Spatiotemporal Control of Pre-existing Alkene Geometry: A Bio-Inspired Route to 4-Trifluoromethyl-2H-chromenes

Article abstract of DOI:10.1021/acs.orglett.7b03859

Routes to prepare C4-trifluoromethyl analogues of the 2H-chromene scaffold are scarce: this is particularly striking given the importance of fluorine in pharmaceutical development. To address this limitation, a facile strategy has been developed that is reliant on catalytic, geometric isomerization of easily accessible allylic alcohols (up to >95:5) followed by intramolecular cyclization via Pd catalysis (up to 96%). This concise biomimetic approach emulates the photoisomerization/cyclization cascade inherent to phenylpropanoid biosynthesis.

Full text of DOI:10.1021/acs.orglett.7b03859

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Spatiotemporal Control of Pre-existing Alkene Geometry: A Bio-
Inspired Route to 4‑Trifluoromethyl‑2H‑chromenes
Svenja I. Faßbender,† Jan B. Metternich,† and Ryan Gilmour*
Organisch-Chemisches Institut, Westfalische Wilhelms-Universitat Munster, Corrensstr. 40, 48149 Munster, Germany
̈
̈
̈
̈
S
* Supporting Information
ABSTRACT: Routes to prepare C4-trifluoromethyl analogues
of the 2H-chromene scaffold are scarce: this is particularly
striking given the importance of fluorine in pharmaceutical
development. To address this limitation, a facile strategy has
been developed that is reliant on catalytic, geometric
isomerization of easily accessible allylic alcohols (up to
>95:5) followed by intramolecular cyclization via Pd catalysis
(up to 96%). This concise biomimetic approach emulates the photoisomerization/cyclization cascade inherent to
phenylpropanoid biosynthesis.
patiotemporal control of pre-existing alkene geometry is a
S
challenging problem in organic synthesis, particularly when
the efficiency of subsequent transformations is predicated on a
predefined configuration.1 An amalgamation of thermodynamics
and microscopic reversibility further compounds this problem,
resulting in a general deficiency in the synthetic arsenal for this
fundamental transformation. Unlike positional isomerization,2
there is a conspicuous absence of general catalysts to affect
geometric isomerization in a specific direction.3 However, the
current renaissance of organic photochemistry provides a
platform for development of efficient methods to facilitate
selective alkene isomerization at a defined stage of a synthetic
sequence: light-induced spatiotemporal control.4 This is
attributable to the initial excitation process that circumvents
the principle of microscopic reversibility5 and ensures that the
subsequent reaction coordinate is energetically downhill.
Seminal mechanistic insights into photocatalytic alkene isomer-
ization of styrenes and stilbenes6−8 provide expansive blueprints
from which to develop synthetically powerful methods. While
the direct excitation of alkenes is possible, photosensitization
using a small-molecule catalyst is often more efficient and allows
for selective activation of a specific functional group. This
strategy has been effectively demonstrated using an Ir(III)
complex9 and by this laboratory using (−)-riboflavin to expedite
the E→Z isomerization of activated olefins via an energy transfer
manifold.10 In this latter case, initial isomerization can be coupled
to a second step exploiting photoinduced single-electron transfer
to oxidatively cyclize cinnamic acids directly to coumarins.11 This
process mimics the initial stage of coumarin biosynthesis via the
phenylpropanoid pathway.12
Figure 1. General, biomimetic strategy based on coumarin biosynthesis
and selected natural 2H-chromene derivatives.
