Boronic acid catalysis as a mild and versatile strategy for direct carbo- and heterocyclizations of free allylic alcohols

Article abstract of DOI:10.1002/anie.201201620

BAC to the future: Boronic acid catalysis (BAC) was applied to the direct activation of alcohols leading to the preparation of carbocycles (see scheme), benzofurans, tetrahydrofurans, pyrrolidines, pyrans, piperidines, and various polycyclic compounds. The reactions proceed under mild conditions that circumvent the use of reactive leaving groups like halides. Copyright

Full text of DOI:10.1002/anie.201201620

Angewandte
Chemie
DOI: 10.1002/anie.201201620
Homogeneous Catalysis
Boronic Acid Catalysis as a Mild and Versatile Strategy for Direct
Carbo- and Heterocyclizations of Free Allylic Alcohols**
Hongchao Zheng, Sina Ghanbari, Shinji Nakamura, and Dennis G. Hall*
The ability to activate simple electrophilic functional groups
directly without recourse to intermediary functionalities like
halides and pseudohalides (e.g., sulfonates, oxyphosphonium)
has become an important goal of modern organic chemistry.[1]
Boronic acid catalysis (BAC) is emerging as a mild and
effective strategy for the direct, covalent activation of
alcohols and carboxylic acids.[2] Important reactions such as
direct esterification and amidation,[3] intramolecular anhy-
dride formation,[4] imine hydrolysis,[5] epoxide opening,[6] the
Biginelli reaction,[7] Diels–Alder[8] and dipolar cycloaddi-
tions,[9] aldol condensations,[10] ene reactions,[11] Friedel–
Crafts alkylations,[12] and transpositions of allylic/propargylic
alcohols[13] have all been performed recently under BAC.
Faster reactions, milder reaction conditions, and increased
selectivity are some of the benefits provided by BAC. These
attributes are highlighted in recent reports from McCubbin
and co-workers on mild Friedel–Crafts-type alkylations by
activation of allylic and benzylic alcohols with electron-poor
Scheme 1. Boronic acid catalysis of 1,3-transposition of allylic alcohols
boronic acids.[12] Inspired by this approach, we devised mild
(a)[13] and the proposed carbo- and heterocyclizations (b).
and effective reaction conditions for stereoselective 1,3-
transpositions of allylic and propargylic alcohols in the
absence of an external nucleophile (Scheme 1a).[13] The
successful development of these reactions unveiled boronic
acids 1a and 1b as superior catalysts for the activation of
hydroxy groups. We reasoned that under conditions favoring
a strong polarization or full ionization of allylic alcohols,
a suitably placed nucleophilic functionality on the substrate
could lead to the formation of cyclic products under unusually
mild reaction conditions (Scheme 1b). Indeed, the pKa of
boronic acids is in the range of 5–9,[14] which is significantly
higher than that of the strong protic acids usually required in
cationic cyclizations. Herein, we demonstrate the versatility
and mildness (low temperature, functional group tolerance)
of boronic acid catalysis over traditional methods employing
strong Lewis and Brønsted acids.[15]
In the first round of optimization, several solvents were
evaluated in the Friedel–Crafts cyclization of model alcohol
2a catalyzed by 1a (Table 1). Although some nonpolar
solvents like toluene and hexanes were quite effective, the
highly polar aprotic solvent nitromethane was superior
(entry 8). The catalyst was effective even at room temper-
ature (entry 9), however a lower yield was obtained when
using a lower catalyst loading (entries 10–11). A comparison
with pentafluorophenylboronic acid (entry 12) confirmed that
1a (entry 9) is a superior catalyst.[16] Furthermore, a control
run without molecular sieves (entry 13) showed that there
were no advantages to using this dehydrating agent or any
other such as MgSO4 (data not shown). It is noteworthy that
the reaction in entry 9 was repeated on a larger scale (1 g of
2a) to give a 95% yield of 3a with a 96% recovered yield of
the catalyst 1a. Thus, 1a is stable to protodeboronation under
the reaction conditions and is recyclable.
