Furanosteroid studies. Improved synthesis of the A,B,C,E-ring core of viridin

Article abstract of DOI:10.1016/j.tetlet.2012.01.070

The tetracyclic core skeleton of the furanosteroid viridin is prepared in nine steps from readily available materials. The key step in the synthesis is a facile acid-promoted cyclodehydration of an aryloxyketone to prepare the benzofuran moiety. From this intermediate, the known target skeleton is prepared in four steps. This new synthesis is a six-step improvement over the previously reported one.

Full text of DOI:10.1016/j.tetlet.2012.01.070

Tetrahedron Letters 53 (2012) 1620–1623
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Furanosteroid studies. Improved synthesis of the A,B,C,E-ring core of viridin
⇑
Kristen C. Mascall, Peter A. Jacobi
Department of Chemistry, Dartmouth College, Hanover, NH 03755, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The tetracyclic core skeleton of the furanosteroid viridin is prepared in nine steps from readily available
materials. The key step in the synthesis is a facile acid-promoted cyclodehydration of an aryloxyketone to
prepare the benzofuran moiety. From this intermediate, the known target skeleton is prepared in four
steps. This new synthesis is a six-step improvement over the previously reported one.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 9 December 2011
Revised 13 January 2012
Accepted 18 January 2012
Available online 28 January 2012
Keywords:
Furanosteroids
Cyclodehydration
Benzofuran
Mukaiyama-aldol
The furanosteroids make up a small class of natural products
known for their novel pentacyclic skeleton—in addition to the
usual steroidal scaffold, there is a furan ring fused at the C4 and
C6 positions. This ring introduces considerable strain in the skele-
ton, and is the site of biological activity of these compounds. The
parent members of the furanosteroid class of natural products
are viridin (1) and wortmannin (2) (Fig. 1).1
O
O
AcO
MeO
OH
MeO
O
O
H
O
O
O
O
O
O
viridin (1)
wortmannin (2)
furanosteroid
skeleton
Isolated in 1945 and 1957, respectively,2a these natural prod-
ucts were long known for their high antifungal and anti-inflamma-
tory activity.2b More recently, wortmannin (2) was recognized as a
potent bimodal inhibitor of phosphoinositide-3 kinases (PI-3K),2c,d
a family of enzymes involved in the intracellular signaling pathway
essential for the growth and development of cells. Wortmannin (2)
has a strong affinity for the ATP-pocket of the enzyme, and binds
tightly through a series of hydrogen bonds. Although this interac-
tion is sufficient to inhibit enzymatic activity, the structure–
activity studies have shown that C20 of the furan ring in 2 binds
irreversibly to a lysine residue of the PI-3 kinase, rendering the
inhibition irreversible (Fig. 2).2e,f This position on the furan ring
is made highly electrophilic due to the flanking carbonyl groups.
It is believed that inhibitors of PI-3Ks related to 2 can be developed
into remedial agents for diseases characterized by rapid cell prolif-
eration, such as cancer.
Figure 1. Parent members of furanosteroid class, viridin (1) and wortmannin (2).
O
O
AcO
20 O
AcO
MeO
MeO
ring
opening
O
O
H
O
H
O
2
O
O
OH
20
HN
3
NH2
K802
K802
(p110 PI-3 kinase)
(p110 PI-3 kinase)
Figure 2. Covalent PI-3K inhibition of wortmannin (2).
synthetic targets in the ongoing discovery and development of
anti-cancer drugs.3 Although viridin (1) and wortmannin (2) have
been known for over half a century, only one synthesis of 1 has
been reported,4 and two syntheses of 2 (one de novo).5 In other
noteworthy work, the pentacyclic skeleton of viridin has been
prepared in nine steps via an ortho-quinone Diels–Alder reaction.6
Indeed, the furanosteroid skeleton itself is a synthetic challenge.7
Previously, the viridin core skeleton 4 has been prepared in 15
steps from commercially available materials, utilizing an intramo-
lecular Diels–Alder/retro Diels–Alder (DA) reaction between a
However, 2 is unselective and interacts with many other biolog-
ical targets, and the enhanced electrophilicity of the furan ring
makes 2 very susceptible to hydrolysis.1b These toxicity and insta-
bility issues make the natural products unsuitable for clinical use.
