Gallium-Catalyzed Scriabine Reaction

Article abstract of DOI:10.1021/acs.orglett.8b03104

γ-Aryl enol acetates are easily obtained from diacetoxy alkenes and electron-rich arenes at room temperature using GaCl₃ as catalyst. The products can then be converted into β-aryl aldehydes. This method represents the first broadly applicable catalytic version of the Scriabine reaction. DFT computations shed light on the mechanism of this transformation.

Full text of DOI:10.1021/acs.orglett.8b03104

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
Gallium-Catalyzed Scriabine Reaction
Manish Pareek,† Christophe Bour,*,† and Vincent Gandon*,†,‡
†
́
́
́
́
Institut de Chimie Moleculaire et des Materiaux d’Orsay, CNRS UMR 8182, Universite Paris-Sud, Universite Paris-Saclay,
̂
Batiment 420, Orsay 91405 cedex, France
‡
́
́
Laboratoire de Chimie Moleculaire (LCM), CNRS UMR 9168, Ecole Polytechnique, Universite Paris-Saclay, route de Saclay,
Palaiseau 91128 cedex, France
S
* Supporting Information
ABSTRACT: γ-Aryl enol acetates are easily obtained from diacetoxy alkenes and electron-rich arenes at room temperature
using GaCl3 as catalyst. The products can then be converted into β-aryl aldehydes. This method represents the first broadly
applicable catalytic version of the Scriabine reaction. DFT computations shed light on the mechanism of this transformation.
Scheme 1. Scriabine Synthesis of Cyclamen Aldehyde
β-Aryl-saturated aldehydes (dihydrocinnamaldehydes) are of
fundamental interest in organic chemistry, as they are used as
synthons in the preparation of natural products and
pharmaceuticals or as ingredients of fragrances.1 Various
strategies have been reported to synthesize such compounds,
but the most obvious route, the Michael addition of arenes to
α,β-unsaturated aldehydes, remains a difficult task (eq 1).2
It is mostly limited to highly nucleophilic heteroarenes
(indoles, pyrroles, furans), aniline derivatives, and arenes
displaying directing groups for C−H activation.3 For other
arenes, efficient approaches are scarce and involve conditions
that prevent the polymerization of the conjugated aldehyde
such as the use of H-type zeolites as catalysts.2 This
polymerization issue has long been described. In 1961, Igor
Scriabine reported that dihydrocinnamaldehydes could be
obtained in low yields by condensing at low temperature (−10
°C to −40 °C) electron-rich benzenes to α,β-unsaturated
aldehydes, using the arene as solvent, an equimolar amount of
TiCl4 compared to the aldehyde, and 2 mol % of BF3·OEt2.4
An example of this approach to the synthesis of cyclamen
aldehyde, obtained in 36% yield, is shown in Scheme 1 (eq 2).
Interestingly, in the same paper, Scriabine demonstrated that
much better yields could be obtained by replacing the α,β-
unsaturated aldehyde by its diacetoxy derivative (eq 3).
However, the reaction conditions remained essentially the
same, with a large excess of arene used as solvent, a slight
excess of TiCl4 (1.1 equiv), and 2 mol % of BF3·OEt2.
synthesis of para-substituted aromatic aldehydes by condensa-
tion of substituted benzenes, used as solvent, and methacrolein
diacetyl acetal, was promoted by 1.5 equiv of TiCl4 at −10 °C.6
This need for a stoichiometric or an excess amount of Lewis
acid suggests that the two acetates must be activated and that
their basicity precludes the regeneration of the catalyst. In this
regard, we have previously reported that the use of calcium-7
and gallium-derived Lewis acids8 could catalyze the con-
densation of relatively weak nucleophiles, such as arenes, to
carbon−carbon multiple bonds, even in the presence of
strongly coordinating functionalities that tend to trap metal
ions. Thus, we decided to turn our attention to this overlooked
way of producing β-aryl-saturated aldehydes from diacetoxy
alkenes. Herein, we report the first broadly applicable catalytic
version of the Scriabine reaction,9 as well as mechanistic
considerations regarding the activation of such substrates
obtained by DFT computations.
Proper reaction conditions were developed using diacetox-
ypropene 1a and 1.5 equiv of anisole (Table 1). The use of the
Due to the interest of the perfume industry for this reaction,
which leads, for instance, to cyclamen or lily aldehydes, a
number of patents have been deposited with attempts to turn
this reaction catalytic, but the claims are focused on specific
compounds and seem to lack generality.5 Still in 2015, the
Received: September 28, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b03104
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Table 1. Screening of the Reaction Conditions
Table 2. Scope of the GaCl3-Catalyzed Scriabine Reaction
a
entry sub.
R1
R2
prod.