The scope of substrates was explored using the optimal
reaction conditions: 20 mol% catalyst 1a in nitromethane
(Table 2). Most substrates provided cyclic products at room
temperature but reaction times can be reduced significantly
by performing the reactions at 508C. Both benzopyrans 3a
and 3b were obtained in high yield, thus demonstrating that
a single phenyl group in 2b is sufficient to promote effective
activation by the catalyst (entries 1 and 2). Remarkably, the
unactivated arene substrate 2c was successfully cyclized, thus
affording a 72% yield of product 3c with catalyst 1c
(entries 3–5). This result suggests that the design of boronic
acids that are even more electronically impoverished could
[*] H. Zheng, S. Ghanbari, S. Nakamura, Prof. Dr. D. G. Hall
Department of Chemistry, Gunning-Lemieux Chemistry Centre,
University of Alberta
Edmonton, Alberta, T6G 2G2 (Canada)
E-mail: dennis.hall@ualberta.ca
Homepage: http://www.chem.ualberta.ca/~dhall/
[**] This research was generously funded by the Natural Sciences and
Engineering Research Council (NSERC) of Canada, and the
University of Alberta. H.Z. thanks the Alberta Ingenuity Foundation
for a Graduate Scholarship and the University of Alberta for a QEII
Scholarship. S.N. thanks the Japan Society for the Promotion of
Science and the Global COE Program (Nagoya University) for
financial support.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201201620.
Angew. Chem. Int. Ed. 2012, 51, 6187 –6190
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6187

.
Angewandte
Communications
Table 1: Optimization of solvent and reaction conditions in the boronic
acid catalyzed cyclization of model allylic alcohol 2a.[a]
Table 2: Substrate scope in the boronic acid catalyzed intramolecular
Friedel–Crafts cyclizations of allylic alcohols.[a]
Entry
Cat. loading
[mol%)
Solvent[b]
T
[8C]
t
[h]
Yield[c]
[%]
1
2
3
4
5
6
7
8
20
20
20
20
20
20
20
20
20
10
5
DMF
EtOAc
THF
50
50
50
50
50
50
50
50
RT
RT
RT
RT
RT
48
48
48
48
48
48
48
48
60
60
60
60
60
<5
<5
19
23
30
55
67
92
97
Entry
1
Catalyst
T
[8C]
t
[h]
Product[b]
Yield[c]
[%]
1,4-dioxane
CH3CN
toluene
hexanes
CH3NO2
CH3NO2
CH3NO2
CH3NO2
CH3NO2
CH3NO2
1a
RT
RT
50
60
48
48
3a
3b
3c
97
79
50
9
10
11
12
13
67
8
20
20
68[d]
97[e]
2
3
1a
1a
[a] Reaction conditions: The boronic acid catalyst (0.04 mmol) was
added to a solution of alcohol substrate (0.2 mmol) in nitromethane
(1 mL) containing activated 4 A molecular sieves (M.S.; 200 mg). The
solution was stirred at the indicated temperature for the indicated period
of time. [b] Dry solvents were employed. [c] Yields of isolated product
purified by column chromatography on silica gel. In most cases, other
materials are minor amounts of leftover substrate, as well as elimination
and 1,3-transposition products. [d] Using pentafluorophenylboronic acid
(20 mol%) as the catalyst. [e] No molecular sieves were used.
DMF=N,N’-dimethylformamide, THF=tetrahydrofuran.
4
5
1b
1c
50
50
48
48
3c
3c
60
72
6
7
1a
1c
50
50
48
48
3d
21
3d
60
further increase the substrate scope of these Friedel–Crafts-
type cyclizations. For example, the use of the new and more
active catalyst, 2,3-difluoro-4-methylpyridiniumboronic acid
(1c), was necessary for effecting a medium ring closure
affording 3d in a reasonable yield of 60% compared to only
21% with catalyst 1a (entries 6 and 7).