Thus, structural analogs of the furanosteroids are attractive
⇑
Corresponding author. Tel.: +1 603 646 3495; fax: +1 603 646 3946.
E-mail address: peter.jacobi@dartmouth.edu (P.A. Jacobi).
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2012.01.070

K. C. Mascall, P. A. Jacobi / Tetrahedron Letters 53 (2012) 1620–1623
1621
Mukaiyama-aldol
condensation
acid-promoted
MeCuCNLi
tBuOK
cyclodehydration
11
OMe
OMe
15
toluene, -78 °C to rt
DMSO, rt
HO
O
OMe
65%
99%
O
OH
9
OMe
6a
O
O
syn-4
OR
O
+
O
CO2Et
5
8, K2CO3
acetone, rt
93%
CF3SO3H
CH2Cl2, 0 °C to rt
OMe
6a
EtO2C
OMe
O
O
O
88%
Br
CO2Et
O
EtO2C
7
O
OMe
OMe
8
O
OH
EtO2C
O-alkylation
9
Scheme 3. Second generation cyclodehydration approach.
7
Scheme 1. Synthetic analysis of viridin model 4.
TMSI, CH2Cl2
tethered alkyne and oxazole as the key step.8 The final step of the
synthesis was a Mukaiyama-aldol cyclization which gave tetracy-
cle 4, with the desired syn-isomer predominating under kinetic
control due to a more favorable Burgi–Dunitz angle.9
OMe
OH
62 °C, sealed tube
O
O
CO2Et
CO2Et
6a
6b
O
In this letter, a new approach to the viridin core skeleton is
described, which utilizes an acid-promoted cyclodehydration of
an aryloxyketone to form the benzofuran.10 Our synthetic analysis
for 4 is outlined in Scheme 1. We envisioned that the final C1–C10
bond connection would be made via an intramolecular aldol con-
densation of 5, a transformation that was previously used in the
synthesis of 4.8 Aldehyde 5 would be derived in a sequence of steps
from 6a, which would be prepared via the cyclodehydration of
aryloxyketone 7.11 Finally, the cyclodehydration precursor 7 would
be prepared by O-alkylation of an appropriately functionalized
naphthol.
silyl triflate, 2,6-lut.
CH2Cl2, 0 °C
DIBAL
CH2Cl2, -78 °C
OR
OR
5c (98%)
5d (100%)
5e (94%)
5f (81%)
O
O
CO2Et
6c (78% from 6a)
6d (83% from 6a)
6e (78% from 6a)
6f (76% from 6a)
O
Ac2O, DMAP
DIBAL
CH2Cl2, -78 °C
6b
pyridine, rt
OH
5b
OAc
6g
O
O
67% from 6a
CO2Et
In the first synthetic approach to 4, we chose naphthol 12 as the
starting point, since this material could easily be prepared in two
steps from 2-naphthol (10) as reported in the literature.12 We pro-
posed that the necessary C4 methyl group could be installed at a
a) R = Me. b) R = H. c) R = TMS. d) R = TES.
e) R = TBS. f) R = TIPS. g) R = Ac
later stage. Alkylation of 12 with
a
-bromoketone 813 proceeded
Scheme 4. Synthesis of aldol cyclization precursors.
under standard conditions to give aryloxyketone 13. Cyclodehy-
dration of 13 was studied extensively using various acids and reac-
tion conditions, but it became apparent that 13 was a poor
substrate for cyclization. Of the acids tested,14 only PPA and
CF3SO3H induced cyclodehydration of 13, but benzofuran 14 was
consistently obtained in unsatisfactory yield (Scheme 2). Optimiza-
tion studies were performed with these acids but there was no
improvement in yield.
We postulated that the cyclodehydration reaction could be im-
proved by starting the synthesis with a more electron-rich naph-
thol such as 9, where the C4 methyl group was already in place.
A search of the literature revealed that naphthols with the required
substitution pattern of 9 were relatively uncommon. Despite this
apparent lack of precedent, the focus shifted to the preparation
of 9. After some synthetic studies, we were pleased to find that this
compound could be prepared in three steps from 2-naphthol (10).
