless, we reasoned that the use of a mild
proton source could avoid the decomposition
pathway and allow access to this versatile
intermediate. We decided to test our hypothesis with
compounds (R)-3x and (R)-3y - prepared via C−Br
bond hydrogenolysis of compounds (R)-3l and (R)-3p
(see SI). Although compounds (R)-3x and (R)-3y
could be prepared by the direct Heck-Matsuda
To get further insights about these challenging
Heck-Matsuda arylations, we also decided to
investigate the enantioselectivity of these reactions
through DFT studies. The analysis of the aryl
palladium migratory insertion into the alkene has
shown to be crucial to rationalize the stereochemical
outcome of the Heck reactions on many occasions.
arylation
of
allylic
alcohol
(Z)-1
using
[14,27–29,50–52]
Our investigation took into account the
benzenediazonium tetrafluoroborate, we found out
that this particular reaction leads to a lower yield of
the desired Heck product. On the other hand, the
reaction using the benzenediazonium salt allowed us
to isolate and characterise the regioisomer formed
from arylation on the α-position (see SI). To our
delight, the use of Amberlyst®-15, a mild proton
source proved effective for the hydrolysis of the
dimethylacetals to provide the respective aldehydes.
For example, acetal (R)-3x was hydrolyzed with
Amberlyst®-15 followed by a subsequent reduction
with NaBH4 affording the alcohol (R)-8 in high yield
and er (Scheme 5). The arylated alcohol (R)-8
constitutes an advanced intermediate in the total
synthesis of fluoxetine as described by Trost and co-
workers.[48]
four permutations between the orientations of
different substrate components for both (Z)-1a and
(E)-1a (Scheme 7). Therefore, we investigated the
enantioselectivity of the reaction (i.e.: arylation at
either the Re or Si face of a given alkene carbon), and
the relative orientation of the ligands during the
migratory insertion (i.e.: alkene cis or trans to the
oxazoline at the stage of the corresponding σ-
arylpalladium). For comparison, we also investigated
a less computationally demanding model ligand in
which one of the oxazolines was replaced by a
fluorine substituent.
As in previous investigations, we assumed that
Curtin-Hammett conditions apply to the migratory
insertion and, that the redox relay is efficient at
converting the migratory insertion products, alkyl
palladium complexes, into the reaction products
(aldehydes or ketones).[14,50–52] The redox relay
process was previously investigated by Sigman and
coworkers and, in general, it is an efficient process
with both PyOx and PyraBox ligands.[15,52] In essence,
this transformation consists of a series of syn β-
hydride eliminations and palladium hydride
migratory insertions. Under these assumptions, the
product distribution should be determined by the
difference in free energy of activation of the
corresponding migratory insertion transition states.
The computational method employed was M06-
L/def2-TZVP//M06-L/set1, in which set1 is a basis
set composed of 6-31G(d,p) for H, C, N, O, F and Cl
atoms and LANL2DZ pseudopotential with its
associated basis set for Pd. Solvent effects were
included both during the optimization and at single
point calculations by the SMD model (more
information about the method as well as the xyz
Scheme 5. Reduction of acetal (R)-3x to alcohol (R)-8.
To further demonstrate the synthetic potential of
the developed method, and to ascertain our
stereochemical assignments, we also employed the
advanced intermediates dimethylacetals (R)-3x and
(R)-3y in the total synthesis of fluoxetine and
atomoxetine respectively (Scheme 6). The synthesis
of (R)-fluoxetine and (R)-atomoxetine were
completed in 4 steps from intermediates (R)-3x and
(R)-3y in reasonably good overall yields as indicated
in Scheme 6. The comparison of their optical rotation
values with the corresponding ones reported in the
6
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
coordinates for the transition states are described in
the SI).
Both for (Z)-1a and (E)-1a, the lowest energy
transition states found do correlate with the absolute
stereochemistry of the corresponding products
(Scheme 7). In addition, given the 2.8 kcal·mol–1 free
energy difference between TS3 and TS2, high levels
of enantio-induction are expected for this reaction.
Indeed, this observation agrees qualitatively with
high experimental enantioselectivity obtained for this
transformation (er: 95:5, ∆∆G‡ ≈ 1.84 kcal·mol–1
estimated from: ∆∆G‡ = –RT ln(e.r.)). Similarly, the
2.4 kcal·mol-1 free energy difference between TS6
and TS8 also qualitatively agrees with the
experimental enantioselectivity (e.r.: 92:8, ∆∆G‡ ≈
1.52 kcal·mol–1).
