Metal-mediated cyclization of aryl and benzyl glycidyl ethers: A complete scenario

Article abstract of DOI:10.1021/ja8062887

Enantiomerically pure aryl and benzyl glycidyl ethers undergo stereospecific cyclizations leading to 3-chromanols or to tetrahydrobenzo[c]oxepin-4-ols under mild conditions in the presence of a catalytic amount of FeBr₃/3AgOTf. The same process

Full text of DOI:10.1021/ja8062887

Published on Web 11/20/2008
Metal-Mediated Cyclization of Aryl and Benzyl Glycidyl Ethers: A Complete
Scenario
Roc´ıo Marcos,† Carles Rodr´ıguez-Escrich,† Clara I. Herrer´ıas,†,‡ and Miquel A. Perica`s†,§,
*
Institute of Chemical Research of Catalonia (ICIQ), AV. Pa¨ısos Catalans 16, 43007 Tarragona, Spain, and
Departament de Qu´ımica Orga`nica, UniVersitat de Barcelona (UB), 08028 Barcelona, Spain
Received August 8, 2008; E-mail: mapericas@iciq.es
Table 1. Lewis Acid Mediated Cyclization of Aryl Glycidyl Ethers 1
Glycidyl systems, readily available in enantiomerically pure form
by Sharpless epoxidation,1 are versatile scaffolds that can be
submitted to many different selective transformations.2 Intramo-
lecular processes involving their epoxide ring are particularly
interesting, since they can provide access to a variety of substituted
heterocyclic systems both in racemic and in enantiomerically pure
form.3 In this context, it has been reported that aryl glycidyl ethers
1, when treated with a catalytic amount of AuCl3 in the presence
of AgOTf in 1,2-dichloroethane at high temperatures for prolonged
periods of time, lead to 3-chromanols 2 as the only reaction
products.4 On the other hand, we have shown that enantiomerically
pure benzyl glycidyl ethers 3, when treated with BF3 ·Et2O in
dichloromethane at low temperature for short periods of time lead
to either 4,5-disubstituted tetrahydrobenzo[c]oxepin-4-ols 4 or to
4-diarylmethyl-1,3-dioxolanes 5 through stereospecific processes,
depending on the substitution pattern of the benzyl residue.5
In spite of the close similarity of these reactions, it has been reported4
that a Lewis acid such as BF3 ·Et2O fails to induce the conversion of
1 into 2 and that the Au(III) catalyst was absolutely required for the
reaction to take place. According to this, the intermediacy of an
arylgold(III) species formed through an auration step6 has been
proposed as a mechanistic alternative for this transformation. We were
intrigued by the possibility of this mechanistic duality, and given the
interest of the considered reactions and of the products arising
therefrom, we decided to reinvestigate this set of processes from the
perspective of its mechanistic nature.
to 3-Chromanols
entry
product
R1
R2
R3
R4
conda
yield [%]
1
2
3
4
5
6
7
8
2a
2a
2b
2c
2d
2d
2e
2f
Ph
OMe
H
OMe
A
B
79
92
61
92
88
97
76
73
95
99
94d
91e
18
17
Ph
Ph
Ph
H
H
Me
H
OMe
H
H
H
Me
Bb
Bc
B
C
p-BrPh
Ph
Ph
OMe
H
H
H
I
t-Bu
OMe
H
H
B
Cb
B
9
2g
2g
2h
2h
2i
10
11
12
13
14
C
D
E
D
E
H
H
H
H
H
H
t-Bu
H
H
H
2i
a Cond. A: BF3 ·Et2O (30 mol%), CH2Cl2, -78 or -55 °C, 30 min.
Cond. B: FeBr3 (10 mol%), CH2Cl2, 20 °C, 30 min. Cond. C: FeBr3/
3AgOTf (10 mol%), CH2Cl2, 20 °C, 30 h. Cond. D: FeBr3/3AgOTf (10
mol%), ClCH2CH2Cl, 80 °C, 48 h. Cond. E: FeBr3/3AgOTf (10 mol%),
ClCH2CH2Cl, 140 °C, microwaves, 40 min. b 30 mol% catalyst, 60 min.
c Reaction time was 60 min. d 57% conversion. e 53% conversion.
