Oxoammonium-Mediated Allylsilane–Ether Coupling Reaction

Article abstract of DOI:10.1002/ejoc.202100026

A new C(sp³)?H functionalization reaction consisting of the oxidative α-allylation of allyl- and benzyl- methyl ethers has been developed. The C?C coupling could be carried out under mild conditions thanks to the use of cheap and green oxoammonium salts. The scope of the reaction was studied over 27 examples, considering the nature of the substituents on the two coupling partners.

Full text of DOI:10.1002/ejoc.202100026

Full Papers
doi.org/10.1002/ejoc.202100026
1
2
3
4
5
6
7
8
9
Oxoammonium-Mediated Allylsilane–Ether Coupling
Reaction
Federica Carlet,[a, b] Greta Bertarini,[a, b] Gianluigi Broggini,[b] Alexandre Pradal,*[a] and
Giovanni Poli*[a]
Dedicated to Professor Franco Cozzi in honor of his 70th birthday.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
A new C(sp3)À H functionalization reaction consisting of the
oxidative α-allylation of allyl- and benzyl- methyl ethers has
been developed. The CÀ C coupling could be carried out under
mild conditions thanks to the use of cheap and green
oxoammonium salts. The scope of the reaction was studied
over 27 examples, considering the nature of the substituents on
the two coupling partners.
Introduction
Metal-catalyzed C(sp3)À H functionalizations allow the prepara-
tion of increasingly complex scaffolds in a reduced number of
steps,[1] which is in accordance with some of the “green
chemistry” principles. Though metal catalysis has proven to be
very effective,[2] metal-free protocols are not less interesting,
and may offer complementary reactivities. Among the organic
reagents capable of effecting oxidative C(sp3)À H functionaliza-
tion through formal hydride abstraction (Figure 1), 2,3-dichloro-
5,6-dicyano-1,4-benzoquinone (DDQ) has been very often used
for the α-functionalization of amines and ethers.[3]
Figure 1. Examples of oxidizing agents used for C(sp3)À H functionalization
by formal hydride abstraction.
benium ions, as opposed to iminium ions. As a matter of fact,
only a limited number of oxoammonium promoted trans-
formations involving the α-C(sp3)À H functionalization of ethers
(essentially benzyl-, allylethers and isochromanes) has been so
far reported.[8]
In the frame of our current interest in CÀ C coupling
reactions, we focused our recent research investigations on the
oxidative allysilane-ether coupling reaction. Such a CÀ C cou-
pling was pioneered by Hayashi and Mukaiyama in 1987, who
used DDQ as the oxidizing agent in the presence of catalytic
amounts of LiClO4 (Scheme 1, equation 1).[9] This transformation
represents an oxidative variant of the Hosomi-Sakurai allylation
(Scheme 1, equation 2),[10] and it has been so far reported only
in very limited examples, always using DDQ as the oxidant.[11]
Although DDQ has allowed to accomplish a great number
of synthetically relevant transformations,[4] this reagent is rather
toxic,[5] and the development of alternative green and user-
friendly oxidants is highly desirable. Conversely, oxoammonium
⊕
⊖
salts, and in particular, those derived from TEMPO (TEMPO X ,
⊕
⊖
⊕
⊖
alias T X , 1a–1e or 4-AcNHTEMPO BF4 1f) show a very low
toxicity, and lend themselves as ideal and green agents for
oxidative CÀ H bond functionalizations.[6] Most of the studies
carried out on oxidative CÀ C bond formation with these
oxoammonium salts focused on the direct α-C(sp3)À H function-
alization of amines.[7] On the other hand, the development of
the corresponding direct α-C(sp3)À H functionalization of ethers
has progressed more slowly, probably due to the energetically
less accessible oxidation potential needed to generate oxocar-
⊕
In view of the reported ability of TEMPO reagents to
oxidize allyl ethers to the corresponding carbonyl compounds
in the presence of water (Scheme 1, equation 3),[12] we
envisaged to replace water with an allyl nucleophile, with the
aim to end-up with a green and synthetically useful oxoammo-
nium-mediated allylation reaction (Scheme 1, equation 4).
[a] F. Carlet, G. Bertarini, Dr. A. Pradal, Prof. G. Poli
Faculté des Sciences et Ingénierie, CNRS,
Institut Parisien de Chimie Moléculaire, IPCM
Sorbonne Université,
4 place Jussieu, 75005 Paris, France
E-mail: alexandre.pradal@sorbonne-universite.fr
giovanni.poli@sorbonne-universite.fr
[b] F. Carlet, G. Bertarini, Prof. G. Broggini
Dipartimento di Scienza e Alta Tecnologia
Università dell’Insubria
Results and Discussion
We started our investigation by choosing the allylation of
cinnamyl methyl ether 2a as model reaction, and adopting
reaction conditions analogous to those reported by
Leadbeater[12b] in the oxoammonium-promoted oxidation of
allylethers to the corresponding aldehydes, replacing water by
Via Valleggio 11, 22100 Como, Italy
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/ejoc.202100026
Part of the “Franco Cozzi’s 70th Birthday” Special Collection.
