Gold(I)-Catalyzed [2 + 2 + 2] Cyclotrimerization of 1,3-Diarylpropargyl Acetals

Article abstract of DOI:10.1021/acs.orglett.7b00411

A gold-nitrone catalyzed [2 + 2 + 2] cyclotrimerization of 1,3-diarylpropargyl acetals into cyclohexylidene products (up to 74% yield) is reported. The trimerization is proposed to proceed through allenic intermediates via gold-catalyzed 1,3-alkoxy rearra

Full text of DOI:10.1021/acs.orglett.7b00411

Letter
pubs.acs.org/OrgLett
Gold(I)-Catalyzed [2 + 2 + 2] Cyclotrimerization of 1,3-
Diarylpropargyl Acetals
Helgi Freyr Jonsson, Sigvart Evjen, and Anne Fiksdahl*
́
Department of Chemistry, Norwegian University of Science and Technology, NO-7491 Trondheim, Norway
*
S Supporting Information
ABSTRACT: A gold-nitrone catalyzed [2 + 2 + 2] cyclo-
trimerization of 1,3-diarylpropargyl acetals into cyclohexyli-
dene products (up to 74% yield) is reported. The trimerization
is proposed to proceed through allenic intermediates via gold-
catalyzed 1,3-alkoxy rearrangement. The presence of catalytic
amounts of different nitrones, tuning of the Au(I) catalyst
activity, was essential for controlled regio-/chemoselective
cyclotrimerization. A linear nitrone−O−Au(I)−P coordination
mode was shown (X-ray analysis) for a catalytic active phosphane−gold−nitrone complex, representing a group of Au(I)
catalysts with specific properties.
2
g
he propargyl ester−gold approach, based on gold-
(Scheme 1g), a novel unusual [2 + 2 + 2] cyclotrimerization of
nonterminal diarylpropargyl acetals was observed. 2,4,6-Triaryl-
1,3,5-tricyclohexylidene products were selectively formed
(Scheme 1h). Diarylpropargyl substrates have until recently
not been applied in synthetic studies, as they have been though3t
to be unviable substrates, affording complex reaction mixtures.
Conventional transition-metal-catalyzed alkyne [2 + 2 + 2]
cyclotrimerization via metallocyclopentadiene complexes repre-
sents a useful transformation to provide highly substituted
T
catalyzed activation of propargyl esters, has been applied
1
in a variety of cycloaddition reactions. Studies on gold(I)-
catalyzed reactions of the corresponding propargyl acetals have
1
m−q
been scarce.
The Fiksdahl group has, however, carried out a
number of studies on chemoselective gold(I)-catalyzed cyclo-
2
addition of propargyl acetals (I, Scheme 1b−g), showing that
Scheme 1. Au(I)-Catalyzed Cycloadditions of Propargyl
2
4a
4b
Substrates
arenes. Gold nanoparticles also catalyze alkyne trimerization.
However, the [2 + 2 + 2] cycloadditions suffer from regio- and
chemoselectivity limitations. Allenes cyclotrimerize under
4c
metal-free conditions as well. On the basis of a different
stepwise mechanism via a gold dimeric cationic intermediate,
4d
the gold-catalyzed allenamide cyclotrimerization is reported to
match the outcome of alkyne [2 + 2 + 2] trimerization reactions.
To the best of our knowledge, our presently reported reaction
is the first intermolecular regioselective [2 + 2 + 2]
cyclotrimerization of propargyl substrates, affording densely
substituted cyclohexylidene products. We hereby report the
results on gold−nitrone-catalyzed transformations of diary-
lpropargyl acetals.