To validate this biomimetic strategy, 4-trifluoromethylated
analogs were identified as being of particular interest due to
the prominence of fluorine in bioactive molecule design (Figure
2).15 A structure search revealed that relatively few examples of
CF3-containing 2H-chromenes have been reported;16 this is in
stark contrast with the corresponding 4-Et scaffold which
Because sequential isomerization/cyclization processes are
pervasive in Nature (Figure 1, top),13 it was envisaged that a
photocatalytic isomerization approach may facilitate biomimetic
synthesis of a structurally related 2H-chromene scaffold. This
motif constitutes the core of a plethora of bioactive natural
products and pharmaceuticals (Figure 1, bottom), thus develop-
ment of concise synthetic routes has been intensely pursued.14
Received: December 11, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b03859
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Figure 3. Reaction progress monitoring of Z-1 to E-1 by HPLC using a
CHIRACEL OJ H column, hexane/iPrOH = 90:10, 0.5 mL·min−1; tR(Z-
1) = 10.2 min, tR(E-1) = 22.1 min.
alcohols19 suggested that H to CF3 substitution may be
advantageous in suppressing this reactivity. From the reaction
design perspective, intrinsic directionality [Z-1→E-1] would be
conditional on a selective excitation manifold (Figure 2, bottom).
As the starting Z-allylic alcohol is planar, it would differ in
excitation efficiency from that of the product E-allylic alcohol.
This is due to 1,3-allylic strain20 in the product distorting the π-
system and reducing conjugation. The consequence is that Z-1
can be excited to the postulated diradical intermediate,10a which
can collapse to the starting material or product. Because the
reverse reaction is inefficient due to the loss of conjugation in the
chromophore, high selectivity for E-1 is expected. As triplet
energy transfer (ET) was expected to be operational in the initial
isomerization, common sensitizers were systematically explored
under irradiation at 365 or 450 nm (Table 1, Z-1→E-1).
Reactions were performed in the solvent indicated for 24 h.
Benzophenone (entry 2) and anthracene (entry 6) were
identified as being highly efficient (94:6 and >95:5, respectively).
Based on considerations of cost and simplicity, the study was
advanced using anthracene (Sigma-Aldrich, 100 g = $55.50).
Acetonitrile was an excellent reaction medium, and catalyst
loadings as low as 2 mol % continued to deliver excellent Z/E
ratios (>95:5) as determined by 1H NMR analysis of the crude
mixture (entry 9). Ir(ppy)3 featured prominently in the study by
Weaver9 was ineffective in the isomerization of Z-1, proving a
degree of complementarity to these studies.
Figure 2. Top: Sci-Finder hits for the three chromene systems indicated
(performed 11/30/2017). Center: A conceptual overview of this study.
Bottom: Design principles employed in the initial isomerization.
provides a useful benchmark due to similar steric volumes of CF3
and Et (39.8 and 38.9 Å3, respectively) (Figure 2, top).17 A two-
step sequence reliant on the efficient isomerization of an easily
prepared Z-configured allylic alcohol,18 and concomitant
cyclization, provides facile access to 4-CF3 2H-chromenes.
Selective excitation of the planar starting material would be
ensured due to 1,3-allylic strain that would manifest itself in the
twisted, deconjugated chromophore of the product. This would
endow the reaction with an intrinsic directionality.
While this was sufficient impetus for reaction development, a
study describing the photosensitized degradation of cinnamyl
Table 1. Optimization of the Photocatalytic Isomerization
c
entry
catalyst
ET (kJ/mol)
catalyst loading (mol %)
solvent
scale (mmol)
E/Z ratio
a
1
2
3
4
5
6
7
8
9
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0:100
a
benzophenone
Ir(ppy)3
286.6
235.6
223.0
190.3
178.6
170.8
144.7
178.6
5
5
5
5
5
5
5
2
94:6
b
<5:95
a
fluorenone
eosin Y
<5:95
b
16:84
a
anthracene
Rose Bengal
methylene blue
anthracene
>95:5
b
<5:95
b
<5:95
a
>95:5
a
b
Reactions were performed at ambient temperature under UV light irradiation (365 nm) for 24 h. Reactions were performed at ambient
c
1
temperature under vis light irradiation (450 nm) for 24 h. Determined by H NMR spectroscopy.