Eur. J. Org. Chem. 2021, 2162–2168
2162
© 2021 Wiley-VCH GmbH

Full Papers
doi.org/10.1002/ejoc.202100026
an allylsilane [freshly prepared T BF4 (1a),[13] then dichloro-
to form the final homoallyl ether (Scheme 2). Worthy of note, as
⊕
⊖
1
2
3
4
5
6
7
8
9
⊕
⊖
°
methane (0.2 M) and allyltrimethylsilane (2.0 equiv.), at 45 C in
an allyl silane is expected to react with T BF4 , the kinetics of
generation of the oxocarbenium ion has to be sensibly higher
than that of the subsequent coupling.
a
sealed tube]. Gratifyingly, reacting allyltrimethylsilane
(2.0 equiv.) with the oxoammonium salt 1a, gave the desired
homoallyl ether 4aa in 61% yield after 1 h (Table 1, entry 1),
together with some traces of cinnamaldehyde 5. The mecha-
nism of this reaction is expected to involve hydride transfer
While decreasing the amount of allyltrimethylsilane to 1.5
and 1.2 equivalents diminished the yield of 4aa without
decreasing the yield of cinnamaldehyde 5, further increase in
the loading of the nucleophile to 2.5, 3.0 or 4.0 equivalents did
not give extra improvements in the yield (see Supporting
Information).
⊕
⊖
from the ether to T BF4 to generate an oxocarbenium ion and
TEMPOH (step a),[14] which are likely to be in equilibrium with
the corresponding mixed acetal adduct (step b). Finally, an SE2’
addition of the allylsilane to the oxocarbenium ion is expected
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
We then studied the influence of the solvent and we found
that, of all the solvents tested (dichloromethane, 1,2-dichloro-
ethane, acetonitrile, chloroform, methyl-tert-butylether, 1,4-
dioxane, toluene, diethyl ether, HFIP, acetone), dichloromethane
gave the best results. In addition, changing the concentration
(0.17 M, 0.25 M, 0.5 M and 1 M) did not improve the result (see
Supporting Information).
Performing the reaction at room temperature and stopping
the experiment after 1 h gave a lower conversion of 2a, a lower
yield of 4aa, as well as a higher yield of cinnamaldehyde 5. No
desired product 4aa was formed in 1 h when the reaction was
°
carried out at 0 C (Table 1, entry 3).
Oxoammonium salts bearing other counter-anions [triflate
(1b), bistriflimide (1c), hexafluorophosphate (1d) and chloride
(1e)[15]] were also tested. However, none of them gave better
⊕
⊖
yields in 4aa than T BF4 (Table 1, entries 4–7). The better
efficiency of the oxoammonium salt bearing the tetrafluorobo-
rate counter-anion with regard to the others may be due to its
°
better solubility in dichloromethane at 45 C. Bobbitt’s salt 1f
was found to be insoluble in dichloromethane and did not
provide any conversion of the substrate 2a in this solvent
(Table 1, entry 8).
Scheme 1. Previous works related to the oxoammonium-mediated allysilane-
ether coupling reaction, and the actual work.
Table 1. Optimization of the reaction conditions.
Entry[a]
Z
Hydride abstractor
T [ C]
Conversion [%][b]
Yield 4aa [%][b]
Yield 5 [%][b]
°
1
2
SiMe3
SiMe3
1a
1a
1a
1b
1c
1d
1e
1f
1a
1a
1a
1a
Ph3CBF4
Ph3CBF4
DDQ
45
rt
93
86
64
90
85
>99
25
61
48
0
21
20
37
0
9
18
48
0
3
4
5
6
7
8
9
SiMe3
SiMe3
SiMe3
SiMe3
SiMe3
SiMe3
Sn(n-Bu)3
Sn(n-Bu)3
Si(OEt)3
SiMe3
0
45
45
45
45
45
45
45
45
45
45
45
45
8
14
17
0
0
0
43
0
25
0
28
0
0
0
0
28
31
0
0
67
10[c]
11[d]
12[c]
13
14[c]
15
>99
>99
>99
>99
>99
>99
SiMe3
SiMe3
SiMe3
12
°
[a] Typical reaction conditions: 2a (0.5 mmol), 3a (2.0 equiv.), 1a–1f (1.2 equiv.), solvent (2.5 mL), were heated at 45 C in a sealed tube, 1 h. [b] Yields were
determined by quantitative NMR analysis, using benzoic acid as the internal standard. [c] Substrate 2a and 1a were stirred for 1 h at 45 C in a sealed tube
before introducing the allylating reagent. [d] The reaction was heated under stirring for 2.5 h; rt: room temperature; DDQ: 2,3-Dichloro-5,6-dicyano-1,4-
benzoquinone.
