BF2OBn·OEt2: A Lewis acid, its use in a regio- and stereoselective opening of trisubstituted epoxides, and its application towards amphidinolide C and F

Article abstract of DOI:10.1002/ejoc.201201331

The generation of a new Lewis acid (BF₂OBn·OEt ₂) has been reported, and its usefulness has been demonstrated in the regio- and stereoselective ring-opening of trisubstituted epoxides. This Lewis acid is one in a series of new Lewis acids generated from BF₃· OEt₂ that display varying levels of Lewis acidity. When paired with a modified Shi epoxidation protocol, highly functionalized propionate units, such as those found in a wide variety of natural products, can be accessed. In conjunction with a Mukaiyama oxidative cyclization employing our second generation catalyst Co(nmp)₂, this procedure ultimately culminated in the shortest and highest yielding route towards the methyl-substituted trans-tetrahydrofuran (trans-THF) fragment present in amphidinolide C, C2, and F. The new Lewis acid BF₂OBn·OEt₂ was generated by an anionic redistribution between BF₃·OEt₂ and benzyloxytrimethylsilane and used in a regio- and stereoselective epoxide ring-opening reaction. The utility of this reaction was demonstrated by the conversion of commercially available 2,3-dihydrofuran into the C-1-C-9 fragment of amphidinolide C and F in five steps and 49 % overall yield. Copyright

Full text of DOI:10.1002/ejoc.201201331

Job/Unit: O21331
/KAP1
Date: 10-12-12 16:52:24
Pages: 6
FULL PAPER
DOI: 10.1002/ejoc.201201331
BF2OBn·OEt2: A Lewis Acid, Its Use in a Regio- and Stereoselective Opening
of Trisubstituted Epoxides, and Its Application towards Amphidinolide C and F
Nicholas A. Morra[a] and Brian L. Pagenkopf*[a]
Keywords: Natural products / Asymmetric catalysis / Regioselectivity / Cyclization / Lewis acids / Epoxidation
The generation of a new Lewis acid (BF2OBn·OEt2) has been
reported, and its usefulness has been demonstrated in the
regio- and stereoselective ring-opening of trisubstituted ep-
oxides. This Lewis acid is one in a series of new Lewis acids
generated from BF3·OEt2 that display varying levels of Lewis
acidity. When paired with a modified Shi epoxidation proto-
col, highly functionalized propionate units, such as those
found in a wide variety of natural products, can be accessed.
In conjunction with a Mukaiyama oxidative cyclization em-
ploying our second generation catalyst Co(nmp)2, this pro-
cedure ultimately culminated in the shortest and highest
yielding route towards the methyl-substituted trans-tetra-
hydrofuran (trans-THF) fragment present in amphidinol-
ide C, C2, and F.
alyst Co(nmp)2 has resulted in a general and high-yielding
synthesis of the trans-THF rings starting from their corre-
sponding 4-pentenols.[3] With that in mind, the initial retro-
synthetic disconnection involves the formation of the trans-
THF ring in 2 through the cyclization of the methyl-substi-
tuted pentenol 3 (see Scheme 1). This key synthetic interme-
diate would be secured by a regio- and stereoselective ring
opening of epoxide 4, which would be accessed through a
Shi epoxidation of diene 5.
Introduction
The synthesis of amphidinolide C (1, see Figure 1) has
been approached by many groups, and although this has
resulted in the completion of several fragments, no total
synthesis has been reported to date.[1,2] In particular, the C-
1–C-9 fragment has attracted considerable synthetic atten-
tion because of its stereochemical complexity. It contains a
methyl-substituted trans-tetrahydrofuran (trans-THF) ring,
an anti-diol moiety, and an exocyclic olefin that is part of
an unusual diene system. An efficient and concise synthesis
of the methyl-substituted trans-THF ring is central to the
total synthesis of amphidinolide C. Herein, we describe a
short synthesis for the C-1–C-7 fragment of amphidinol-
ide C, which features a new Lewis acid and an aerobic oxi-
dative cyclization to form the trans-THF ring.
Scheme 1. Retrosynthetic analysis.
Figure 1. Amphidinolide C.
Results and Discussion
Our recent work on the Mukaiyama oxidative cyclization
reaction and the development of the second generation cat-
The synthesis began with unsaturated alcohol 5, which
was prepared by a 1,2-metallate rearrangement of dihy-
drofuran in a one-pot procedure (see Scheme 2).[4] The re-
sulting alcohol 5 was functionalized with a variety of pro-
tecting groups to give 6a–6c, which underwent epoxidation
by treatment with meta-chloroperoxybenzoic acid
(mCPBA) to generate a set of racemic epoxides that were
used to evaluate the subsequent ring-opening reaction.