As before, oxidative dearomatization of 10 gave 11 (cf. Scheme
2),12a and by exploiting the imbedded enone functionality, the C4
methyl group was introduced via conjugate addition of MeCuCNLi
to give ketone 15. This intermediate is on the same oxidation level
as the target naphthol, and a simple keto-enol tautomerization
followed by the elimination of a molecule of MeOH would give 9.
This transformation was serendipitously achieved in quantitative
t
yield using BuOK in DMSO. Thus, 9 can be prepared efficiently
on gram scale in good overall yield (41% from 10). Alkylation with
8 gave the key aryloxyketone 7, and gratifyingly, when 7 was
treated with CF3SO3H (5 equiv), benzofuran 6a was obtained in
high yield (Scheme 3).
Next, the focus of the synthesis turned to the Mukaiyama-aldol
cyclization to form the final C1–C10 bond as shown in the retro-
synthetic scheme (54). This key step was well-precedented, since
it was successfully used to prepare syn- and anti-4 in a favorable
4:1 ratio from aldehyde 5e (R = TBS). For this particular transfor-
mation, many reagents and conditions were screened, and it was
found that TiCl4 in CH2Cl2 at rt gave the best result.15 As an exten-
sion of this study, we chose to investigate the substrate effect in
the reaction by varying the protecting group ‘R’ on the phenol. In
addition to the usual class of Mukaiyama-aldol reactants, where
PhI(OAc)2
NaBH4
OMe
OMe
MeOH
63%
MeOH
47%
OMe
O
OH
OH
10
11
12
8, K2CO3
OH
OH
various
acids
TMSOTf
acetone, rt
EtO2C
O
OMe
OMe
14
avg. yield ~40%
95%
CH2Cl2, rt, 5 h
O
O
O
O
13
CO2Et
O
O
anti-4
syn-4
Scheme 2. First generation cyclodehydration approach.
Scheme 5. Isomerization of syn-4 with TMSOTf.

1622
K. C. Mascall, P. A. Jacobi / Tetrahedron Letters 53 (2012) 1620–1623
Table 1
Table 2
Intramolecular Mukaiyama-aldol cyclization studies of 5b–f
Cyclization studies of aldehyde 5a
OH
OH
OH
OH
O
O
+
+
O
O
O
O
OMe
OR
5b-f
O
O
O
O
O
O
5a
anti-4
syn-4
anti-4
syn-4
b) R = H. c) R = TMS. d) R = TES. e) R = TBS. f) R = TIPS.
Reagent
% Conversiona
syn-4b
anti-4b
5bb
Substrate
% Conversiona
syn-4b
anti-4b
5bb
1
TiCl4
BBr3
TMSI
TMSOTf
TiCl4/NaI
TiCl4/LiI
BBr3/NaI
TiCl4/TMSI
TiCl4/TMSI
TiCl4/TMSI
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
8
—
d
d
2c
3
1
2
3
4
5b
5c
5d
5e
5f
5c
5d
5e
5b
5e
0
100
100
100
100
100
100
90
—
21
55 (40)
80 (46)
89 (55)
—
—
40
(11)
—
—
7
100
72
34
—
4
5
6
7
15d
29
20
16
75
0
15
29
20
8
25 (20)
—
11 (10)
20 (11)
11 (11)
—
—
10
5
—
6c
7c
8c
9d
10d
100
100
40
—
8e
9f
10g
50 (30)
—
17 (8)
—
(12)
50
33
100
100
(22)
100
—
Reagents and conditions: reagent, CH2Cl2, rt; reaction progress monitored by TLC.
a
% conversion based on unreacted starting material.
Reagents and conditions: TiCl4, CH2Cl2, rt; reaction progress monitored by TLC.
Ratios determined by 1H NMR integrations; isolated yield in parentheses.
b
a
% conversion based on unreacted starting material.
c
Reaction run at À78 °C.
Ratios determined by 1H NMR integrations; isolated yields in parentheses. Low
b
d
Predominant decomposition.
isolated yields are due to the instability of syn-4 and anti-4 at room temperature.
e
c
TiCl4 was added one minute after adding TMSI.
TiCl4 and 5a were stirred together at rt for 35 min before adding TMSI.
TMSI and 5a were stirred together at rt for 35 min before adding TiCl4.