°
Eur. J. Org. Chem. 2021, 2162–2168
www.eurjoc.org
2163
© 2021 Wiley-VCH GmbH

Full Papers
doi.org/10.1002/ejoc.202100026
1
2
3
4
5
6
7
8
9
⊖
Scheme 3. New water-free based protocol to T⊕BF4
.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
PhI(OAc)2 (1.0 equiv.),[18] gave, after further 2 h stirring, the
desired water-free oxoammonium salt 1a in 94% yield. Chloro-
form revealed to be the solvent of choice, as it well dissolves
Scheme 2. Plausible mechanism of the oxoammonium-mediated allysilane-
ether coupling reaction.
⊕
⊖
TEMPO, TEMPOH HBF4, as well as PhI, without dissolving T BF4 .
⊕
⊖
Gratifyingly, use of T BF4 obtained from this new water-free
protocol in the above allylation reaction gave the desired
product 4aa with reproducible 56–59% isolated yields
(Scheme 3) (See Supporting information).
We then inspected the nature of the allylation reagent.
Replacing allyltrimethylsilane for allyltributylstannane[16] did not
afford the desired product 4aa (Table 1, entry 9) but only trace
With the optimized reaction conditions in hand, the scope
of the reaction was then evaluated, starting with examination
of the nature of the ether function linked to the cinnamyl
structure (Scheme 4, top left box). Cinnamyl methyl ether 2a
provided the desired allylated product 4aa in 57% isolated
yield, and scale-up of this reaction to 10 mmol gave 61% yield.
Ethyl and iso-propyl ethers 2b and 2c gave the corresponding
allylated products 4ba and 4ca in 28% and 44% yield,
respectively. On the other hand, tert-butyl ether 2d provided
cinnamaldehyde 5 in 72% yield, and only traces amount of the
desired allylated compound 4da. The anhydrous formation of
the aldehyde 5 from the t-butyl ether 2d can be accounted for
on the basis of a competitive Me3CÀ O bond breaking from the
oxocarbenium ion intermediate, to form t-butyl cation, which in
turn will undergo an E1 elimination to isobutene.[19] The fact
that this alternative reactivity appears only in the case of the t-
butyl ether 2d suggests that the balance “allylation vs CÀ O
bond breaking” is rather delicate and very much dependent on
the stabilization of the carbocation resulting from the CÀ O
bond breaking (Scheme 5).
Treatment of cinnamyl acetate 2e under the above
optimized reactions left the substrate unchanged. This absence
of reactivity is in line with the fact that the oxcarbenium ion
intermediate is destabilized by the acetate function.[14a] Sub-
stitution at the aromatic ring was then evaluated (Scheme 4,
bottom left and top right boxes). The presence of electron poor
substituents on the aromatic ring was tolerated, as shown by
the formation of the p-bromophenyl, p-nitrophenyl and p-
trifluoromethyl adducts 4fa, 4ga and 4ha in moderate to good
yields. The substitution at the ortho, meta and para positions on
the aromatic ring was then evaluated with the p-, o-, and m-
tolyl derivatives, which gave the corresponding allylated
adducts 4ja, 4ka, and 4la in 46%, 51%, and 35% yield,
respectively. A methylenedioxy bridge on the aromatic ring
allowed also to obtain the corresponding allylation product 4ia
in 58% yield.
⊕
⊖
amounts of O-allylTEMPO. Furthermore, treating 2a with T BF4
°
in dichloromethane at 45 C in a sealed tube for 1 h before the
introduction of allyltrimethylstananne – to avoid the possible
undesired reaction between allyltrimethylstannane and the
oxoammonium salt – afforded 4aa in 25% yield, together with
several unidentified byproducts (Table 1, entry 10). Using
allyltriethoxysilane afforded only cinnamylaldehyde 5 in a low
⊕
⊖
yield (Table 1, entry 11). Finally, treating 2a with T BF4
°
(1.2 equiv.) in dichloromethane at 45 C in a sealed tube for 1 h
before the introduction of allyltrimethylsilane (2.0 equiv.)
afforded a roughly equimolar mixture of the desired homoally-
lether 4aa and cinnamylaldehyde 5 in low yields (Table 1,
entry 12).
For the sake of comparison, other hydride abstracting
reagents were also tested. In the event, tritylium tetrafluorobo-
rate generated only allyltriphenylmethane in 31% yield (Table 1,
entry 13), and essentially the same result was obtained by
mixing 2a with tritylium tetrafluoroborate for 1 h before
introducing allyltrimethylsilane (Table 1, entry 14). The use of
DDQ gave mainly cinnamaldehyde 5 (Table 1, entry 15). In the
end, no other hydride abstracting reagent proved to be more
⊕
⊖
efficient than the oxoammonium salt T BF4 for our model
transformation.