R
yield (%)
0
entry
catalyst
solvent
conv (yield) (%)
b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1a
1b
1c
1d
1e
1f
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
Me
-
H
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Ca(NTf2)2/nBu4PF6
Ca(NTf2)2/nBu4PF6
Ga(OTf)3
GaCl3
toluene
HFIP
toluene
toluene
toluene
toluene
toluene
1,2-DCE
1,2-DCE
THF
MeNO2
PhF
63
99 (24)
60
75
60
60
70
99 (66)
99 (35)
b
e
2b/2b′
2c
2a
2d/2d′
2e
2f
2g
2h
2i
2j
4-Me/2-Me
4-iPr
54
f
b
63
g
b
4-OMe
83
c
h
b
3,4-Me2/2,3-Me2
2,5-Me2
55
In(OTf)3
i
b
74
49
InCl3
FeCl3
GaCl3
a
j
b
2-OMe, 5-CO2Me
3-Me, 4-OMe
3,4-(OMe)2
2,4,6-Me3
β-naphthyl
4-Ph
g n
,
79
88
78
ko
,
c
IPr·GaCl3 /AgSbF6
g p
,
b
GaCl3
GaCl3
GaCl3
GaCl3
GaBr3
GaI3
99
a
a
jq
52 ,
99 (29)
99 (74)
99 (83)
99 (78)
99 (70)
99 (44)
g q
,
2k
44
75
87
64
81
82
89
d
g q
,
H
2l/2l′
2m
2n
2o
2p
2-thienyl/3-thienyl
4-OMe
-
-
-
-
l
Me
Et
nPr
H
j
4-OMe
4-OMe
4-OMe
4-OMe
m
g
g
Ga2Cl4
a
b
c
GC. Complex mixture. IPr = 1,3-bis(2,6-diisopropylphenyl)-
Et
2q
imidazol-2-ylidene.
a
When the nucleophiles are solids (entries 11 and 12), 250 μL of PhF
b
c
d
was added. Isomeric ratio: 1.2:1. Isomeric ratio: 3:1. Isomeric ratio
6:1. E/Z ratio: 4.1:1/1.9:1. E/Z ratio: 2.1/1. E/Z ratio: >20:1. E/
e
f
g
h
Ca(NTf2)2/nBu4PF6 catalytic mixture10 led to an incomplete
conversion of 1a at rt after 1 h and to a complex mixture of
products (entry 1). Full conversion was reached in
hexafluoroisopropanol (HFIP),7c yet the desired enol acetate
2a was isolated in a low 24% yield (entry 2). Various Lewis
acids of the gallium, indium, and iron series also proved
nonselective in toluene (entries 3−7). In 1,2-dichloroethane,
GaCl3 allowed the synthesis of 2a in 66% yield (entry 8). The
more electrophilic [IPr·GaCl2][SbF6] complex11 was found to
be a less selective catalyst (entry 9). Using GaCl3 in THF or
MeNO2 was not fruitful (entries 10 and 11), but in PhF, a
good 74% yield was obtained (entry 12).
In the absence of solvent, the yield could be increased to
83% provided 4 equiv of anisole was used (entry 13). Under
such conditions, other gallium halides led to lower yields
(entries 14−16). Thus, the reaction can be efficiently
conducted using GaCl3 as catalyst in PhF with 1.5 equiv of
nucleophile or directly in 4 equiv of the nucleophile.
We chose to use the “solvent-free” conditions to study the
scope of the GaCl3-catalyzed reaction, which is shown in Table
2. The E/Z ratio of the enol acetates is indicated, but since
they are meant to be converted into the corresponding
aldehydes,4,12 this point is not essential. Diacetoxypropene 1a
was first used as acceptor (entries 1−13). While no reaction
was observed with deactivated aromatics13 and benzene (entry
1), toluene led to the para and ortho isomers of the desired
enol acetate in a 1.2:1 ratio (entry 2). Only the para isomer
was obtained using cumene or anisole (entries 3 and 4). The
83% yield obtained with anisole is actually the same as the one
reported by Onaka using a zeolite as catalyst.2 The zeolite-
catalyzed reaction leads to the aldehyde directly, yet as an
85:15 para/ortho mixture of regioisomers, so both approaches
have their own advantages. Disubstituted benzenes (entries 5−
9) and a trisubstituted one (entry 10) also furnished the
expected products. With the solid nucleophiles methyl p-
i
j
k
Z ratio: 6:1/2.7:1. E/Z ratio: 6.7:1. E/Z ratio: 4:1. E/Z ratio: 13:1.