The reaction enantioselectivity can be understood
once one recognizes that the chiral palladium
complex differentiates prochiral alkenes’ faces
depending upon their substitution at the β-aryloxy
position. This differentiation is based primarily on
steric hindrance between the allylic hydrogens and
the bulky substituent at the oxazoline moiety.
Arylation proceeds at the α-aryloxy position and thus
the stereoisomer formed depends upon the double
bond geometry. The preference for transition states in
which the aryl group is proximal to the pyrazine
moiety can be understood both in terms of steric
t
hindrance between Bu group and aryl group and in
terms of a favourable C-H···π interaction between the
pyrazine C-H bond (TS3 and TS6).[50,51]
Scheme 7. Relative free energies for migratory insertion
on the (Z)-1a and (E)-1a.
Conclusion
In summary, we have developed an efficient
enantioselective Heck-Matsuda arylation of acyclic
enol ethers. The arylated products were prepared in
moderate to good yields and in high enantiomeric
ratios. Competing β-aryloxy eliminations were
avoided in most cases via in situ methanol
ketalization of the intermediate β-aryloxyaldehydes.
The dimethyl acetals obtained can be effectively
transformed into their corresponding methyl esters by
mild oxidation with mCPBA/BF3·OEt2 or to the
primary alcohols by a two-step sequence involving
hydrolysis with Amberlyst®-15 and immediate
reduction with NaBH4. The method demonstrated its
synthetic
potential
by
the
synthesis
of
enantioenriched bioactive compounds and valuable
core structures, like (R)-fluoxetine, (R)-atomoxetine,
and a functionalized benzo[c]chromene. The relative
free energies of activation of the carbopalladation
step were calculated with DFT and are in good
agreement with the product distribution of the Heck-
Matsuda reaction. The enantiodivergent nature of the
arylation, regarding the E and Z enol ethers, could be
also understood as a natural consequence of the
transition states these olefins go through.
7
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
Experimental Section
[10] G. Yue, K. Lei, H. Hirao, J. (Steve) Zhou, Angew.
Chem., Int. Ed. Engl. 2015, 54, 6531–6535.
[11] J. A. Carmona, V. Hornillos, P. Ramírez-López, A.
Ros, J. Iglesias-Sigüenza, E. Gómez-Bengoa, R.
Fernández, J. M. Lassaletta, J. Am. Chem. Soc. 2018,
140, 11067–11075.
[12] Z.-M. Chen, M. J. Hilton, M. S. Sigman, J. Am.
Chem. Soc. 2016, 138, 11461–11464.
[13] H. H. Patel, M. S. Sigman, J. Am. Chem. Soc. 2016,
138, 14226–14229.
General Procedure for Enantioselective Heck-
Matsuda Reactions: Pd(TFA)2 (4.99 mg, 5 mol%,
0.015 mmol), ligand (S)-PyraBox L1 (5.95 mg, 6
mol%, 0.018 mmol), and methanol (1.5 mL or 0.75
mL) were added to a 4 mL screw-top vial containing
a magnetic stirrer. The resulting light-orange solution
was then stirred at 40°C for 15 min to form the
catalyst complex. In another 4 mL vial, was added the
olefin (Z)-1a (0.30 mmol), CaCO3 (0.33 mmol, 33.0
mg, 1.1 equiv) or ZnCO3 (0.30 mmol, 37.6 mg, 1
equiv), diazonium salt (0.60 mmol, 2 equiv) and the
solution of catalyst complex, rinsing the vial
containing the catalyst solution with MeOH (1.5 mL
or 0.75 mL). The vial was then capped and stirred for
1-8 h. (Caution! Pressure might develop inside the
reaction flask due to the N2 released by the reaction).
The reaction vessel was then cooled to room
temperature, carefully opened and the reaction
mixture was poured onto a pad of silica gel (230–400
mesh; column height ~3 cm; diameter ~2 cm)
[14] C. Zhang, B. Tutkowski, R. J. DeLuca, L. A. Joyce,
O. Wiest, M. S. Sigman, Chem. Sci. 2017, 8, 2277–
2282.
[15] C. C. Oliveira, A. Pfaltz, C. R. D. Correia, Angew.
Chem., Int. Ed. Engl. 2015, 54, 14036–14039.
[16] T. Y. Chaudhari, A. Hossian, M. K. Manna, R. Jana,
Org. Biomol. Chem. 2015, 13, 4841–4845.