B
DOI: 10.1021/acs.orglett.7b03859
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
a
Scheme 1. Exploring Substrate Scope
Scheme 3. Concise Synthesis of 2H-Chromenes
a
Reactions were performed with 0.1 mmol of substrate in MeCN
using 5 mol % catalyst loading under UV-light irradiation (365 nm) for
b
24 h at ambient temperature. Reactions were performed with 0.2
mmol of substrate in MeCN using 5 mol % catalyst loading under UV-
light irradiation (365 nm) for 24 h at ambient temperature.
c
d
Determined by 1H NMR spectroscopy. Not determined due to
significant product degradation.
Scheme 2. Exploring the Difluoromethyl Group
a
Isomerization reactions performed with 0.1 mmol of substrate in
MeCN using 5 mol % of anthracene under UV light irradiation (365
nm) for 24 h at ambient temperature. Cyclizations were performed
with 0.15 mmol of substrate in toluene using Pd(OAc)2 (5 mol %),
dppf (5 mol %), and NaOtBu (2.0 equiv). E/Z ratios determined by
1H NMR spectroscopy.
Cognizant of the aforementioned study describing the
photosensitized rearrangements of allylic alcohol derivatives,
reaction progress monitoring of the transformation [Z-1→E-1]
was performed by 19F NMR spectroscopy (see SI for details).
Reaction proceeded smoothly and cleanly, with only signals
corresponding to the substrate (δF = −54.27) and product (δF =
−67.23) observed. Reaction progress analysis by HPLC provided
additional independent verification of the efficient, selective
nature of the isomerization catalyzed by anthracene (Figure 3)
and indicates that the CF3 group effectively impedes photo-
degradation pathways.
To establish the scope and limitations of the method, a series
of trifluoromethylated allylic alcohols (R1 = CF3) were exposed
to the standard isomerization conditions and compared with the
corresponding ethyl systems (R1 = Et) due to their similar steric
volumes (Scheme 1).17 Initially, primary alcohols (R2 = H, 1a
and 1b) were exposed to the general conditions. While the ethyl
derivative gave rise to a 78:22 2b/1b mixture in only 48% yield,
exchanging Et with CF3 enhanced the yield and selectivity
(quant., 2a/1a >95:5). The bulkier gem-dimethyl substrates (1c
and 1d) were processed with excellent E/Z selectivities (>95:5
2/1), but the efficiency of the transformation was heavily
dependent on the nature of R1. Only a 50% material recovery was
recorded in the case of 2d, R1 = Et, whereas the corresponding
CF3 derivative 2c was processed quantitatively. The p-OCH3
derivatives (1e and 1f) were markedly different in their
isomerization efficiency and selectivity (R1 = CF3, quant., 2e/
1e >95:5 vs R1 = Et, yield and E/Z ratio not determined due to
degradation). This variance in yield between R1 = Et and R1 =
CF3 persisted throughout the series with the p-Me (1g vs 1h), p-
Cl (1i vs 1j), and p-F (1k vs 1l) substituted derivatives giving
similar E/Z ratios but markedly different yields (Δ up to 50%).
This supports the notion that the CF3 group occupies a
comparable steric volume to Et, which influences selectivity.10b
Degradative photochemical pathways that plague the carbogenic
scaffolds are suppressed by the CF3 group.
Due to the importance of the difluoromethyl group as an
effective hydrogen bond donor in structural and medicinal
chemistry,21 we explored a model substrate in which R1 = CHF2
(Scheme 2). Isomerization was highly selective (E/Z >95:5) with
89% yield. With an optimized set of isomerization conditions to
C
DOI: 10.1021/acs.orglett.7b03859
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
L.-G.; Yao, Z.-K.; Yu, Z.-X. Org. Lett. 2013, 15, 4634. (c) Chen, C.;
Dugan, T. R.; Brennessel, W. W.; Weix, D. J.; Holland, P. L. J. Am. Chem.
Soc. 2014, 136, 945. (d) Timsina, Y. N.; Biswas, S.; Rajanbabu, T. V. J.
Am. Chem. Soc. 2014, 136, 6215. (e) Larsen, C. R.; Erdogan, G.;
Grotjahn, D. B. J. Am. Chem. Soc. 2014, 136, 1226.