[a] The University of Western Ontario, Department of Chemistry,
London, Ontario N6A 5B7, Canada
Fax: +1-519-661-3022
E-mail: bpagenko@uwo.ca
Homepage: www.uwo.ca/chem
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201201331.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O21331
/KAP1
Date: 10-12-12 16:52:24
Pages: 6
N. A. Morra, B. L. Pagenkopf
FULL PAPER
bution, in which one of the fluorines was replaced with a
less electronegative group. We had previously achieved suc-
cess with this strategy, when applied to the more Lewis
acidic species BF2OTf·OEt2 and BF2OMs·OEt2 (see
Scheme 3), which were used in the direct reduction of esters
to ethers.[7] Thus, the treatment of BF3·OEt2 with (benz-
yloxy)trimethylsilane (TMSOBn) generated the modified
Lewis acid BF2OBn·OEt2 (11), which displayed a 19F NMR
signal at –151.5 ppm that is indicative of decreased Lewis
acidity in comparison to BF3·OEt2 (see Scheme 3).[8]
Scheme 2. Synthesis of trisubstituted epoxides rac-7a, rac-7b, and
rac-7c (TBS = tert-butyldimethylsilyl).
The conversion of the monoepoxides 7a–7c to the desired
alcohols required a regioselective hydride delivery to the
more hindered carbon. Initially, we explored the Hutchins
protocol, which proceeds by an apparent SN2 reaction and
occurs with the inversion of stereochemistry.[5] Unfortu-
nately, treatment of epoxide 7a under the standard condi-
tions (BF3·OEt2, NaCNBH3) gave results consistent with a
Scheme 3. Generation of two new Lewis acids by anionic redistri-
bution of BF3·OEt2.
Gratifyingly, 11 displayed sufficient Lewis acidity to fa-
premature epoxide ring opening, and reactions of the pre- cilitate efficiently the desired SN2 conversion of racemic ep-
sumed carbocation intermediate gave rise to an unfavorable oxide rac-7c to alcohol rac-8, without promoting the
mixture of diastereomers through either SN1 hydride deliv- troublesome side reactions encountered by using BF3·OEt2
ery to give rac-8 (see Table 1, Entry 1), elimination to give (see Table 1, Entry 11). The p-methoxybenzyl-protected
rac-9 (see Table 1, Entry 2), or a pinacol-like hydride shift (PMB-protected) epoxide rac-7c gave the most consistent
to give rac-10 (see Table 1, Entries 3 and 4). Reactions car- results for the epoxide ring-opening reaction.
ried out in diethyl ether with a slow addition of the Lewis
The Shi epoxidation protocol appeared to be an ideal
acid resulted in an increase in yield, but there was no fur- candidate for securing an enantioselective route to epoxide
ther improvement in the diastereomeric ratio (dr, see 7c, as this reaction generally shows a strong preference for
Table 1, Entries 5–8). A variety of Lewis acids were scre- unactivated trisubstituted olefins over monosubstituted
ened to achieve the desired transformation, but these ones.[9] Although the enantiomer of the Shi catalyst (12),
attempts were unsuccessful (see Table 1, Entries 9 and 10).[6] which is derived from l-fructose, would be required for this
Ultimately, we was decided to modify the BF3·OEt2 by at- application, it is accessible in five steps from inexpensive l-
tenuating its Lewis acidity through an anionic redistri- sorbose.[10] Unfortunately, initial attempts at the enantiose-
Table 1. Epoxide opening results.
Entry
Epoxide
Lewis acid
[4 equiv.]
Solvent
Addition time
[h]
Yield
9
%
dr of 8
(anti/syn)
8
%
10
%
1
2
3
4
5
6
7
8
7a
7b
7a
7a
7c
7c
7c
7c
7c
7c
7c
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
BF3·OEt2
InBr3
THF
CH2Cl2
THF
THF
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
–
3
0.5
3
–
0.5
3
4
–
4
4
20
–
–
99
–
–
–
–
–
–
–
–
–
25
15
–
–
–
–
–
–
–
2:1
–
50
65
23
51
66
90
0[a]
0[a]
91
2:1
2:1
2:1
2:1
2:1
2:1
–
9
10
11
BEt3
BF2OBn·OEt2 (11)
–
–
–
Ͼ20:1
[a] Starting material recovered.