⊕
⊖
However, further tests using different batches of T BF4
under our optimized reaction conditions revealed some
irreproducibility issues, the yield of 4aa oscillating between 35
and 61%, with proportionally increasing amounts of cinnamal-
dehyde. Since the protocol we followed to prepare T BF4
performed in water, we suspected that precipitation from this
solvent (see Supporting information) may be the cause of the
formation of trace amounts of undesired TEMPOH HBF4 in the
reagent. To solve this problem, we developed a water-free
[8a]
⊕
⊖
is
⊕
⊖
protocol to prepare T BF4 , employing a purification technique
that excluded the presence of residual TEMPOH HBF4.[17]
After several trials, we found that treating TEMPO with HF4B
The geometry and the substitution of the alkene moiety
was also investigated (Scheme 4, bottom right box).
°
OEt2 (1.1 equiv.) at 0 C in CHCl3 for 1 h, followed by addition of
Eur. J. Org. Chem. 2021, 2162–2168
www.eurjoc.org
2164
© 2021 Wiley-VCH GmbH

Full Papers
doi.org/10.1002/ejoc.202100026
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Scheme 4. Influence of the substitution at the aromatic ring and at the alkene moiety on cinnamyl methyl ether substrates. [a] Typical reaction conditions:
°
2a–p (0.5 mmol), 3a (2.0 equiv.), 1a (1.2 equiv.), CH2Cl2 (2.5 mL), 45 C in a sealed tube, 1–5 h. [b] reaction performed on a 0.28 mmol scale. [c]
Cinnamaldehyde 5 was isolated with 72% yield. [d] No conversion was observed after 24 h of reaction. Trace amounts of allyl-TEMPO were detected by 1H
NMR.
Since oxocarbenium ion A may be so stabilized to be
unproductive, the reaction evolves exclusively toward diene 6
through cation B. In line with the above hypothesis, in situ
spectrometric mass analysis (see SI for details) detected ions
compatible with structures A and B. Repetition of the above
experiment in the absence of the oxoammonium salt 1a,
performed as a blank test, gave no conversion, thus ruling out
an SN1 mechanism for such a reactivity (Scheme 7).
Scheme 5. Allylation vs CÀ O bond breaking as a function of the nature of
the ether.
Substitutions other than the 2-arylpropenyl moiety were
then considered for the starting allylic ether (Scheme 8). Thus,
methyl ether 2p, bearing a highly conjugated phenylpentadien-
yl substituent gave the corresponding adduct 4pa in 38% yield.
(E)-2-Dodecenyl methyl ether 2q could be also allylated,
providing the corresponding allylated product 4qa in 41%
yield.
The ethers 2r and 2s deriving from O-methylation of the
monoterpenes (À )-myrtenol and (À )-perillyl alcohol afforded
the corresponding allylated products 4ra and 4sa as 63:37 and
54:46 diastereoisomeric mixtures and in 68% and 56% yield,
respectively.
In the event, the trisubstituted alkene 2m generated the
expected homoallyl ether 4ma in 34% yield with a small loss of
the E/Z ratio (88:12 starting from >99:1 for 2m). The 1,1-
disubstituted alkene 2-phenyl-3-methoxypropene 2n was un-
reactive. Interestingly, treatment of (E)-1,3-diphenyl-3-meth-
oxypropene 2o with our optimized reaction conditions afforded
the demethoxylated 1,5-diene
(Scheme 6).
6 in 87% isolated yield
Thus, it appears that on passing from primary to secondary
allylic ethers, isohypsic allylation (methoxy departure) competes
with oxidative allylation (hydrogen departure) of the substrate.
This behavior parallels the Lewis acid catalyzed allylation of allyl
ethers.[20] Such a reactivity change may be due to the fact that
substrate 2o can give rise to cation A (via the classical hydride
abstraction), (Scheme 7, red path), as well as to the 1,3-
diphenylallyl cation B (via OÀ C cleavage), (Scheme 7, blue path).
Scheme 6. Reductive allylation of (E)-1,3-diphenyl-3-methoxypropene.
Eur. J. Org. Chem. 2021, 2162–2168
www.eurjoc.org
2165
© 2021 Wiley-VCH GmbH

Full Papers
doi.org/10.1002/ejoc.202100026
1
2
3
4
5
6
7
8
Scheme 7. Mechanistic proposal for the isohypsic allylation of substrate 2o.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Scheme 9. Scope of the reaction with benzylic methyl ethers. [a] typical
reaction conditions: 7a–7h (0.5 mmol), 3a (2.0 equiv.), 1a (1.2 equiv.),
°
Solvent (2.5 mL), 45 C in a sealed tube, 6–24 h. [b] Yields: spectroscopic
yields, in parentheses, were determined by quantitative 1H NMR using
benzoic acid as an internal standard; isolated yields are out of parentheses.
[c] Conversion was only 13%. [d] p-homoallylanisole 9 was also isolated in
5% yield.