Reaction run at À78 °C.
f
d
TMSOTf instead of TiCl4.
g
R = silyl, we were interested in the cyclization potential of methyl
ether aldehyde 5a and phenol aldehyde 5b. With ample quantities
of 6a readily available, a small library of cyclization substrates was
prepared as outlined in Scheme 4. The silyl-protected phenol
aldehydes 5c–f were prepared by demethylating 6a with TMSI,16
then protecting the unstable phenol as the appropriate silyl ether
under standard conditions. Facile DIBAL reduction of the ester
group gave the desired aldehydes. By maintaining a low reaction
temperature and quenching at À78 °C, no over-reduction of
the ester was observed, even in the presence of excess DIBAL
reagent.
Phenol aldehyde 5b is susceptible to air oxidation, and it had to
be synthesized and used directly without purification. We found it
expedient to demethylate 6a, and then convert it to the corre-
sponding acetate 6g. In the subsequent step, treatment with DIBAL
reduced the ester cleanly and removed the acyl-protecting group
in the same pot to give 5b. Attempts to reduce the ester in 6b di-
rectly resulted in either incomplete conversion at À78 °C, or
over-reduction at higher temperature. Furthermore, demethyla-
tion of 5a was unsuccessful, resulting in decomposition.
Using the conditions developed for 5e, the Mukaiyama-aldol
reaction of 5b–f was studied (Table 1). We were pleased to find
that the ring closure occurred in every case (entries 2–5), except
for 5b, which gave no reaction (entry 1). Furthermore, the size of
the silyl group did affect the product ratios. In general, desilylation
was predominant with substrates bearing smaller silyl protecting
groups (5c, d), while more favorable results were achieved with
the larger groups. Thus, 5f (R = TIPS) gave an improved ratio of
syn/anti-4 (89:11; entry 5),17 compared to 5e (80:20; entry 4). Fi-
nally, lowering the reaction temperature was detrimental, giving
either complete desilylation (entries 6 and 7), or sluggish cycliza-
tion (entry 8).
Since 5b was unreactive toward TiCl4, other Lewis acids were
screened, and the only one that induced ring closure was TMSOTf
(entry 9). However, the product distribution was unfavorable
(syn/anti = 1:2), and the isolated yield was poor (33%). Although
these results were not synthetically useful, we were satisfied to
know that 5b was able to undergo cyclization.
The last substrate to be tested in the Mukaiyama-aldol cycliza-
tion was 5a (Table 2). Cyclization of this aldehyde would provide
the shortest route to tetracycle 4, since protecting group manipu-
lation would be avoided. Under the standard cyclization
conditions, no reaction was observed (entry 1). Other reagents
resulted in decomposition. When TiCl4 was combined with NaI or
LiI, a trace of anti-4 was detected by NMR (entries 5 and 6). The
combination of TiCl4 and TMSI gave an encouraging syn- to anti-4
ratio of 50:25 (entry 8). No reaction was observed when TiCl4
and 5a were allowed to stir together for 35 min before adding TMSI
(entry 9). Conversely, when TMSI and 5a were stirred together for
35 min before adding TiCl4 (entry 10), some cyclization was ob-
served, but the ratio of desired to undesired isomer was worse than
that observed in entry 8. To date, cyclization of 5a with TMSI/TiCl4
as shown in entry 8 gives the best ratio of syn- to anti-4.
The isomerization of syn-4 to its anti-isomer was also briefly
investigated. It was already shown that the ratio obtained from
cyclization of 5e was a kinetic one, and subjecting either isomer
to the standard reaction conditions (TiCl4, CH2Cl2, rt) resulted in
no equilibration.8 However, we noticed that cyclizations of 5b,
5c, and 5e with TMSOTf appeared to favor the formation of the
anti-isomer, and it was hypothesized that this reagent was promot-
ing equilibration of syn-4 and anti-4, presumably through 5b. This
hypothesis is supported by DFT calculations, which predict that
anti-4 is more stable in CH2Cl2 than syn-4 by approximately
2 kcal/mol, mainly due to differences in solvation energy.18 In the
event, when syn-4 was treated with TMSOTf at ambient tempera-
ture, the anti-isomer was almost immediately detectable by TLC
(Scheme 5). After 5 h, TLC and proton NMR analysis indicated that
all of the syn-isomer was converted to anti-4, although the conver-
sion was not quantitative (decomposition poses a problem in these
systems after prolonged reaction times). Nevertheless, we have
shown that we can equilibrate syn-4 to its thermodynamic isomer
with TMSOTf, supporting our hypothesis.