Studies on Reaction Conditions and Reactivity
Introductory studies on the Au(I)-catalyzed reactions between
equimolar amounts of diarylpropargyl acetal 1a and nitrone 2a
in DCM gave full conversion in 2 h at room temperature in the
presence of 5 mol % of JohnPhos−Au(I) catalyst I (Table 1,
entry 1). The reaction showed that the cyclotrimerization
cyclohexylidene 3a product was formed as the major product
they are versatile substrates for a diverse range of novel Au(I)-
catalyzed cycloadditions with multiple bond reactants, leading to
a series of polyfunctionalized carbo- and heterocyclic products
via highly reactive gold(I) carbenoid intermediates (II). The
activating effect provided by the C2-vinylalkoxy group in the
MeO-vinylgold carbenoid (II) facilitates the reactions and
explains the high reactivity and ability of propargyl acetals to
undergo cycloadditions. In the course of our recent studies on
(
60%) by a [2 + 2 + 2] propargyl cycloaddition. The product
was formed as a mixture of three stereoisomers. Au(I) catalyst
oxidation of alkynes, including propargyl substrates, is known to
Received: February 13, 2017
[
3 + 3] cycloadditions with nitrones, giving oxazine products
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b00411
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
5,2g
Table 1. Studies on [2 + 2 + 2] Cyclotrimerization
take place with nitrone.
However, no competing oxidation
byproducts were seen in the present study. As the nitrone 2a
was not incorporated in the product, it did not seem to take part
in the reaction. Unexpectedly, complex undefined product
mixtures were obtained by Au(I) I catalyzed reaction of
diarylpropargyl acetal 1a in the absence of nitrone 2a (Table 1,
entry 2). No formation of the cyclic trimer 3a took place. The
fact that nitrone 2a was not incorporated in the trimeric product
and that complex product mixtures were obtained in the absence
of nitrone 2a would suggest that Au catalyst I was too active to
afford selective trimerization. A relevant hypothesis would be
that reduced Au catalyst activity may be due to nitrone−Au
coordination, thereby allowing chemoselective trimerization to
take place. The Au−nitrone hypothesis was supported by the
fact that the rate and yield of the reactions were essentially
unchanged by reducing the amount of nitrone 2a from 1.0 to
0
.05 equiv (60−65% yield of 3a in 2 h, Table 1, entry 3). Hence,
controlled Au(I)-catalyzed trimerization reactions of propargyl
acetal 1a only took place in the presence of (catalytic amounts
of) nitrone 2a.
Reactivity studies of acetals 1a−d in the presence of Au(I)
complexes I−III (0.05 equiv) and nitrones (2a−c, 0.1 equiv)
were carried out. The Au-complex counterions (± nitrone), the
substrate concentration, and the temperature (Table 1) were
varied. The reactivity and reaction time of MeO-, CF -, and Cl-
3
phenylpropargyl acetals 1b−d differed, but the yields remained
almost constant (54−60%, 2−24 h, entry 4). The aromatic
nitrones, pyridine N-oxide 2b and 8-methylquinoline 1-oxide 2c,
were likewise able to tune the activity of Au catalyst I, affording
somewhat higher reactivity (>64% yield, entries 5 and 6) than
nitrone 2a. The isomer ratios of products 3a−d (entries 3−6)
1
were approximately 5:3:3, as shown by H NMR. Selected
isomers of 3b and 3c were isolated and fully characterized
(
Scheme 2b).
Scheme 2. (a) Outline of the Overall Triple C−C Bond
Formation Sequence for Au(I)−nitrone Catalyzed [2 + 2 + 2]
Cyclotrimerization of Propargyl Acetals 1. (b) X-ray
Structure Analysis of Cyclotrimerization Products 3b and 3c
a
General reactions conditions: 1a−c (1 equiv) in DCM (approx.10
mg/mL; c = 33 mM), Au I−III (5 mol %), nitrone (2a−c, 5 mol %)
were stirred at rt before being quenched with NEt . Total yield of
three isomers. Full conversion into complex product mixture; trimeric
product (3) not detected. The reaction performed with crystalline
b
3
c
The Au(I) complex counterion slightly affected the reactivity,
d
−
as the more weakly coordinating anion, NTf , gave higher
2
Au(I)I−nitrone 2b complex (5 mol %) afforded a similar yield of 3b as
yields of product 3b than SbF6− (74% vs 64%, entries 7 and 5).
e
when the catalyst was generated in situ. Au(I) catalyst formed by
f
g
Comparable results were obtained by the less bulky Ph P (II) to
3
counterion exchange. No reaction observed. Six times higher
h
the originally used JohnPhos ligand (entries 4 and 9). The
NHC−Au(I) catalyst III was also tested, as a less active catalyst
could be expected to give controlled trimerization, even without
nitrone tuning of the Au(I) catalyst. However, complex product
concentration of 1b; c = 210 mM. 50% diluted conc. of 1b; c = 17
mM. Dimer 4 was formed in the absence of nitrone. No cyclotrimeric
triacetate was formed. Nitrone was replaced by benzotriazole (5 mol
i
j
%
).