Scheme 8. Scope of the reaction with alkyl- and alkenyl-substituted allyl
methylethers. [a] typical reaction conditions: 2p–2s (0.5 mmol), 3a
°
(2.0 equiv.), 1a (1.2 equiv.), Solvent (2.5 mL), 45 C in a sealed tube, 45 min-
2 h.
The allylation of benzylic methyl ethers was then tested
(Scheme 9).[21] When benzyl methylether 7a was treated with
allyl silane under the same reaction conditions as previously
used for allylic substrates, the reaction turned out to be rather
sluggish, and an overnight period was necessary to obtain full
conversion. Furthermore, the resulting allylated product 8aa
was very volatile, and only a spectroscopic (1H-NMR) yield of
47% could be determined.
The methyl ether bearing a p-trifluoromethyl group on the
aryl ring 7b gave a very low conversion (15%) after 24 h
reaction, and the allylated compound 8ba was isolated with a
very low yield (8%). This result may be in line with the lower
reactivity of the electron-poor substrate, which is less prone to
hydride abstraction. Allylation of 2-(methoxymethyl)naphtalene
7c and piperonyl methyl ether 7d were more successful,
affording the corresponding allylated products 8ca and 8da in
60% and 53% isolated yields. In the case of 1-methoxy-4-
(methoxymethyl)benzene 7e, the desired product 8ea was
Scheme 10. Scope of the reaction as a function of the substitution of the
allylsilane. [a] 2a: R=(E)-Ph-CH=CH-; 7d: R=piperonyl; [b] 3b: R1 =H,
R2 =CH2OAc; 3c: R1 =Ph, R2 =H. [c] Typical reaction conditions: 2a or 7d
°
(0.5 mmol), 3b–c (2.0 equiv.), 1a (1.2 equiv.), solvent (2.5 mL), 45 C in a
sealed tube, 1–3 h30 for substrate 2a or 24 h for substrate 7d. Isolated
yields.
1
isolated in 22% yield (43% H-NMR), accompanied by a little
bis-MOM-protected catechol ether 7h coming from demeth-
ylvanillin could be converted in only 6 h into the allylated
methylether 8ha in 64% isolated yield.
The scope of the reaction as a function of the substitution
of the allylsilane was then undertaken, starting with an
allylsilane substituted at position 2 (Scheme 10). Accordingly,
reaction between cinnamyl methyl ether 2a and 2-((trimeth-
amount (5%) of the isohypsic allylation product, homoallyl
anisole 9, whose formation may be explained as in the case of
the conversion of 2o into 6 (vide supra). Switching to benzyl
and methoxymethyl (MOM) protection of the para located
phenol function gave better yields [(57%), 32%], [(69%), 57%]
of the respective allylated products 8fa and 8ga. Finally, the
Eur. J. Org. Chem. 2021, 2162–2168
www.eurjoc.org
2166
© 2021 Wiley-VCH GmbH

Full Papers
doi.org/10.1002/ejoc.202100026
versions as well as to the application of our new procedure to
the total synthesis of relevant targets.
1
2
3
4
5
Experimental Section
Water-free preparation of the oxoammonium salt 1a: In a round
bottom flask under inert atmosphere was dissolved TEMPO (2 g,
12.8 mmol, 1.0 equiv.) in 15 mL (C=1.3 M) of dry chloroform. After
6
7
8
9
°
cooling down the mixture to 0 C (ice/water bath), tetrafluoroboric
acid diethylether complex (1.9 mL, 14.1 mmol, 1.1 equiv.) was
added dropwise with a syringe. The mixture is warmed up to room
temperature and stirred at this temperature for 1 h. Diacetoxyiodo-
benzene (4.12 g, 12.8 mmol, 1.0 equiv.) was then added in one
portion and the mixture was stirred at room temperature for 2 h.