In conclusion, tetracycle 4 has been successfully prepared by
the new cyclodehydration approach, with an improvement in
product ratio. This synthesis is more efficient than the previous
route, and provides us with larger quantities of material in order

K. C. Mascall, P. A. Jacobi / Tetrahedron Letters 53 (2012) 1620–1623
1623
7. For synthetic studies toward the furanosteroid class of compounds, see: (a)
Moffatt, J. S. J. Chem. Soc. (C) 1966, 734; (b) Yasuchika, Y.; Kenji, H.; Kanematsu,
K. Chem. Commun. 1987, 515; (c) Broka, C. A.; Ruhland, B. J. Org. Chem. 1992, 57,
4888; (d) Carlina, R.; Higgs, K.; Older, C.; Randhawa, S.; Rodrigo, R. J. Org. Chem.
1997, 62, 2330; (e) Boynton, J.; Hanson, J. R.; Kiran, I. J. Chem. Res. (S) 1999, 638;
(f) Wright, D. L.; Robotham, C. V.; Aboud, K. Tetrahedron Lett. 2002, 43, 943; (g)
Sessions, E. H.; O’Connor, R. T.; Jacobi, P. A. Org. Lett. 2007, 9, 3221; (h) Muller,
K. M.; Keay, B. A. Synlett 2008, 1236; (i) Sundstrom, T. J.; Wright, D. L. Synlett
2010, 2875.
to study these tetracyclic systems in greater detail. Further trans-
formations of 4 and its derivatives are currently under investiga-
tion, and will be reported in due course.
Acknowledgments
Financial support of this work by the National Science Founda-
tion, CHE-0646876, and Dartmouth College is gratefully acknowl-
edged. We are grateful to Professor Russell Hughes of the
Dartmouth College Chemistry Department for carrying out DFT
calculations on compounds syn-4 and anti-4.18
8. Sessions, E. H.; Jacobi, P. A. Org. Lett. 2006, 8, 4125.
9. Burgi, H. B.; Dunitz, J. D. Acc. Chem. Res. 1983, 16, 153.
10. Cagniant, P.; Cagniant, D. In Advances in Heterocyclic Chemistry; Katritzky, A. R.,
Boulton, A. J., Eds.; Academic Press: New York, 1975; Vol. 18, pp 337–473.
11. Representative examples of cyclodehydration of aryloxyketones in synthesis:
(a) Hirai, Y.; Doe, M.; Kinoshita, T.; Morimoto, Y. Chem. Lett. 2004, 33, 136; (b)
Marugán, J. J.; Manthey, C.; Anaclerio, B.; LaFrance, L.; Lu, T.; Markotan, T.;
Leonard, K. A.; Crysler, C.; Eisennagel, S.; Dasgupta, M.; Tomczuk, B. J. Med.
Chem. 2005, 48, 926; (c) Kedrowski, B. L.; Hoppe, R. W. J. Org. Chem. 2008, 73,
5177; (d) Davydenko, I. G.; Slominskii, Y. L.; Kalchenko, O. I.; Gutov, A. V.;
Chernega, A. N.; Tolmachev, A. I. Russ. Chem., Bull. Int. Ed 2008, 57, 159; (e) Kim,
I.; Choi, J. Org. Biomol. Chem. 2009, 7, 2788; (f) Kim, K.; Kim, I. Org. Lett. 2010, 12,
5314.
Supplementary data
Supplementary data (experimental and NMR spectra for all new
compounds) associated with this article can be found, in the online
version, at doi:10.1016/j.tetlet.2012.01.070.
12. (a) Mal, D.; Roy, H. N.; Hazra, N. K.; Adhikari, S. Tetrahedron 1997, 53, 2177; (b)
Dolson, M. G.; Jackson, D. K.; Swenton, J. S. J. Chem. Soc., Chem. Commun 1979,
327.
References and notes
13. Stearns, B. A.; Clark, R. WO 2010/085820 A2, 2010.
14. Representative examples of acids tested include: TiCl4, AlCl3, BF3ÁOMe2, BCl3,
Ga(OTf)3, Bi(OTf)3, CF3CO2H, CF3SO3H, p-TsOH, H2SO4, PPA, P2O5-CH3SO3H
(Eaton’s reagent). Except for CF3SO3H and PPA, these acids resulted in either
dealkylation of 13, decomposition, or no reaction.