(4) (a) Beharry, A. A.; Woolley, G. A. Chem. Soc. Rev. 2011, 40, 4422.
(b) Merino, E.; Ribagorda, M. Beilstein J. Org. Chem. 2012, 8, 1071.
(c) Gautier, A.; Gauron, C.; Volovitch, M.; Bensimon, D.; Jullien, L.;
Vriz, S. Nat. Chem. Biol. 2014, 10, 533. (d) Hu, Y.; Tabor, R. F.;
Wilkinson, B. L. Org. Biomol. Chem. 2015, 13, 2216.
(5) IUPAC definition of microscopic reversibility: In a reversible
reaction, the mechanism in one direction is exactly the reverse of the
mechanism in the other direction. This does not apply to reactions that
begin with a photochemical excitation.
(6) Examples of the cis−trans isomerizations of stilbenes: (a) Ham-
mond, G. S.; Saltiel, J. J. Am. Chem. Soc. 1962, 84, 4983. (b) Hammond,
G. S.; Saltiel, J. J. Am. Chem. Soc. 1963, 85, 2515. (c) Hammond, G. S.;
Saltiel, J.; Lamola, A. A.; Turro, N. J.; Bradshaw, J. S.; Cowan, D. O.;
Counsell, R. C.; Vogt, V.; Dalton, C. J. Am. Chem. Soc. 1964, 86, 3197.
(7) For examples of cis−trans isomerizations of styrenes, see: (a) Arai,
T.; Sakuragi, H.; Tokumaru, K. Chem. Lett. 1980, 9, 261. (b) Arai, T.;
Sakuragi, H.; Tokumaru, K. Bull. Chem. Soc. Jpn. 1982, 55, 2204.
(8) Metternich, J. B.; Gilmour, R. Synlett 2016, 27, 2541.
mimic the photochemical isomerization step in phenylpropanoid
biosynthesis, attention was focused on the final intramolecular
cyclization. Previous mechanistic delineation from this labo-
ratory demonstrated that an ortho-substituent on the aryl ring of
the chromophore is advantageous in ensuring high levels of
stereoselectivity.10b This can be rationalized based on the
augmentation of unfavorable nonbonding interactions in the
product, which further hinders reconjugation of the π-system and
subsequent excitation. Simple ortho-bromo substrates were
prepared which were unlikely to compromise isomerization
efficiency but would allow intramolecular cyclization under Pd
catalysis (Scheme 3). Isomerization selectivities were good to
excellent in almost all cases examined (E/Z >95:5, for p-H, Cl,
and F, E/Z 85:15 for the naphthyl, and E/Z 72:28 for the p-CH3
derivative). Intramolecular cyclization employing Pd(OAc)2 and
dppf, in combination with NaOtBu as a base in toluene (110
°C),22 delivered 4-CF3 2H-chromenes 3n−r in synthetically
useful yields (up to 96%). Importantly, the starting Z-configured
allylic alcohols do not undergo cyclization to afford 2H-
chromenes even at elevated temperatures. Inspired by the
spatiotemporal blueprint of phenylpropanoid biosynthesis, a
direct route to medicinally relevant 2H-chromenes is disclosed.
Efficient and selective isomerization of a pre-existing alkene (E/Z
ratios up to >95:5) is catalyzed by inexpensive, commercially
available anthracene (Sigma-Aldrich, reagent grade 97%, 100 g =
$55.50) and likely proceeds via a triplet energy transfer
manifold.6−8,23 This study contributes to the current interest in
photocatalytic isomerization reactions.24 Installation of the CF3
unit effectively suppresses the photochemical degradation
pathways of allylic alcohols. Combining the initial photocatalysis
step with Pd-mediated cyclization constitutes a rapid approach to
this privileged heterocyclic nucleus.
(9) Singh, K.; Staig, S. J.; Weaver, J. D. J. Am. Chem. Soc. 2014, 136,
5275.