The obtained yellow suspension was then vacuum-filtered with a
Büchner or Hirsch funnel and the solid was washed with 2x50 mL
of chloroform. The yellow solid was collected in a flask and dried
under high vacuum at room temperature for 1 h to give the pure
oxoammonium salt 1a (2.7 g, 87% yield). 1H NMR (CD3CN,
300 MHz) δ 2.53–2.49 (m, 4H), 2.44-2.36 (m, 6H), 1.67 (s, 12H); 13C
NMR (CD3CN, 75 MHz) 105.2 (2 C), 41.5 (2 C), 29.6 (4 C), 17.1; 19F
NMR (CD3CN, 282.3 MHz) δ À 151.6 (s). The characterization data are
in accordance with the ones reported in the literature.[27]
Scheme 11. Deviations from the allylsilane-ether coupling.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
ylsilyl)methyl)-allyl acetate 3b gave the corresponding adduct
4ab in 61% yield. Reaction of piperonyl methyl ether 7d with
3b gave the allylated adduct 8db in 55% yield. Reaction of 2a
with cinnamyltrimethylsilane 3c[22] gave 4ac in 69% yield as a
69:31 mixture of diastereoisomers. Analogously, reaction of 7d
with 3c afforded the corresponding adduct 8dc in 58% and as
74:26 syn/anti diastereoisomeric ratio.[23]
Not unexpectedly, the syn preference found in the CÀ H
allylation of the benzylic ether 7d parallels the syn diastereose-
lectivity observed in the acetal Hosomi-Sakurai reaction.[24]
On the other hand, treating 2a with methallyltrimeth-
ylsilane 3d under the above reaction conditions gave only
methallylTEMPO in 70% yield (with respect to the oxoammo-
nium salt) 1a, while treating 7d with 3d gave the demeth-
oxylated (isohypsic) adduct in 64% yield (with respect to 7d) as
well as methallylTEMPO in 49% yield (with respect to 1a)
(Scheme 11).[25] These results suggest that this allylsilane-ether
CÀ H coupling is rather sensitive to the nature of both the
reaction partners. Indeed, if the allylic silane is much more
nucleophilic than simple allylsilane, as is the case of
metallylsilane,[26] it can compete with the ether substrate in the
General procedure for the allylation of allyl- and benzyl meth-
ylethers: A sealed tube equipped with a magnetic stirrer was
charged with the allylether or the benzylic ether substrate
(1.0 equiv.) and the tube was placed under high vacuum and then
backfilled with argon. The vacuum/argon cycles were repeated
twice and anhydrous dichloromethane (0.2 M) was introduced. To
this solution was added the oxoammonium salt (1.2 equiv.)
followed by the allylsilane (2.0 equiv.). The reaction mixture was
°
stirred at 45 C until conversion was complete as checked by TLC
(usually around 1 h when starting from allylethers and 24 h when
starting from benzylic ethers). The mixture was then cooled to rt
and mixture was filtered over a pad of silica gel with dichloro-
methane. The filtrate was concentrated in vacuo and the residue
was purified by flash column chromatography on silica gel to give
the desired products.
⊕
⊖
interaction with T BF4 , which will behave as an electrophile.
On the other hand, the hydride abstraction next to the ethereal
oxygen atom may be derailed from the allylic/benzylic position
toward the alternative α‘ methyl position.
Acknowledgements
Conclusion
The authors would like to acknowledge CNRS and Sorbonne
Université for financial support. G.B. and F.C. also thank the
Erasmus+ programme for a scholarship. G.B. and G.P. thank
COST Action, CA15106 (CHAOS) for funding. The authors thank
Aurélie Bernard and Claire Troufflard for their help in quantitative
NMR analyses as well as Dalila Seghouane, Cédric Przybylski and
Gilles Clodic for HRMS analyses. Support through the Marie
Skłodowska-Curie Innovative Training Network European Training
Networks (ITN-ETN) CÀ H Activation for Industrial Renewal (CHAIR),
grant agreement no: 860762, is also gratefully acknowledged.
Finally, G.P. warmly thanks Prof. Pier Giorgio Cozzi for helpful
discussions.
In summary, we developed a CÀ H allylsilane-ether coupling
reaction starting from allylic and benzylic methylethers in the
presence of oxoammonium salts. The use of the more soluble
⊕
⊖
T BF4 TEMPO-based oxoammonium salt allowed the genera-
tion of functionalized homoallyl methylethers in moderate to
good yields. The use of cinnamyltrimethylsilane also gave a
relatively good diastereoselectivity in favor of the syn isomers,
which is in accordance with the results reported on the Lewis-
acid mediated allylation of acetals with allyltrimethylsilanes. The
nature of the O-substitution on the other side of the reactive
allylic moiety is crucial. Indeed, while the methyl group appears
ideal to assure enough stability to the transient oxocarbenium
ion, a t-butyl group opens the way to a rapid non-hydrolytic Conflict of Interest
aldehyde formation via CÀ O cleavage. Future work will be
dedicated to the development of catalytic and enantioselective
The authors declare no conflict of interest.
Eur. J. Org. Chem. 2021, 2162–2168
www.eurjoc.org
2167
© 2021 Wiley-VCH GmbH

Full Papers
doi.org/10.1002/ejoc.202100026
[13] In the beginning of our investigations, the oxoammonium salts were
prepared according a procedure reported in the literature, see ref 8a.
For other articles reporting on the preparation of oxoammonium salt
1a, see: a) M. Shibuya, M. Tomizawa, Y. Iwabushi, J. Org. Chem. 2008, 73,
4750–4752; b) T. Vogler, A. Studer, Adv. Synth. Catal. 2008, 350, 1963–
1967; c) H.-R. Huo, X.-Y. Tang, Y.-F. Gong, Synthesis 2018, 50, 2727–2740.
See also:; d) M. A. Mercadante, C. B. Kelly, J. M. Bobbitt, L. J. Tilley, N. E.
Leadbeater, Nat. Protoc. 2013, 8, 666–676; e) J. M. Bobbitt, J. Org. Chem.