B
DOI: 10.1021/acs.orglett.7b00411
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
mixtures were obtained both in the presence and absence of
quent acetal cleavage and 1,3-alkoxy shift followed by C2′−C1″
bond formation is repeated twice, the last time in a C2″−C1
intramolecular manner. In total, the cyclotrimeric product 3 is
afforded by ring closure, and deauration allows regeneration of
the Au(I) catalyst. Previous NMR studies of 1,3-diarylpropargyl
pivaloyl esters showed that Au-catalyzed 1,3-acyloxy shift and
rearrangement immediately took place to give full conversion3
and quantitative amounts of the corresponding stable allene.
Subsequent allene [2 + 2] cyclodimerization seemed to be a
thermal reaction, as a separate reaction afforded the trans-
nitrone (entries 10 and 11). Likewise, Au(III)Cl (entry 12) was
3
unable to afford trimerization, giving full conversion of 1b into a
number of unidentified products.
An effect of the substrate concentration was expected for this
reaction. Higher concentration would assist intermolecular
reaction, such as dimer and trimer formation. However, the final
intramolecular cyclization step would be favored in diluted
reaction mixtures, which seemed to control the outcome of the
reaction, as higher reactivity and yield were obtained by 50%
substrate dilution (2 h, 71%, entry 14). Moreover, yields
dropped remarkably in six times concentrated solution (27%, 24
3
formation without gold activation. We did not observe a
corresponding simple transformation into one stable allene
1
h, entry 13; versus 64%, 1.5 h, entry 5).
Combining the positive impact of counterion (NTf , entry
intermediate ( H NMR). In contrast, generation of complex
−
mixtures of intermediates was temporarily seen during Au(I)−
nitrone-catalyzed conversion of acetals 1 until the selective
formation of the cyclic products 3 was completed.
2
7) and dilution (entries 14) did not give any additional effect on
the yield of trimeric product 3b (entry 15). An alternative
strategy to control the reaction toward selective trimerization
would be to compensate for the absence of nitrone by decreased
temperature. However, low temperature (−78 °C) failed to
provide cyclotrimerization or any conversion at all (entry 16),
while reactions at 0 to −40 °C gave complex product mixtures
1
H NMR of stereoisomers of products 3 indicated nonsym-
metrical cyclic trimeric structures, as demonstrated, e.g., by
different MeO surroundings, shown by three separate MeO-
signals of each isomer. X-ray analysis of isolated isomers of
products 3b and 3c (Scheme 2b) showed a trans,trans,cis
relationship between the 2,4,6-triarylcyclohexyl substituents,
while the stereochemistry of the three vinyl groups varied.
(
entries 17 and 18), showing that Au catalyst tuning may only be
obtained by nitrone coordination.
The diarylpropargyl acetate 1′, which is known to be less
Studies on the Au−Nitrone Catalyst
2
reactive than the corresponding acetal, failed to give the
1
H and 31P NMR analysis of an equimolar mixture of Au(I) I
corresponding cyclotrimeric triacetate. Low conversion of
acetate 1′ was seen in the presence of nitrone. While full
conversion was seen in the absence of a nitrone, only minor
amounts of a [2 + 2] cyclodimerization cyclobutylidene product
and nitrone 2a indicated that an exchange of the acetonitrile
ligand of the Au(I) catalyst I with nitrone 2a took place to give a
Au(I)−nitrone complex (Scheme 3a). In a previous study,
8a
a
(
4, 16%, one isomer, 24 h, entry 19) were isolated. Previous
Scheme 3. a) Formation of Au(I)−Nitrone 2a Complex,
6a
studies reported formation of both [2 + 2] regioisomers, and
the dimerization was proposed to proceed through an allenic Au
intermediate formed by Au(I)-catalyzed 1,3-O-acyl shift. The
alkylpropargyl moiety seemed to be unsuitable for the reaction,
as cyclotrimerization of methylpropargyl acetal 1″ was
unsuccessful. Full conversion was observed (24 h, entry 20),
but no trimeric product was detected. Benzotriazole−Au(I)
catalysts have been employed in chemoselective propargyl ester
Including NMR Observations. (b) X-ray Structure Analysis of
Au(I)−Nitrone 2a Complex
6
b
transformations. However, when the original Au(I)−nitrones
were replaced with a benzotriazole-coordinated JohPhosAu(I)
complex, no reaction took place (24 h, entry 21).