[a] Reaction conditions: The boronic acid catalyst (0.04 mmol) was
added to a solution of alcohol substrate (0.2 mmol) in nitromethane
(1 mL). The solution was stirred at the indicated temperature for the
indicated period of time. [b] E/Z isomer ratio >20:1 for 3b and 3d.
[c] Yields of isolated product purified by column chromatography on
silica gel. In most cases, other materials obtained are minor amounts of
leftover substrate, as well as elimination and 1,3-transposition products.
Boronic acid catalyzed cationic cyclizations are not
À
limited to C C bond formation. Heteroatom cyclizations
with diols and amino alcohols are possible, as shown with the
isolation of tetrahydrofurans 5a,b (Table 3, entries 1–3),
pyrans 5c,d (entries 4-5), oxepan 5e (entry 6), pyrrolidine
5 f (entry 7), and piperidine 5g (entry 8). Secondary alcohols
were employed successfully (entries 9 and 10), with pyran 5i
formed in excellent yield with very high diastereoselectivity.
The desired heterocyclic products were all obtained in high
yields. The mildness and effectiveness of these BAC proce-
dures are remarkable. When attempted under standard protic
conditions (20 mol% p-TsOH under various reaction con-
ditions, see the Supporting Information), the same trans-
formations led to products 3c (Table 2), 5b, 5 f, and 5i
(Table 3) in very low maximum yields of 21, 27, 13, and 24%,
respectively.
Several other variants of the above cationic cyclizations
demonstrate the versatility of boronic acid catalysis in
promoting ring-forming reactions (Scheme 2). The example
of diol 6 shows that benzylic alcohols are suitable substrates
(Scheme 2a). Diol 8 also led to a high yield of cyclized
product 9 with intact phenolic silyl ethers (Scheme 2a), and
thus bears testimony to the functional group compatibility of
these reaction conditions. Indeed, the p-TsOH-catalyzed
reaction gave product in only 61% as the doubly deprotected
bisphenol (see the Supporting Information). The tertiary
nonstabilized alcohol 10 afforded the desired product 11,
albeit in modest yield using a higher temperature (Sche-
me 2b). The benzopyran 13 was formed in high yield by
cyclization of a phenol as the nucleophile (Scheme 2c). The
remarkable cascade polycyclization of substrate 14 provided
the desired tricycle 15 in high diastereoselectivity
(Scheme 2d), while the use of a ketone as nucleophile led
to an efficient spiroketalization reaction (Scheme 2e).
The scope of substrates of these cyclizations of nucleo-
phile-tethered allylic alcohols is consistent with a mechanism
involving complete or near-complete ionization into an allylic
carbocation. For example, while substrate 4b undergoes
a smooth and high-yielding reaction to give tetrahydrofuran
5b at room temperature [Eq. (1), Scheme 3], the isomeric
allylic alcohol 4b’ is inert and reacts slowly only at a temper-
6188
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 6187 –6190

Angewandte
Chemie
Table 3: Substrate scope in the boronic acid catalyzed heterocyclizations
of allylic alcohols.[a]
Entry
1
T
[8C]
t
[h]
Product[b]
Yield[c]
[%]
RT
24
5a
95
2
3
50
50
16
16
5a
99
89
5b
5c
5d
4
5
50
50
16
48
87[d]
94
6
50
16
5e
62[e]
7
8
50
50
16
16
5 f
88
85
5g
88
9
50
50
16
24
5h
5i
(55:45 d.r.)
90
10
(95:5 d.r.)
[a] Reaction conditions: The boronic acid catalyst 1a (4 mg, 0.02 mmol)
was added to a solution of alcohol substrate (0.2 mmol) in nitromethane
(1 mL). The solution was stirred at the indicated temperature for the
indicated period of time. [b] E/Z isomer ratio >20:1 for 5b and 5d.
[c] Yields of isolated product purified by column chromatography on
silica gel. In most cases, other materials obtained are minor amounts of
leftover substrate, as well as elimination and 1,3-transposition products.