The unsubstituted glycidyl ethers 1 h-i, lacking the activation
of the epoxide ring toward ring-opening provided by the aryl
substituent, are sensitive probes for the efficiency of the different
Lewis acid types in the cyclization process. Taking 1i as a model
substrate, different Lewis acids known to induce the ring opening
of epoxides [BF3 ·Et2O, FeBr3, InCl3, LiClO4, Cu(OTf)2, Cu(ClO4)2,
Zn(OTf)2] were tested under different experimental conditions.8 In
no case the expected 3-chromanol 2i was formed, the only identified
minor reaction products being halohydrins 6i.9 In view of the results
reported by He,4 we decided to explore the use in the reaction of
FeBr3 in combination with AgOTf with the aim of increasing the
electrophilicity of the iron salt. Gratifyingly enough, 2i was
obtained, albeit in a modest 18% yield, under these conditions
(Conditions D, entry 13). It is worth mentioning that, in our hands,
use of the AuCl3/3AgOTf catalyst (5 mol%) under identical
experimental conditions afforded 2i in 11% yield.10 Use of
Conditions D allowed preparation of 2h in 94% yield (57%
conversion, entry 11). Interestingly, reaction times for the prepara-
tion of 2h-i could be drastically reduced (from 48 h to 40 min)
without yield decrease by performing the reactions at 140 °C with
microwave irradiation (Conditions E, entries 12 and 14). The FeBr3/
3AgOTf reagent was subsequently tested under milder conditions
(Conditions C) for substrates 1a-g. Except for the results in entries
6 and 10, only marginal yield increases were recorded.
As the starting point of this research, we examined the Lewis
acid reactivity of a family of aryl glycidyl ethers 1a-i, most of
them in enantiomerically pure form (1h and 1i were racemic). The
results of this study have been summarized in Table 1.
In contrast with the results discussed above,4 but in full
agreement with our previous findings,5 BF3 ·Et2O induces the high
yield (79%), stereospecific cyclization of 1a into 3-chromanol 2a
after a short treatment at -55 °C in dichloromethane (entry 1).
Even more conveniently, the same reaction can be induced by FeBr3
(10 mol%)7 at 20 °C (Conditions B), 2a being obtained in 92%
yield after 30 min (entry 2). It is interesting to realize that for this
particular substrate the use of AuCl3/3AgOTf (2.5 mol%) in 1,2-
dichloroethane, although involving harsher reaction conditions
(50 °C) and a more prolonged reaction time (4 h), leads to lower
yields than the more conventional Lewis acids.4
The use of FeBr3 under the same very mild conditions employed
for 1a allowed the stereospecific, high yield cyclization of aryl
phenylglycidyl ethers 1b-g to 3-chromanols 2b-g. For substrates
bearing less nucleophilic aryloxy fragments, like 1b, increased
amounts of FeBr3 (30 mol%) and extended reaction times (1 h) are
required for complete conversion (entry 3).
† Institute of Chemical Research of Catalonia (ICIQ).
‡ On leave from: Departamento de Qu´ımica Orga´nica, Universidad de Zaragoza,
50009 Zaragoza, Spain
The results obtained with unactivated substrates (1h-i) provide
a clear indication that the cationic Au(III) or Fe(III) species likely
§ Universitat de Barcelona (UB).
9
16838 J. AM. CHEM. SOC. 2008, 130, 16838–16839
10.1021/ja8062887 CCC: $40.75
2008 American Chemical Society

C O M M U N I C A T I O N S
generated in the presence of AgOTf can efficiently mediate the
formation of 2 when halide delivery is avoided (Figure 1).
substrates to diarylmethanes is a more critical process, normally
accompanied by variable degrees of halohydrin formation. Accord-
ing to this, the rearrangement of 3c-d leading to 5c-d is a more
stringent test for possible mediators. With substrate 3c (entries
8-11), it becomes clear that the ability of the four Lewis acids to
induce the rearrangement varies in the order BF3 ·Et2O > FeBr3/
3AgOTf > AuCl3/3AgOTf > FeBr3. When the yield of the
diarylmethane 5 is considered, the FeBr3/3AgOTf reagent represents
an optimal compromise (entry 11). The same trends are observed
for the conversion of 3d into 5d.