[17] W. Khodja, A. Leclair, J. Rull-Barrull, F. Zammattio,
K. V. Kutonova, M. E. Trusova, F.-X. Felpin, M.
Rodriguez-Zubiri, New J. Chem. 2016, 40, 8855–
8862.
[18] D. Cortés-Borda, K. V. Kutonova, C. Jamet, M. E.
Trusova, F. Zammattio, C. Truchet, M. Rodriguez-
Zubiri, F.-X. Felpin, Org. Process Res. Dev. 2016,
20, 1979–1987.
previously washed with
a
50% mixture of
EtOAc/hexanes. The silica pad was then washed with
about 50 mL of the same eluent. The solvent was
removed with a rotary evaporator and the resulting
mixture was purified by column chromatography on
silica gel with ethyl acetate and hexanes as the eluent.
[19] H. Nakamura, K. Yasui, Y. Kanda, P. S. Baran, J.
Am. Chem. Soc. 2019, 141, 1494–1497.
[20] N. J. Race, Q. Yuan, M. S. Sigman, Chem. Eur. J.
2019, 25, 512–515.
[21] Z.-M. Chen, C. S. Nervig, R. J. DeLuca, M. S.
Sigman, Angew. Chem., Int. Ed. Engl. 2017, 56,
6651–6654.
[22] C. Zhang, C. B. Santiago, J. M. Crawford, M. S.
Sigman, J. Am. Chem. Soc. 2015, 137, 15668–15671.
[23] E. W. Werner, T.-S. Mei, A. J. Burckle, M. S.
Sigman, Science 2012, 338, 1455.
Acknowledgements
We thank the financial support of the São Paulo Research
Foundation (FAPESP, grants: 2017/21494-2 [MFW];
2019/02052-4 [RAA]; 2015/01491-3 [AACB] 2014/25770-6, and
2013/07600-3, and 2014/25770-6 [CRDC]), the Coordination for
the Improvement of Higher Education Personnel (PNPD-CAPES
[ECP]) and the Brazilian National Research Council (CNPq,
grants 406643/2018-0, and 306773/2018-0 [CRDC]).
[24] C. Frota, E. C. Polo, H. Esteves, C. R. D. Correia, J.
Org. Chem. 2018, 83, 2198–2209.
[25] C. C. Oliveira, R. A. Angnes, C. R. D. Correia, J.
Org. Chem. 2013, 78, 4373–4385.
[26] R. A. Angnes, J. M. Oliveira, C. C. Oliveira, N. C.
Martins, C. R. D. Correia, Chem. Eur. J. 2014, 20,
13117–13121.
[27] J. de Oliveira Silva, R. A. Angnes, V. H. Menezes
da Silva, B. M. Servilha, M. Adeel, A. A. C. Braga,
A. Aponick, C. R. D. Correia, J. Org. Chem. 2016,
81, 2010–2018.
[28] F. de Azambuja, R. C. Carmona, T. H. D. Chorro, G.
Heerdt, C. R. D. Correia, Chem. Eur. J. 2016, 22,
11205–11209.
[29] S. Kattela, G. Heerdt, C. R. D. Correia, Adv. Synth.
Catal. 2017, 359, 260–267.
[30] J. M. de Oliveira, R. A. Angnes, I. U. Khan, E. C.
Polo, G. Heerdt, B. M. Servilha, V. H. Menezes da
Silva, A. A. C. Braga, C. R. D. Correia, Chem. Eur.
J. 2018, 24, 11738–11747.
[31] R. A. Angnes, L. M. Thompson, M. S. Mashuta, C.
R. D. Correia, G. B. Hammond, Adv. Synth. Catal.
2018, 360, 3760–3767.
[32] G. M. Borrajo-Calleja, V. Bizet, T. Bürgi, C. Mazet,
Chem. Sci. 2015, 6, 4807–4811.
References
[1] A. B. Dounay, L. E. Overman, Chem. Rev. 2003,
103, 2945–2964.
[2] D. Mc Cartney, P. J. Guiry, Chem. Soc. Rev. 2011,
40, 5122–5150.
[3] M. Shibasaki, E. M. Vogl, T. Ohshima, Adv. Synth.
Catal. 2004, 346, 1533–1552.
[4] M. Oestreich, The Mizoroki-Heck Reaction, John
Wiley & Sons, 2009.
[5] M. Oestreich, Angew. Chem., Int. Ed. Engl. 2014, 53,
2282–2285.
[6] X. Wu, J. (Steve) Zhou, Chem. Commun. 2013, 49,
4794–4796.