Reaction Mechanism
gold(I)−nitrone coordination has been suggested to proceed
Alkyne substitution of propargyl substrates has a strong effect of
their reactivity in Au-catalyzed reactions. Terminal propargyl
acetals or esters tend to undergo 1,2-alkoxy or -acyloxy
migration, affording gold carbenoid intermediates (II, Scheme
through Au(I) ligand exchange, while X-ray diffraction studies of
+
a Au(III)Cl −nitrone−H complex failed to show any Au−
3
nitrone coordination through the O atom. Nitrone ligands are,
however, known from Ru−heme complexes and organoboron
compounds. X-ray analysis of our crystalline Au−nitrone 2a
complex, afforded from an equimolar mixture of Au I and
nitrone 2a, confirmed for the first time the O-binding of the
nitrone ligand to the Au(I) center (Scheme 3b), in analogy to
8b
1
g−p,2
8c,d
1
). In contrast, nonterminal propargyl acetals/esters
undergo 1,3-alkoxy shift to generate allenic intermediates,
which may give rise to a series of reactions. [2 + 2]
6
7a−d
cycloaddition of allenamides with alkenes
and [2 + 2]
7e,f
8b
cyclodimerizations are known to take place by gold catalysis.
The present [2 + 2 + 2] cyclotrimerization is proposed to
proceed via controlled regioselective coupling and cyclization of
three propargyl units through allenic intermediates in the
presence of a Au(I)−nitrone complex (Scheme 2a), in some
accordance with a mechanism suggested for Au-catalyzed
reported Ru−porphyrin nitrone complexes. The linear
(nitrone)−O−Au(I)−P-(phosphane) coordination mode is
clearly verified. The catalytic activity of the isolated crystalline
Au−nitrone 2b complex was tested. The trimerization reaction
rate of acetal 1b and obtained yield of trimer 3b (67%, 5% Au−
nitrone complex, 1.5 h, Table 1, entry 5) were similar to the
reaction with Au−nitrone 2b generated in situ (64%, entry 5).
Subsequent acidic hydrolysis of the enol ether moieties of
trimer 3b (10:4:3 isomer mixture) with TsOH gave the
triketone product 5b (46%, Scheme 3a) as one single isomer.
Complete enolization of triketone 5b was observed after months
of storage. Compounds 3b and 5b were surprisingly stable
4d
cyclotrimerization of allenamides. The overall regioselective
trimerization reaction sequence is driven by the allene activation
provided by the methoxy group. The reaction pathway is
believed to be initiated by the Au-catalyzed 1,3-alkoxy shift,
being the driving force for an intermolecular C2−C1′ bond
formation between (gold-)allenic intermediates. The subse-
C
DOI: 10.1021/acs.orglett.7b00411
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
toward modifications, probably due to the densely substituted
nature of the more flexible cyclohexyl(idene) ring system
Lett. 2015, 56, 3318. (e) Cai, S.; Liu, Z.; Zhang, W.; Zhao, X.; Wang, Z.
D. Angew. Chem., Int. Ed. 2011, 50, 11133. (f) Liu, J.; Chen, M.; Zhang,
L.; Liu, Y. Chem. - Eur. J. 2015, 21, 1009. (g) Pagar, V. V.; Jadhav, A. M.;
Liu, R.-S. J. Am. Chem. Soc. 2011, 133, 20728. (h) Gung, B. W.; Bailey,
L. N.; Wonser, J. Tetrahedron Lett. 2010, 51, 2251. (i) Gung, B. W.;
Craft, D. T.; Bailey, L. N.; Kirschbaum, K. Chem. - Eur. J. 2010, 16, 639.
(Scheme 4). Several attempts to afford aromatization were
Scheme 4. Transformation of Cyclic Trimeric Compounds 3
(j) Gung, B. W.; Conyers, R. C.; Wonser, J. Synlett 2013, 24, 1238.