l
m
n
E/Z ratio: 4.7:1. E/Z ratio: 3.9:1. Using 1.5 equiv of Ar−H and
o
PhF as solvent (0.3 M): 77%. Using 1.5 equiv of Ar−H and PhF as
p
solvent (0.3 M): 69%. Using 1.5 equiv of Ar−H and PhF as solvent
q
(0.3 M): 67%. A small amount of dialkylation product has been
isolated as side compound.
anisate, naphthalene, and biphenyl, a low amount of PhF was
added to solubilize the heterogeneous mixture (250 μL)
without preventing the reaction (entries 7, 11, and 12). No
poisoning of the catalyst took place with thiophene, which led
to the 2-thienyl isomer of the enol acetate as the major one
(entry 13). Substituted diacetoxypropenes 1b−f were next
reacted with anisole (entries 14−18), and they also proved
compatible with the reaction conditions. Overall, the products
were isolated with good purity in 49−88% yields. These yields
are comparable to those reported by Scriabine for 2a (85%)
and 2p (88%), lower for 2c (84%), and higher for 2b/2b′
(20%) and 2h (62%).4 The regioselectivity can be easily
controlled by electronic and steric factors, except in the case of
toluene (entry 2) and o-xylene (entry 5). It is worth noting
that the proportion of arene can be lowered to 1.5 equiv when
PhF is used as solvent with somewhat lower, yet appreciable,
yields (entries 8−10, 77%, 69%, 67%, respectively).
As an application of the title reaction, we undertook the
synthesis of dianethole, a natural compound found in fennel
and anise,14 the synthesis of which has not been reported in
the literature (Scheme 2). Starting from 1.5 g of 1b, compound
2m was isolated in 66% yield. The decrease in yield compared
to Table 2, entry 14, can be explained by an increase of
temperature when working at such a scale. Methanolysis12 led
to the β-aryl aldehyde 3m, which decomposes rapidly when
stored at rt. It was rapidly engaged in a Horner−Wadsworth−
Emmons reaction, which provided dianethole in 42% yield.
B
DOI: 10.1021/acs.orglett.8b03104
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
equivalent to a Michael addition of simple arenes to α,β-
unsaturated aldehydes, which is very difficult to achieve under
homogeneous conditions. DFT computations suggests that a
double activation of the diacetoxy alkene by two molecules of
Lewis acid would be more efficient than with just one. In spite
of the strongly coordinating nature of the substrate, the use of
GaCl3 as catalyst allows the reaction to turn over.
Scheme 2. Gram-Scale Reaction, Methanolysis, and
Application to the Synthesis of Dianethole
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.8b03104.
1
Experimental procedures, characterization data, H and
13C NMR spectra for all new compounds, coordinates,
and energy of the computed species (PDF)
To rationalize the catalytic role of GaCl3 in the title reaction,
DFT computations were carried out using the Gaussian 09
software package, the ωB97XD functional, and the 6-31+G**
basis set for all atoms. The full energy profile is provided in the
Supporting Information. The teaching of this study is that the
approach of anisole to the terminal carbon of diacetoxypro-
pene 1a complexed to one or two GaCl3 units by its carbonyl
groups promotes the elimination of GaCl3(OAc)− (Figure 1).
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: vincent.gandon@u-psud.fr.
*E-mail: christophe.bour@u-psud.fr.
ORCID
Christophe Bour: 0000-0001-6733-5284
Vincent Gandon: 0000-0003-1108-9410
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This project was funded by an ANR grant (ANR-15-CE07-
0003). We thank Julien Coulomb (Firmenich SA, Corporate
R&D Division, Geneva, Switzerland) and David Leboeuf
(Paris-Sud University) for useful discussions.
REFERENCES
■
́
(1) See inter alia: (a) Alacid, E.; Najera, C. Adv. Synth. Catal. 2007,
349, 2572. (b) Chen, M.; Wang, J.; Chai, Z.; You, C.; Lei, A. Adv.
Synth. Catal. 2012, 354, 341. (c) Vellakkaran, M.; Andappan, M. M.
S.; Kommu, N. Green Chem. 2014, 16, 2788 and the references
therein. .
(2) See: Hayashi, D.; Narisawa, T.; Masui, Y.; Onaka, M. Bull. Chem.
Soc. Jpn. 2016, 89, 460 and the references therein. .
(3) For selected examples, see: (a) Austin, J. F.; MacMillan, D. W. C.
J. Am. Chem. Soc. 2002, 124, 1172. (b) Paras, N. A.; MacMillan, D. W.
Figure 1. Computed free energy profile (ΔG298, kcal/mol) and
́
C. J. Am. Chem. Soc. 2002, 124, 7894. (c) Bah, J.; Franzen, J. Chem. -
transition states (selected distances in Å).