[7] R. C. Carmona, O. D. Köster, C. R. D. Correia,
Angew. Chem., Int. Ed. Engl. 2018, 57, 12067–
12070.
[8] C. M. Avila, J. S. Patel, Y. Reddi, M. Saito, H. M.
Nelson, H. P. Shunatona, M. S. Sigman, R. B. Sunoj,
F. D. Toste, Angew. Chem., Int. Ed. Engl. 2017, 56,
5806–5811.
[9] Y. Reddi, C.-C. Tsai, C. M. Avila, F. D. Toste, R. B.
Sunoj, J. Am. Chem. Soc. 2019, 141, 998–1009.
8
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
[33] C. Wu, J. Zhou, J. Am. Chem. Soc. 2014, 136, 650–
652.
[34] J. Hu, Y. Lu, Y. Li, J. (Steve) Zhou, Chem. Commun.
2013, 49, 9425–9427.
[35] Q. Yuan, M. S. Sigman, J. Am. Chem. Soc. 2018,
140, 6527–6530.
[45] D. García-Cuadrado, P. de Mendoza, A. A. C. Braga,
F. Maseras, A. M. Echavarren, J. Am. Chem. Soc.
2007, 129, 6880–6886.
[46] L. Qi, N. Yamamoto, M. M. Meijler, L. J. Altobell,
G. F. Koob, P. Wirsching, K. D. Janda, J. Med.
Chem. 2005, 48, 7389–7399.
[36] S. Kattela, E. C. de Lucca Jr., C. R. D. Correia,
Chem. Eur. J. 2018, 24, 17691–17696.
[47] E. J. Corey, W. G. Su, J. Am. Chem. Soc. 1987, 109,
7534–7536.
[37] Z.-Z. Jiang, Y.-J. Jiang, J. Du, D. Chen, C.-H. Ding,
B. Xu, X.-L. Hou, Synthesis 2019, s-0037-1610712.
[38] H. H. Patel, M. B. Prater, S. O. Squire, M. S. Sigman,
J. Am. Chem. Soc. 2018, 140, 5895–5898.
[39] M. B. Prater, M. S. Sigman, Israel Journal of
Chemistry 2019, 0, DOI 10.1002/ijch.201900077.
[40] J. G. Taylor, A. V. Moro, C. R. D. Correia, Eur. J.
Org. Chem. 2011, 2011, 1403–1428.
[41] A. Roglans, A. Pla-Quintana, M. Moreno-Mañas,
Chem. Rev. 2006, 106, 4622–4643.
[42] F.-X. Felpin, L. Nassar-Hardy, F. Le Callonnec, E.
Fouquet, Tetrahedron 2011, 67, 2815–2831.
[43] L.-C. Campeau, M. Parisien, A. Jean, K. Fagnou, J.
Am. Chem. Soc. 2006, 128, 581–590.
[48] B. M. Trost, P. L. Fraisse, Z. T. Ball, Angew. Chem.,
Int. Ed. Engl. 2002, 41, 1059–1061.
[49] Y. Li, Z. Wang, K. Ding, Chem. Eur. J. 2015, 21,
16387–16390.
[50] Y. Dang, S. Qu, Z.-X. Wang, X. Wang, J. Am. Chem.
Soc. 2014, 136, 986–998.
[51] L. Xu, M. J. Hilton, X. Zhang, P.-O. Norrby, Y.-D.
Wu, M. S. Sigman, O. Wiest, J. Am. Chem. Soc.
2014, 136, 1960–1967.
[52] M. J. Hilton, L.-P. Xu, P.-O. Norrby, Y.-D. Wu, O.
Wiest, M. S. Sigman, J. Org. Chem. 2014, 79,
11841–11850.
[44] D. García-Cuadrado, A. A. C. Braga, F. Maseras, A.
M. Echavarren, J. Am. Chem. Soc. 2006, 128, 1066–
1067.
9
This article is protected by copyright. All rights reserved.

10.1002/adsc.201901471
Advanced Synthesis & Catalysis
FULL PAPER
Enantioselective Heck Arylation of Acyclic
Alkenol Aryl Ethers: Synthetic Applications and
DFT Investigation of the Stereoselectivity
Adv. Synth. Catal. Year, Volume, Page – Page
Ellen Christine Polo,ǂ,a Martí Fernández Wang,ǂ,a
Ricardo Almir Angnes,b Ataualpa A. C. Braga,b
and Carlos Roque Duarte Correia*a
10
This article is protected by copyright. All rights reserved.