(k) Shapiro, N. D.; Toste, F. D. J. Am. Chem. Soc. 2008, 130, 9244.
(l) Shapiro, N. D.; Shi, Y.; Toste, F. D. J. Am. Chem. Soc. 2009, 131,
11654. (m) Zi, W.; Wu, H.; Toste, F. D. J. Am. Chem. Soc. 2015, 137,
3225. (n) Zhang, G. Z.; Zhang, L. M. J. Am. Chem. Soc. 2008, 130,
12598. (o) Navarro, C.; Shapiro, N. D.; Bernasconi, M.; Horibe, T.;
Toste, F. D. Tetrahedron 2015, 71, 5800. (p) Liu, F.; Wang, Y.; Ye, W.;
Zhang, J. Org. Chem. Front. 2015, 2, 221. (q) Marion, S.; Díez-
́ ́
Gonzalez, S.; de Fremont, P.; Noble, A. R.; Nolan, S. P. Angew. Chem.,
unsuccessful. Hence, neither heating nor treatment of 3b with
Int. Ed. 2006, 45, 3647.
4c
9
NaOH (80 °C, 170h), I , HOAc or CH NO /Pd/C (90 °C)
2
3
2
(2) (a) Sperger, C. A.; Tungen, J. E.; Fiksdahl, A. Eur. J. Org. Chem.
2011, 2011, 3719. (b) Iqbal, N.; Sperger, C. A.; Fiksdahl, A. Eur. J. Org.
Chem. 2013, 2013, 907. (c) Iqbal, N.; Fiksdahl, A. J. Org. Chem. 2013,
78, 7885. (d) Siah, M. H.-S.; Kaur, M.; Iqbal, N.; Fiksdahl, A. Eur. J. Org.
Chem. 2014, 2014, 1727. (e) Siah, M. H.-S.; Hogsnes, M. C.; Iqbal, N.;
Fiksdahl, A. Tetrahedron 2016, 72, 1058. (f) Evjen, S.; Fiksdahl, A. Eur.
j. Org. Chem. 2016, 2016, 2858. (g) Evjen, S.; Fiksdahl, A. Tetrahedron
gave any conversion at all. Likewise, dehydrogenation/oxidation
10
of 5b with hypervalent iodine(V) failed to give any reaction.
In conclusion, we have shown that 1,3-diarylpropargyl acetals
undergo regio-/chemoselective Au(I)−nitrone-catalyzed [2 + 2
+
2] cyclotrimerization to afford cyclohexylidene products (up
to 74% yield). The presence of (catalytic amounts of) different
nitrones was essential for successful selective cyclotrimerization.
The crystalline phosphane−Au(I)−nitrone 2a complex per-
formed similar catalytic activity as the corresponding Au(I)−
nitrone 2a catalyst formed in situ. X-ray analysis of the Au(I)−
nitrone 2a complex confirmed the linear nitrone−O−Au(I)−P
coordination mode of the crystalline catalyst. Acidic hydrolysis
of the enol ether moieties of trimer 3b afforded triketone 5b.
The [2 + 2 + 2]-cyclotrimerization approach readily allows
chemoselective preparation of densely substituted and poly-
functionalized cyclohexyl(idene) products in the presence of
Au(I)−nitrone complexes, which represent an interesting group
of Au(I) catalysts with specific properties. Further studies on
cyclotrimerizations as well as Au(I)−nitrone complexes are in
progress in our laboratories.
2
016, 72, 3270.
3) Su, Y.; Zhang, Y.; Akhmedov, N. G.; Petersen, J. L.; Shi, X. Org.
Lett. 2014, 16, 2478.
4) (a) Broere, D. L. J.; Ruijter, E. Synthesis 2012, 44, 2639. (b) Leyva-
Perez, S. A.; Oliver-Meseguer, J.; Cabrero-Antonino, J. R.; Rubio-
Marques, P.; Serna, P.; Al-Resayes, S. I.; Corma, A. ACS Catal. 2013, 3,
1865. (c) Tao, X.; Kehr, G.; Daniliuc, C. G.; Erker, G. Angew. Chem., Int.
Ed. 2017, 56, 1376. (d) Hernandez-Diaz, C.; Rubio, E.; Gonzalez, J.