1. For reviews, see: (a) Hanson, J. R. Nat. Prod. Rep. 1995, 12, 381; (b) Wipf, P.;
Halter, R. J. Org. Biomol. Chem. 2005, 3, 2053.
2. (a) Brian, P. W.; McGowan, J. C. Nature (London) 1945, 156, 144; (b) Vischer, E.
B.; Howland, S. R.; Raudnitz, H. Nature 1950, 165, 528; (c) Powis, G.;
Bonjouklian, R.; Berggren, M. M.; Gallegos, A.; Abraham, R.; Ashendel, C.;
Zalkow, L.; Matter, W. F.; Dodge, J.; Grindey, G.; Vlahos, C. J. Cancer Res. 1994,
54, 2419; (d) Wymann, M. P.; Bulgarelli-Leva, G.; Zvelebil, M. J.; Pirola, L.;
VanHaese-Broeck, B.; Waterfield, M. D.; Panayotou, G. Mol. Cell. Biol. 1996, 16,
1722; (e) Norman, B. H.; Paschal, J.; Vlahos, C. J. Bioorg. Med. Chem. Lett. 1995, 5,
1183; (f) Norman, B. H.; Shih, C.; Toth, J. E.; Ray, J. E.; Dodge, J. A.; Johnson, D.
W.; Rutherford, P. G.; Schultz, R. M.; Worzalla, J. F.; Vlahos, C. J. J. Med. Chem.
1996, 39, 1106.
3. Wipf, P.; Minion, D. J.; Halter, R. J.; Berggren, M. I.; Ho, C. B.; Chiang, G. G.;
Kirkpatrick, L.; Abraham, R.; Powis, G. Org. Biomol. Chem. 1911, 2004, 2.
4. Anderson, E. A.; Alexanian, E. J.; Sorensen, E. J. Angew. Chem., Int. Ed. 2004, 43,
1998.
5. (a) Sato, S.; Nakada, M.; Shibasaki, M. Tetrahedron Lett. 1996, 37, 6141; (b)
Mizutani, T.; Honzawa, S.; Tosaki, S.-Y.; Shibasaki, M. Angew. Chem., Int. Ed.
2002, 41, 4680.
15. Sessions, E. H. Ph.D. Dissertation, Dartmouth College, 2006.
16. We found that commercially available TMSI worked best in this reaction. Other
reagents such as BBr3 (McOmie, J. F. W.; Watts, M. L.; West, D. E. Tetrahedron,
1968, 24, 2289) and in-situ generated TMSI (Olah, G. A.; Narang, S. C.; Balaram
Gupta, B. G.; Malhotra, R. J. Org. Chem. 1979, 44, 1247) did not work as well. For
a review of ether cleavage, see: Bhatt, M. V.; Kulkarni, S. U. Synthesis, 1983, 249.
17. Representative procedure for the Mukaiyama-aldol cyclization of 5f–4.
Aldehyde 5f (0.20 mmol, 1.0 equiv) is dissolved in dry CH2Cl2 (21 mL) at rt
under nitrogen, and TiCl4 (0.42 mmol of a 1.0 M solution in CH2Cl2, 2.1 equiv) is
added dropwise. The resulting dark–green reaction mixture is stirred at rt for
45–50 min (prolonged reaction times cause decomposition), then diluted with
CH2Cl2 and washed with water. The aqueous layer is extracted twice with
CH2Cl2 and the combined organic layers are washed with brine and dried over
MgSO4. After concentration, the crude product mixture is purified by flash
chromatography to give syn-4 and anti-4 as unstable brown oils.
18. Calculations were carried out using Jaguar 7.7 (B3LYP/LACV3P⁄ ⁄ + +); Poisson-
Boltzmann solvation method.
6. (a) Souza, F. E. S.; Rodrigo, R. Chem. Commun. 1947, 1999, 19; (b) Lang, Y.; Souza,
F. E. S.; Xu, X.; Taylor, N. J.; Assoud, A. Rodrigo, R J. Org. Chem. 2009, 74, 5429.