In summary, the results reported here clearly show that the
cyclization and rearrangement reactions of aryl and benzyl glycidyl
ethers are Lewis acid mediated processes. While the FeBr3/3AgOTf
combination appears as the catalyst of choice for this set of
processes, the cationic species presumably formed through the
interaction of AuCl3 with AgOTf is also able to mediate, albeit in
a less efficient manner, the same reactions. From a mechanistic
point of view, the observed reactivity trends clearly indicate that
these cyclizations are of the Friedel-Crafts type.
From the perspective of practical use, cost and availability
considerations11 make iron bromide a most attractive alternative
for these reactions and, in general, for processes where gold and
other expensive metals merely act as Lewis acids.
Figure 1. Reaction pathways of aryl glycidyl ethers with Lewis acids.
Once the ability of Lewis acids to induce the cyclization of aryl
glycidyl ethers 1 into 3-chromanols 2 had been established, we
turned our attention toward benzyl glycidyl ethers 3, with the goal
of determining whether the FeBr3, FeBr3/3AgOTf, and AuCl3/
3AgOTf catalysts could also induce the reactions5 shown in Table
2. Substrates 3a-b were selected as precursors of tetrahydroben-
Table 2. Lewis Acid Mediated Cyclization/Rearrangement of
Benzyl Glycidyl Ethers 3 to Tetrahydrobenzo[c]oxepin-4-ols 4 or to
4-Diarylmethyl-1,3-dioxolanes 5
Acknowledgment. This paper is dedicated to Professor Josep
Font on the occasion of his 70th birthday. This work was funded
by MEC (Grant CTQ2005-02193/BQU), DIUE (Grant 2005SGR225),
Consolider Ingenio 2010 (Grant CSD2006-0003), and ICIQ Foun-
dation. R.M. is also indebted to MEC for a fellowship.
Supporting Information Available: Experimental details and
product characterizations. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
entry
substrate
product
conditionsa
yield [%]
(1) (a) Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.; Masamune, H.;
Sharpless, K. B. J. Am. Chem. Soc. 1987, 109, 5765–5780. For a review
see: (b) Katsuki, T.; Martin, V. S. Org. React. 1996, 48, 1–299.
(2) Aziridines and Epoxides in Organic Synthesis; Yudin, A. K., Ed.; Wiley-
VCH: Weinheim, Germany, 2006.
(3) Recent work: (a) Bandini, M.; Melloni, A.; Umani-Ronchi, A. Angew.
Chem., Int. Ed. 2004, 43, 550–556. (b) Bandini, M.; Melloni, A.; Tommasi,
S.; Umani-Ronchi, A. Synlett 2005, 1199–1222. (c) Smith, B. M.; Skellam,
E. J.; Oxley, S. J.; Gram, A. E. Org. Biomol. Chem. 2007, 5, 1979–1982.
(d) Pulipaka, A. B.; Bergmeier, S. C. Synthesis 2008, 1420–1430. (e) Liu,
Y.; Li, X.; Lin, G.; Xiang, Z.; Xiang, J.; Zhao, M.; Chen, J.; Yang, Z. J.
Org. Chem. 2008, 73, 4625–4629.
1
2
3
4
5
6
7
8
3a
3a
3a
3a
3b
3b
3b
3c
3c
3c
3c
3d
3d
4a
4a
4a
4a
4b
4b
4b
5c
5c
5c
5c
5d
5d
A
B
C
D
B
C
D
Ab
B
C
D
C
D
80
82
70
88
75
74
81
59
25c
72
82
89
96
9
10
11
12
13
(4) Shi, Z.; He, C. J. Am. Chem. Soc. 2004, 126, 5964–5965.