Eur. J. Org. Chem. 2016, 2016, 265.
5) (a) Ji, K.; Nelson, J.; Zhang, L. Beilstein J. Org. Chem. 2013, 9,
925. (b) Lu, B.; Li, C.; Zhang, L. J. Am. Chem. Soc. 2010, 132, 14070.
c) Davies, P. V.; Cremonesi, A.; Martin, N. Chem. Commun. 2011, 47,
79. (d) Xu, M.; Ren, T.-T.; Li, C.-Y. Org. Lett. 2012, 14, 4902.
6) (a) Mauleon, P.; Toste, F. D. In Modern Gold Catalyzed
Synthesis;Hashmi, A. S. K., Toste, F. D., Eds.; Wiley, 2012; p 75.
b) Wang, D.; Zhang, Y.; Cai, R.; Shi, X. Beilstein J. Org. Chem. 2011, 7,
014.
7) (a) Faustino, H.; Bernal, P.; Castedo, L.; Lopez, F.; Mascaren
L. Adv. Synth. Catal. 2012, 354, 1658. (b) Suarez-Pantiga, S.;
Hernandez-Díaz, K.; Rubio, E.; Gonzalez, J. M. Angew. Chem., Int. Ed.
012, 51, 11552. (c) Bernal-Albert, P.; Faustino, H.; Gimeno, G.;
Asensio, G.; Mascarenas, J. L.; Lopez, F. Org. Lett. 2014, 16, 6196.
(
(
́
́
́
(
1
(
3
(
(
1
(
ASSOCIATED CONTENT
Supporting Information
̃
as, J.
■
*
S
́
́
́
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.7b00411.
2
̃
́
Full experimental details, characterization of compounds
(d) Jia, M.; Monari, M.; Yang, Q.-Q.; Bandini, M. Chem. Commun.
2015, 51, 2320. (e) Li, X.-X.; Zhu, L.-L.; Zhou, W.; Chen, Z. Org. Lett.
1
13
(
(
1a−d, 3a−d, 4 and 5b), and H and C NMR spectra
2
012, 14, 436. (f) Suar
M.; Rubio, E.; Gonzalez, J. M. Adv. Synth. Catal. 2012, 354, 1651.
8) (a) Ade, A.; Cerrada, E.; Contel, M.; Laguna, M.; Merino, P.;
́ ́
ez-Pantiga, S.; Hernandez-Díaz, K.; Piedrafita,
PDF)
́
(
AUTHOR INFORMATION
Corresponding Author
■
Tejero, T. J. Organomet. Chem. 2004, 689, 1788. (b) Lee, J.; Twamley,
B.; Richter-Addo, G. B. Chem. Commun. 2002, 380. (c) Kliegel, W.;
Metge, J.; Rettig, S. J.; Trotter, J. Can. J. Chem. 1998, 76, 389.
*E-mail: anne.fiksdahl@chem.ntnu.no.
(d) Kliegel, W.; Metge, J.; Rettig, S. J.; Trotter, J. Can. J. Chem. 1997,
ORCID
7
5, 1830.
Anne Fiksdahl: 0000-0003-2577-2421
Notes
(9) Cossy, J.; Belotti, D. Org. Lett. 2002, 4, 2557.
(10) Nicolaou, K. C.; Montagnon, T.; Baran, P. S.; Zhong, Y.-L. J. Am.
Chem. Soc. 2002, 124, 2245.
The authors declare no competing financial interest.
Crystallographic data for Au(I)−nitrone 2a complex and
products 3b and 3c (CCDC nos. 1531556, 1531557, and
1531558) have been deposited with the Cambridge Crystallo-
graphic Data Centre.
REFERENCES
■
(
1) (a) Zhang, L. M. J. Am. Chem. Soc. 2005, 127, 16804.
b) Obradors, C.; Leboeuf, D.; Aydin, J.; Echavarren, A. M. Org. Lett.
(
2
013, 15, 1576. (c) Conyers, R. C.; Gung, B. W. Chem. - Eur. J. 2013,
19, 654. (d) Conyers, R. C.; Barnes, C. L.; Gung, B. W. Tetrahedron
D
DOI: 10.1021/acs.orglett.7b00411
Org. Lett. XXXX, XXX, XXX−XXX