1998, 63, 9367–9374; f) L. J. Tilley, J. M. Bobbit, S. A. Murray, C. E. Camire,
N. A. Eddy, Synthesis 2013, 45, 326–329; g) S. A. Miller, J. Nandi, N. E.
Leadbeater, N. A. Eddy, Eur. J. Org. Chem. 2020, 108–112.
[14] a) T. A. Hamlin, C. B. Kelly, J. M. Ovian, R. J. Wiles, L. J. Tilley, N. E.
Leadbeater, J. Org. Chem. 2015, 80, 8150–8167; b) C. A. Morales-Rivera,
P. E. Floreancig, P. Liu, J. Am. Chem. Soc. 2017, 139, 17935–17944; see
also:; c) J. E. Nutting, M. Rafiee, S. S. Stahl, Chem. Rev. 2018, 118, 4834–
4885; d) M. V. H. Huynh, T. J. Meyer, Chem. Rev. 2007, 107, 5004–5064.
[15] Oxoammonium 1e was prepared according to the procedure using
silicon tetrachloride reported in: S. Stefan, F. Belaj, T. Madl, R. Pietschnig,
Eur. J. Inorg. Chem. 2010, 289–297.
Keywords: Allylation
· Benzyl ethers · Cinnamyl ethers ·
1
2
3
4
5
6
7
8
9
C(sp3)À H functionalization · Oxoammonium
[1] For a selection of recent reviews, see: a) G. Pototschnig, N. Maulide, M.
Schnürch, Chem. Eur. J. 2017, 23, 9206–9232; b) F. Roudesly, J. Oble, G.
Poli, J. Mol. Catal. A 2017, 426, 275–296; c) J. Le Bras, J. Muzart, Eur. J.
Org. Chem. 2018, 1176–1203; d) R. R. Karimov, J. F. Hartwig, Angew.
Chem. Int. Ed. 2018, 57, 4234–4241; Angew. Chem. 2018, 130, 4309–
4317; e) S. St John-Campbell, J. A. Bull, Adv. Synth. Catal. 2019, 361,
3662–3682; f) P. Gandeepan, T. Müller, D. Zell, G. Cera, S. Warratz, L.
Ackermann, Chem. Rev. 2019, 119, 2192–2452.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
[2] For a recent review, see: N. Dastbaravardeh, M. Christakakou, M. Haider,
M. Schnürch, Synthesis 2014, 46, 1421–1439.
[3] For a selection of recent reviews, see: a) Y. Qin, L. Zhu, S. Luo, Chem.
Rev. 2017, 117, 9433–9520; b) B. V. Varun, J. Dhineshkumar, K. R.
Bettadapur, Y. Siddaraju, K. Alagiri, K. R. Prabhu, Tetrahedron Lett. 2017,
58, 803–824; c) S.-R. Guo, P. S. Kumar, M. Yang, Adv. Synth. Catal. 2017,
358, 2–25; d) A. E. Wendlandt, S. S. Stahl, Angew. Chem. Int. Ed. 2015, 54,
14638–14658; Angew. Chem. 2015, 127, 14848–14868; e) A. Gini, T.
Brandhofer, O. García Mancheño, Org. Biomol. Chem. 2017, 15, 1294–
1312.
[4] For a selection of recent papers, see: a) H. Kim, D. Lee, Synlett 2015, 26,
2583–2587; b) A. Lee, R. C. Betori, E. A. Crane, K. A. Scheidt, J. Am. Chem.
Soc. 2018, 140, 6212–6216; c) Y. Mao, M. Cao, X. Pan, J. Huang, J. Li, L.
Xu, L. Liu, Org. Chem. Front. 2019, 6, 2028–2031; d) Z. Džambaski, B. P.
Bondžić, Org. Biomol. Chem. 2019, 17, 6420–6425; e) R. Xiong, M. I.
Hussain, Q. Liu, W. Xia, Y. Xiong, Tetrahedron 2020, 76, 130798; f) E. Lee-
Ruff, F. J. Ablenas, Can. J. Chem. 1989, 699; g) For a recent example
using 2,5-dimethylbenzoquinone as the oxidant, see: T.-C. Wang, L.-F.
Fan, Y. Shen, P.-S. Wang, L.-Z. Gong, J. Am. Chem. Soc. 2019, 141, 10616–
10620.
[16] For a DDQ-mediated allylation of benzylamines with allyltributylstan-
nane, see: R. Xiong, M. I. Hussain, Q. Liu, W. Xia, Y. Xiong, Tetrahedron
2020, 76, 130798.
1
⊖
[17] H NMR analysis (dry CD3CN) of T⊕BF4 obtained by following the
published protocol (Ref. 8a) revealed the presence of small amounts of
TEMPOH.HBF4.
[18] PhI(OAc)2 has been used as oxidant for TEMPOH.HX in the Epp-
Widlanski oxidation of alcohols a) A. De Mico, R. Margarita, L. Parlanti, A.