(5) Islas-Gonza´lez, G.; Benet-Buchholz, J.; Maestro, M. A.; Riera, A.; Perica`s,
M. A. J. Org. Chem. 2006, 71, 1537–1544.
(6) Kharasch, M. S.; Isbell, H. S. J. Am. Chem. Soc. 1931, 53, 3053–3059.
(7) For the use of FeBr3 in aza-Prins cyclizations, see: Carballo, R. M.; Ram´ırez,
M. A.; Rodr´ıguez, M. L.; Mart´ın, V. S.; Padro´n, J. I. Org. Lett. 2006, 8,
3837–3840.
(8) InBr3: (a) Bandini, M.; Cozzi, P. G.; Melchiorre, P.; Umani-Ronchi, A. J.
Org. Chem. 2002, 67, 5386–5389. SnCl4: (b) Taylor, S. K.; Hockermann,
G. L.; Karrick, G. L.; Lyle, S. B.; Schramm, S. B. J. Org. Chem. 1983, 48,
2449. (c) Taylor, S. K.; Davisson, M. E.; Hissom, B. R., Jr.; Brown, S. L.;
Pristach, H. A.; Schramm, S. B.; Harvey, S. M. J. Org. Chem. 1987, 52,
425–429. Metal perchlorates: (d) Chini, M.; Crotti, P.; Flippin, L. A.;
Macchia, F. J. Org. Chem. 1990, 55, 4525–4562. For reviews, see: (e)
Lewis Acids in Organic Synthesis; Yamamoto, H., Ed.; Wiley-VCH:
Weinheim, Germany, 2006.
a Cond. A: BF3 ·Et2O (30 mol%), CH2Cl2, -78 °C, 15 min. Cond. B:
FeBr3 (10 mol%), CH2Cl2, 20 °C, 60 min. Cond. C: AuCl3/3AgOTf (2.5
mol%), ClCH2CH2Cl, 50 °C, 4 h. Cond. D: FeBr3/3AgOTf (20 mol%),
CH2Cl2, 20 °C, 30 min. b Reaction at -35 °C for 5 min. c 29% of the
corresponding bromohydrin was also obtained. Reaction time was 90 min.
zo[c]oxepin-4-ols 4a-b via ortho activation of the benzyl sub-
stituent, while 3c-d were chosen as precursors to 4-diarylmethyl-
1,3-dioxolanes 5 via ipso activation of the benzyl fragment. The
results of the study have been summarized in Table 2 along with
those obtained with BF3 ·Et2O, when available.5
(9) For the formation of fluorohydrins from epoxyethers with BF3 · Et2O, see:
Islas-Gonza´lez, G.; Puigjaner, C.; Vidal-Ferran, A.; Moyano, A.; Riera,
A.; Perica`s, M. A. Tetrahedron Lett. 2004, 45, 6337–6341.
When the stereospecific cyclization of 3a is considered (entries
1-4), it becomes clear that Au(III) is also able to induce conversion
into 4a, although FeBr3/3AgOTf and the other tested Lewis acids
are more efficient mediators than AuCl3/3AgOTf for this process.
The less reactive 3b displayed similar behavior.
While the cyclization of substrates 3 bearing an ortho-activated
benzyl substituent to oxepinols 4 is normally very clean,5 the
corresponding Lewis acid induced rearrangement of ipso-activated
(10) The FeBr3/3AgOTf and AuCl3/3AgOTf catalytic systems have also been
tested in the stereospecific and regioselective intermolecular addition of
1,3,5-trimethoxybenzene to (2S,3S)-3-phenylglycidyl methyl ether. The
former gave the desired product 7 in 60% yield, whereas the latter was
less active (ca. 40%; see SI for details). Further studies concerning related
intermolecular addition are in course.
(11) For recent reviews, see: (a) Enthaler, S.; Junge, K.; Beller, M. Angew.
Chem., Int. Ed. 2008, 47, 3317–3321. (b) Correa, A.; Garc´ıa-Mancheno,
O.; Bolm, C. Chem. Soc. ReV. 2008, 37, 1108–1117.
JA8062887
9
J. AM. CHEM. SOC. VOL. 130, NO. 50, 2008 16839