Eur. J. 2014, 20, 1066. Metal-catalyzed C−H activation: (d) Yang, L.;
Correia, C. A.; Li, C.-J. Org. Biomol. Chem. 2011, 9, 7176. (e) Zhang,
Z.; Tang, M.; Han, S.; Ackermann, L.; Li, J. J. Org. Chem. 2017, 82,
664. (f) Li, J.; Zhang, Z.; Ma, W.; Tang, M.; Wang, D.; Zou, L.-H.
Adv. Synth. Catal. 2017, 359, 1717. Lewis acid catalysis: (g) Carmona,
This step is much more efficient with two GaCl3 instead of
one, with the barriers being 17.0 and 25.6 kcal/mol,
respectively. Besides, the addition of anisole is strongly
endergonic with only one GaCl3 (13.3 kcal/mol instead of
−5.3 kcal/mol with two GaCl3). GaCl3(OAc)− can then serve
as base to deprotonate the resulting Wheland-type inter-
mediate at a low free energy cost of 3.2 kcal/mol. From the
final products complexed to GaCl3, going back to 1a·2GaCl3 is
exergonic by 11.9 kcal/mol (not shown), which corroborates
the catalytic activity of GaCl3.
In conclusion, we have developed catalytic conditions that
allow the synthesis of γ-aryl enol acetates from diacetoxy
alkenes and of the β-aryl-saturated aldehydes thereof. This
approach avoids the use of a very large excess of nucleophile
and a stoichiometric amount of TiCl4. It is synthetically
́
D.; Lamata, M. P.; Sanchez, A.; Viguri, F.; Oro, L. A. Tetrahedron:
Asymmetry 2011, 22, 893. (h) Hu, X.; Martin, D.; Melaimi, M.;
Bertrand, G. J. Am. Chem. Soc. 2014, 136, 13594. (i) Li, W.; Werner,
́
T. Org. Lett. 2017, 19, 2568. (j) Liu, Z.; Vidovic, D. J. Org. Chem.
2018, 83, 5295.
(4) Scriabine, I. Bull. Soc. Chim. Fr. 1961, 1194.
(5) See inter alia: (a) Snowden, R.; Birkbeck, A.; Womack, G.
WO2006/120639 (Firmenich), 2006. (b) Shirai, M.; Yoshida, Y.;
Sadaike, S. EP 1 574 509 (UBE), 2005. (c) Doi, T.; Yoshida, Y.;
Sajiki, E.; Fujitsu, S. US 2010/0029541 (to UBE), 2010.
(6) Zhou, C.; Su, W. J. Chem. Res. 2015, 39, 555.
(7) (a) Morcillo, S. P.; Lebœuf, D.; Bour, C.; Gandon, V. Chem. -
Eur. J. 2016, 22, 16974. (b) Lebœuf, D.; Marin, L.; Michelet, B.;
Perez-Luna, A.; Guillot, R.; Schulz, E.; Gandon, V. Chem. - Eur. J.
C
DOI: 10.1021/acs.orglett.8b03104
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
2016, 22, 16165. (c) Qi, C.; Hasenmaile, F.; Gandon, V.; Leboeuf, D.
ACS Catal. 2018, 8, 1734.
(8) (a) Michelet, B.; Thiery, G.; Bour, C.; Gandon, V. J. Org. Chem.
2015, 80, 10925. (b) Michelet, B.; Tang, S.; Thiery, G.; Monot, J.; Li,
H.; Guillot, R.; Bour, C.; Gandon, V. Org. Chem. Front. 2016, 3, 1603.
(9) A report from Firmenich SA mentions some attempts to achieve
the Lewis acid-catalyzed addition of santene to allylidene diacetate,
but the yields were low. See: Birkbeck, A. A.; Marquet, X.; Millet, P.;
Pamingle, H. Eur. J. Org. Chem. 2014, 2014, 7582.
(10) (a) Begouin, J.-M.; Niggemann, M. Chem. - Eur. J. 2013, 19,
8030. (b) Lebœuf, D.; Gandon, V. Synthesis 2017, 49, 1500.
(11) Tang, S.; Monot, J.; El-Hellani, A.; Michelet, B.; Guillot, R.;
Bour, C.; Gandon, V. Chem. - Eur. J. 2012, 18, 10239.
(12) See for instance: Archambeau, A.; Nguyen, D.-V.; Meyer, C.;
Cossy, J. Chem. - Eur. J. 2016, 22, 6100.
(13) No reaction took place with (trifluoromethyl)benzene, methyl
benzoate, or benzonitrile.
(14) (a) Albert-Puleo, M. J. Ethnopharmacol. 1980, 2, 337.
(b) Jordan, V. C. Cancer Res. 2009, 69, 1243.
D
DOI: 10.1021/acs.orglett.8b03104
Org. Lett. XXXX, XXX, XXX−XXX