Vescovi, G. Piancatelli, J. Org. Chem. 1997, 62, 6974–6977; b) J. B. Epp,
T. S. Widlanski, J. Org. Chem. 1999, 64, 293–295.
[19] A study dedicated to the anhydrous conversion of ethers into
aldehydes will be reported elsewhere.
[20] a) A. Hosomi, T. Imai, M. Endo, H. Sakurai, J. Organomet. Chem. 1985,
285, 95–107; b) A. Hosomi, M. Endo, H. Sakurai, Chem. Lett. 1978, 499–
502.
[5] Toxicology data of DDQ can be found on the MSDS sheets, see also:
a) L. Liu, P. E. Floreancig, Org. Lett. 2010, 12; b) 4686-4689. Toxicity
concerns essentially come from the potential release of HCN when DDQ
is in contact with water.
[21] For an example of oxoammonium mediated dehydrogenative alkylation
of benzyl ethers with active methylenes, see ref 8a.
[22] Cinnamyltrimethylsilane was prepared from cinnamyl alcohol according
to the protocol of Szabó et al.: N. Selander, J. R. Paasch, K. J. Szabó, J.
Am. Chem. Soc. 2011, 133, 409–411.
[23] The assignment of the syn and anti diastereoisomers was done by
comparison with the reported data in the literature: J. C. T. Reddel, K. E.
Lutz, A. B. Diagne, R. J. Thomson, Angew. Chem. Int. Ed. 2014, 53, 1395–
1398; Angew. Chem. 2014, 126, 1419–1422.
[24] A. Hosomi, M. Ando, H. Sakurai, Chem. Lett. 1986, 365–368.
[25] Also, 2a and 7d did not react when put in the presence of geranyl
trimethylsilane 3e under the above reaction conditions, while treat-
ment of 7d with 3f gave only an intractable mixture of products. See
the Supporting Information for the synthesis of these two allylsilanes.
[26] An additional methyl group at the developing carbenium center in an
allyl silane causes a striking reactivity increase: H. Mayr, G. Hagen, J.
Chem. Soc. Chem. Commun. 1989, 91–92.
[6] For reviews, see: a) O. García Mancheño, T. Stopka, Synthesis 2013, 45,
1602–1611; b) N. E. Leadbeater, J. M. Bobbitt, Aldrichimica Acta 2014, 47,
65–74.
[7] For references on TEMPO+ mediated amine α-C(sp3)À H couplings, see:
a) J. Wang, S. Yang, Tetrahedron Lett. 2016, 57, 3444–3448; b) C. Yan, Y.
Liu, Q. Wang, Org. Lett. 2015, 17, 5714–5717; c) C. Yan, Y. Liu, Q. Wang,
RSC Adv. 2014, 4, 60075–60078; d) S. Sun, C. Li, P. E. Floreancig, H. Lou,
L. Liu, Org. Lett. 2015, 17, 1684–1687.
[8] a) H. Richter, O. García Mancheño, Eur. J. Org. Chem. 2010, 4460–4467;
b) Z. Zhang, Y.-Q. Tu, X.-M. Zhang, F.-M. Zhang, S.-H. Wang, Org. Chem.
Front. 2019, 6, 2275–2279.
[9] Y. Hayashi, T. Mukaiyama, Chem. Lett. 1987, 1811–1814.
[10] a) A. Hosomi, H. Sakurai, Tetrahedron Lett. 1976, 1295–1298; b) A.
Hosomi, M. Endo, H. Sakurai, Chem. Lett. 1976, 941–942; c) A. Hosomi, H.
Sakurai, J. Am. Chem. Soc. 1977, 99, 1673–1675; d) A. Hosomi, Acc.
Chem. Res. 1988, 21, 200–206.
[11] For the allylation of unsaturated chromanes and isochromanes, see:
a) D. J. Clausen, P. E. Floreancig, J. Org. Chem. 2012, 77, 6574–6582; b) Y.
Cui, L. A. Villafane, D. J. Clausen, P. E. Floreancig, Tetrahedron 2013, 69,
7618–7626. For an intramolecular variant, see: c) A. K. Ghosh, X. Cheng,
Tetrahedron Lett. 2012, 53, 2568–2570.
[12] a) P. P. Pradhan, J. M. Bobbitt, W. F. Bailey, J. Org. Chem. 2009, 74, 9524–
9527; b) C. B. Kelly, J. M. Ovian, R. M. Cywar, T. R. Gosselin, R. J. Wiles,
N. E. Leadbeater, Org. Biomol. Chem. 2015, 13, 4255–4259.
[27] M. Shibuya, M. Tomizawa, Y. Iwabushi, J. Org. Chem. 2008, 73, 4750–
4752.
Manuscript received: January 11, 2021
Revised manuscript received: February 20, 2021
Accepted manuscript online: February 23, 2021
Eur. J. Org. Chem. 2021, 2162–2168
www.eurjoc.org
2168
© 2021 Wiley-VCH GmbH