(10) (a) Metternich, J. B.; Gilmour, R. J. Am. Chem. Soc. 2015, 137,
11254. (b) Metternich, J. B.; Artiukhin, D. G.; Holland, M. C.; von
Bremen-Kuhne, M.; Neugebauer, J.; Gilmour, R. J. Org. Chem. 2017, 82,
̈
9955.
(11) Metternich, J. B.; Gilmour, R. J. Am. Chem. Soc. 2016, 138, 1040.
(12) Yao, R.; Zhao, Y.; Liu, T.; Huang, C.; Xu, S.; Sui, Z.; Luo, J.; Kong,
L. Plant Mol. Biol. 2017, 95, 199.
(13) Matern, U.; Luerm, P.; Kreusch, D. Biosynthesis of Coumarins. In
̈
Comprehensive Natural Product Chemistry, Vol. 1: Polyketides and Other
Secondary Metabolites; Sankawa, U., Ed.; Pergamon: Oxford, 1999.
(14) Majumdar, N.; Paul, N. D.; Mandal, S.; de Bruin, B.; Wulff, W. D.
ACS Catal. 2015, 5, 2329.
ASSOCIATED CONTENT
* Supporting Information
■
(15) (a) Purser, S.; Moore, P. R.; Swallow, S.; Gouverneur, V. Chem.
Soc. Rev. 2008, 37, 320. (b) Meanwell, N. A. J. Med. Chem. 2011, 54,
2529.
(16) SciFinder Scholar, 2017; SciFinder search was performed with
substructure search type and exact specification of the chromene 4-
position substitution (accessed November 30, 2017).
(17) Leroux, F. ChemBioChem 2004, 5, 644.
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.7b03859.
Experimental procedures and characterization data (PDF)
(18) Substrates were prepared via a nonselective Horner−Wads-
worth−Emmons reaction and separated (see SI). Only the Z-isomer was
used for the study to facilitate analysis.
(19) Lee, G. A.; Israel, S. H. J. Org. Chem. 1983, 48, 4557.
(20) Hoffmann, R. W. Chem. Rev. 1989, 89, 1841.
(21) Zafrani, Y.; Yeffet, D.; Sod-Moriah, G.; Berliner, A.; Amir, D.;
Marciano, D.; Gershonov, E.; Saphier, S. J. Med. Chem. 2017, 60, 797.
(22) Arcadi, A.; Cacchi, S.; Fabrizi, G.; Marinelli, F.; Verdecchia, M.
Synlett 2006, 2006, 909.
(23) (a) Dexter, D. L. J. Chem. Phys. 1953, 21, 836. (b) Porter, G.;
Wilkinson, F. Proc. R. Soc. London, Ser. A 1961, 264, 1.
(24) (a) Cai, W.; Fan, H.; Ding, D.; Zhang, Y.; Wang, W. Chem.
Commun. 2017, 53, 12918. (b) Zhan, K.; Li, Y. Catalysts 2017, 7, 337.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: ryan.gilmour@uni-muenster.de.
ORCID
Ryan Gilmour: 0000-0002-3153-6065
Author Contributions
†S.I.F. and J.B.M. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We acknowledge financial support from the WWU Munster,
Deutsche Forschungsgemeinschaft (Excellence Cluster EXC
1003) and Fonds der Chemischen Industrie (Fellowship to
J.B.M.).
■
̈
REFERENCES
(1) Pearson, C. M.; Snaddon, T. ACS Cent. Sci. 2017, 3, 922.
(2) Vasseur, A.; Bruffaerts, J.; Marek, I. Nat. Chem. 2016, 8, 209.
■
(3) E→Z alkene isomerization via organometallic catalysis: (a) Punner,
̈
F.; Schmidt, A.; Hilt, G. Angew. Chem., Int. Ed. 2012, 51, 1270. (b) Zhuo,
D
DOI: 10.1021/acs.orglett.7b03859
Org. Lett. XXXX, XXX, XXX−XXX