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O21331
/KAP1
Date: 10-12-12 16:52:24
Pages: 6
Regio- and Stereoselective Opening of Epoxides
lective synthesis of epoxide 7c resulted in a disappointing Table 3. Optimization of oxidative cyclization.
yield (61%) and selectivity, giving a 3:1 ratio of monoepox-
ide 7c to bis(epoxide) 13 (see Table 2, Entry 1).
Table 2. Synthesis of epoxide 7c.
Entry Catalyst
Loading
[mol-%]
Temp.
[°C]
Time
[h]
% Yield 16
Entry
Addition
time [h]
12
Recovered 6c
7c
13
1
2
3
4
5
6
7
8
9
10
14
15
15
15
15
15
15
15
15
15
15
15
15
15
15
55
55
45
35
22
35
35
55
55
55
16
16
16
16
16
16
16
1
0
[mol-%]
[%]
[%]
[%]
10
55
81
1
2
3
2
4
6
4
4
4
4
20
20
20
20
35
25
20
14
32
45
62
0
0
11
61
50
40
23
75
74
60
20
11
7
67 (85[a]
65[b]
)
)
4[a]
5[b]
6[b]
7[b]
8
15
15[c]
15[c]
10[c]
5[c]
80[b]
13
11
8
91
1
16
94[b]
77 (92[a]
[a] Reaction conducted at –10 °C. [b] Catalyst added in four por-
tions over 4 h.
[a] Yields based on recovered starting material. [b] Reaction per-
formed on a 15 mmol scale. [c] Catalyst was preactivated.
Attempts were made to improve the selectivity by in-
creasing the addition time of the base and Oxone® to 4 or
6 h. This did lead to an increase in the selectivity, but there
was a decrease in overall yield (see Table 2, Entries 2 and
3). A lower reaction temperature resulted in a decrease in
both the yield and selectivity (see Table 2, Entry 4). Because
the Shi catalyst could decompose during the course of the
reaction,[11] it was added portionwise. At higher catalyst
loadings, complete conversion was finally achieved (see
Table 2, Entries 5 and 6) with selectivities of an approxi-
mate 7:1 ratio of the desired product 7c to bis(epoxide) 13.
Further attempts to decrease the amount of catalyst re-
sulted in incomplete conversions (see Table 2, Entry 7).
With a concise route towards pentenol 8 secured, we fo-
cused our attention on the oxidative cyclization to form the
trans-THF ring (see Table 3). The first generation catalyst
Co(modp)2 (14) had previously been shown to be incompat-
ible with the easily oxidized PMB group,[3] and attempts to
cyclize 8 with this catalyst were unsuccessful (see Table 3,
Entry 1). Using the standard oxidation conditions, the sec-
ond generation catalyst Co(nmp)2 (15) also afforded little
success in forming trans-THF ring 16 (see Table 3, Entry 2).
To reduce the amount of overoxidation byproducts formed
during the course of the reaction, lower reaction tempera-
tures were examined, and as a result, an optimal yield of
81% was obtained when the reaction was conducted at
35 °C (see Table 3, Entries 3–5). It is noteworthy that even
at room temperature a comparable yield based on recovered
starting material (BORSM) of 85% was observed. Upon
scale-up of the cyclization at a lower temperature, the yields
were uncharacteristically erratic (see Table 3, Entry 6), and
we speculated that the peroxide used during the activation
of the catalyst could be contributing to the overoxidation
byproducts.
Thus, an alternative protocol was performed, whereby
the catalyst was activated prior to the introduction of the
pentenol to ensure that no peroxides were present. Initial
reactions using the preactivated Co(nmp)2 (15) provided
significant advantages in terms of the reproducibility of the
yield (see Table 3, Entry 7), but overoxidation remained
problematic. Eventually, careful monitoring of the reaction
by TLC resulted in a surprising finding, that is, the reaction
went to completion after just 1 h (see Table 3, Entry 8).
Similar reactions typically require more than 12 h. Further
optimization of the reaction conditions showed that a cata-
lyst loading of 10 mol-% resulted in the highest yield of
16 (94%) and the cleanest reaction, whereas lower catalyst
loadings led to incomplete conversions (see Table 3, En-
tries 9 and 10). These optimized conditions proved repro-
ducible on a multigram scale.
Conclusions
In summary, we have reported on a highly selective Shi
epoxidation of a skipped diene. This was followed by a re-
ductive epoxide ring-opening reaction that was mediated by
the new Lewis acid BF2OBn·OEt2 to provide a compound
containing a useful propionate unit. Implementation of
these procedures for the synthesis of the C-1–C-9 fragment
of amphidinolide C and F will be disclosed elsewhere.