[d] 5% of SN2 product (eight-membered ring) was isolated. [e] 30% of
1,3-hydroxy transposition product was isolated. Ts=4-toluenesulfonyl.
Scheme 2. Other examples of boronic acid catalyzed cyclizations of
nonallylic (a, b) and allylic alcohols (c–e).
ature of 808C [Eq. (2)]. Because ionization of 4b is greatly
facilitated by the benzylic nature of this alcohol, this outcome
is consistent with an SN1’ mechanism. Given that 4b’ does not
cyclize readily, cyclization of 4b must be significantly faster
than the 1,3-allylic hydroxy transposition giving isomer 4b’
through an SN2’-like mechanism.[13] Likewise, reverse trans-
position of 4b’ into 4b is obviously not occurring under these
reaction conditions (see the Supporting Information for more
experiments). Altogether these results support the formation
of carbocation-like intermediates, of which formation is
highly favored in the polar aprotic solvent nitromethane.
In summary, we have reported the application of boronic
acid catalysis for the direct carbo- and heterocyclizations of
allylic alcohols. The versatility of this concept was convinc-
ingly demonstrated by the broad range of cyclic and polycyclic
products which can be obtained in high yields with this new
type of catalysis. In addition to avoiding the use of reactive
leaving groups like sulfonates or halides, BAC provides
operationally simple reactions using air-stable catalysts under
Scheme 3. Mechanistic control experiments with alcohols 4b, 4b’.
very mild reaction conditions compared to traditional Lewis
or protic acid catalysis.
Experimental Section
Typical procedure for the boronic acid catalyzed cyclization of allylic
alcohols: 4-(2,2-Diphenylvinyl)chroman (3a; Table 2, entry 1):
2,3,4,5-Tetrafluorophenylboronic acid (1a; 8 mg, 0.04 mmol) was
added to a solution of (E)-5-phenoxy-1,1-diphenylpent-2-en-1-ol (2a;
Angew. Chem. Int. Ed. 2012, 51, 6187 –6190
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6189

.
Angewandte
Communications
66 mg, 0.2 mmol) in CH3NO2 (1 mL) at room temperature. The
resulting solution was stirred at room temperature for 60 h. Upon
evaporation of the solvent under reduced pressure, the residue was
purified by silica gel column chromatography (EtOAc/hexanes =
1:20) to give the chroman 3a (60.5 mg, 97%) in pure form. A larger
scale reaction (1 g of 2a) gave a 95% yield of 3a with a 96%
Maki, K. Ishihara, H. Yamamoto, Tetrahedron 2007, 63, 8645 –
8657; e) R. Al-Zoubi, O. Marion, D. G. Hall, Angew. Chem.
2008, 120, 2918 – 2921; Angew. Chem. Int. Ed. 2008, 47, 2876 –
2879; f) K. Arnold, A. S. Batsanov, B. Davies, A. Whiting, Green
Chem. 2008, 10, 124 – 134; g) M. Tommaso, Angew. Chem. 2010,
122, 6992 – 6995; Angew. Chem. Int. Ed. 2010, 49, 6840 – 6843.
[4] A. Sakakura, T. Ohkubo, R. Yamashita, M. Akakura, K.
Ishihara, Org. Lett. 2011, 13, 892 – 895.
[5] G. Rao, M. Philipp, J. Org. Chem. 1991, 56, 1505 – 1512.
[6] X.-D. Hu, C.-A. Fan, F.-M. Zhang, Y.-Q. Tu, Angew. Chem. 2004,
116, 1734 – 1737; Angew. Chem. Int. Ed. 2004, 43, 1702 – 1705.
[7] A. Debache, B. Boumoud, M. Amimour, A. Belfaitah, S.
Rhouati, B. Carboni, Tetrahedron Lett. 2006, 47, 5697 – 5699.
[8] H. Zheng, D. G. Hall, Tetrahedron Lett. 2010, 51, 3561 – 3564.
[9] H. Zheng, R. McDonald, D. G. Hall, Chem. Eur. J. 2010, 16,
5454 – 5460.
1
recovered yield of the catalyst 1a. H NMR (500 MHz, CDCl3): d =
7.48–7.24 (m, 10H), 7.22–7.10 (m, 2H), 6.93–6.81 (m, 2H), 6.12 (d, J =
10.2 Hz, 1H), 4.34 (dt, J = 10.9, 4.1 Hz, 1H), 4.12–4.02 (m, 1H), 3.78–
3.66 (m, 1H), 2.12–2.00 ppm (m, 2H). 13C NMR (125 MHz, CDCl3):
d = 154.4, 142.7, 142.0, 139.8, 131.6, 129.7, 129.6, 128.6, 128.2, 127.8,
127.37, 127.35, 127.3, 124.9, 120.4, 116.8, 65.1, 35.5, 29.6 ppm. IR
À1
~
(Microscope): n = 3056, 3023, 2947, 2876, 1603, 1580, 1487, 1450 cm
.
HRMS (EI) for C23H20O: calcd. 312.1514; found 312.1517. Products
3a–3d, 5a–5i, 7, 9, 11, 13, 15, and 17 were prepared using a similar
procedure and then fully characterized (see the Supporting Informa-
tion).
[10] K. Arnold, A. S. Batsanov, B. Davies, C. Grosjean, T. Schuetz, A.
Whiting, K. Zawatzky, Chem. Commun. 2008, 3879 – 3881.
[11] M. Li, T. Yang, D. J. Dixon, Chem. Commun. 2010, 46, 2191 –
2193.
Received: February 29, 2012
Published online: May 15, 2012
[12] a) J. A. McCubbin, H. Hosseini, O. V. Krokhin, J. Org. Chem.
2010, 75, 959 – 962; b) J. A. McCubbin, O. V. Krokhin, Tetrahe-
dron Lett. 2010, 51, 2447 – 2449.
[13] H. Zheng, M. Lejkowski, D. G. Hall, Chem. Sci. 2011, 2, 1305 –
1310.
[14] Boronic Acids—Preparation and Applications in Organic Syn-
thesis Medicine and Materials (Ed.: D. G. Hall), 2nd ed., Wiley-
VCH, Weinheim, 2011.
[15] E. Emer, R. Sinisi, M. G. Capdevila, D. Petruzziello, F.
De Vincentiis, P. G. Cozzi, Eur. J. Org. Chem.
2011, 647 – 666.
[16] The inferior efficacy of pentafluoro- versus tetra-
fluorophenylboronic acid (1a) is not due to
a difference in stability because both can be
recovered intact after the reaction. One explan-
ation is the possibility for additional substrate
Keywords: allylic compounds · boron · carbocycles · cyclization ·
heterocycles
.
[1] For example, the ACS Green Chemistry Institute Pharmaceutical
Roundtable ranked “alcohol activation for nucleophilic activa-
tion” the second most important priority area for green
chemistry research: D. J. C. Constable, P. J. Dunn, J. D. Hayler,
G. R. Humphrey, J. L. Leazer, R. J. Linderman, K. Lorenz, J.
Manley, B. A. Pearlman, A. Wells, A. Zaks, T. Y. Zhang, Green
Chem. 2007, 9, 411 – 420.
[2] I. Georgiou, G. Ilyashenko, A. Whiting, Acc. Chem. Res. 2009,
42, 756 – 768.
[3] a) K. Ishihara, S. Ohara, H. Yamamoto, J. Org. Chem. 1996, 61,
4196 – 4197; b) P. Wipf, X. Wang, J. Comb. Chem. 2002, 4, 656 –
660; c) K. Arnold, B. Davies, R. L. Giles, C. Grosjean, G. E.
Smith, A. Whiting, Adv. Synth. Catal. 2006, 348, 813 – 820; d) T.
À
activation from the relatively acidic ortho C H
bond of 1a (and 1b, 1c).
6190 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 6187 –6190