Experimental Section
(Z)-1-Methoxy-4-[(4-methylhepta-3,6-dienyloxy)methyl]benzene
(6c): To a suspension of NaH (780 mg, 32.5 mmol, 1.3 equiv.) in
THF (150 mL) at 0 °C was added freshly prepared p-meth-
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O21331
/KAP1
Date: 10-12-12 16:52:24
Pages: 6
N. A. Morra, B. L. Pagenkopf
FULL PAPER
oxybenzyl bromide (PMBBr, 6.53 g, 32.5 mmol, 1.3 equiv.) fol-
lowed by the addition of alcohol 5 (2.15 g, 25 mmol, 1.0 equiv.).
The ice bath was removed, and after approximately 16 h, the reac-
tion was poured into a half-saturated solution of NH4Cl (100 mL)
in ice water (200 mL). The resulting mixture was stirred for 5 min,
and then the aqueous layer was extracted with EtOAc
(3ϫ150 mL). The combined organic extracts were washed with
brine, dried with MgSO4, and then filtered through a thin pad of
packed Celite. The solvent was removed under reduced pressure,
and the crude oil was purified by flash chromatography (10 %
EtOAc/hexane) to yield PMB ether 6c (5.54 g, 22.5 mmol, 90%) as
a colorless oil; Rf = 0.42 (10 % EtOAc/hexane). 1H NMR
(600 MHz, CDCl3): δ = 7.25 (d, J = 8.8 Hz, 2 H), 6.87 (d, J =
8.8 Hz, 2 H), 5.72 (ddt, J = 16.7, 10.1, 6.4 Hz, 1 H), 5.22 (t, J =
6.7 Hz, 1 H), 5.03–4.96 (m, 1 H), 4.44 (s, 2 H), 3.78 (s, 3 H), 3.42
(t, J = 7.0 Hz, 2 H), 2.76 (d, J = 6.4 Hz, 2 H), 2.31 (q, J = 7.0 Hz,
2 H), 1.68 (s, 3 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 159.1,
135.9, 135.1, 130.6, 129.2, 121.8, 115.2, 113.7, 72.5, 69.8, 55.3, 36.5,
28.5, 23.4 ppm. HRMS: calcd. for C16H22O2 246.1619; found
246.1615.
BF2OBn·OEt2 solution was stable at room temp., or it was refriger-
ated for several weeks. Solvents other than diethyl ether caused
decomposition, and, therefore, diethyl ether was used for the char-
acterization and reactions. 19F NMR (375 MHz, Et2O): δ =
–151.5 ppm. Trifluorotoluene (–63.9 ppm) was used as an internal
standard.
(3S,4R)-1-(4-Methoxybenzyloxy)-4-methylhept-6-en-3-ol (8): To a
flask charged with NaCNBH3 (255 mg, 4.0 mmol, 4.0 equiv.) in di-
ethyl ether (15 mL) was added epoxide 7c (262 mg, 1.0 mmol,
1.0 equiv.). To the vigorously stirred solution was added a solution
of BF2OBn·OEt2 (ca. 1.0 m, 4.0 mL, 4.0 mmol, 4.0 equiv.) by sy-
ringe pump over 4 h. After the addition was complete, the reaction
was stirred for 15 min and then was poured into a half-saturated
solution of aqueous sodium hydrogen carbonate (100 mL). The
mixture was transferred to a separatory funnel, and the aqueous
layer was extracted with EtOAc (3ϫ50 mL). The combined organic
extracts were washed with brine, dried with MgSO4, and filtered
through a thin pad of packed Celite. The solvent was removed un-
der reduced pressure, and the crude oil was purified by flash
chromatography (30% EtOAc/hexane) to yield alcohol 8 (240 mg,
0.91 mmol, 91%) as a yellow oil; Rf = 0.50 (40% EtOAc/hexane).
1H NMR (600 MHz, CDCl3): δ = 7.24 (d, J = 8.7 Hz, 2 H), 6.86
(d, J = 8.7 Hz, 2 H), 5.79 (ddt, J = 17.3, 10.0, 7.1 Hz, 1 H), 5.04–
4.97 (m, 2 H), 4.45 (m, 2 H), 4.45 (s, 2 H), 3.79 (s, 3 H), 3.71 (dt,
J = 9.5, 4.9 Hz, 1 H), 3.61 (q, J = 6.4 Hz, 2 H), 3.00 (d, J = 2.3 Hz,
1 H), 2.30–2.26 (m, 1 H), 1.90 (dt, J = 13.9, 8.3 Hz, 1 H), 1.73–
1.70 (m, 2 H), 1.64–1.59 (m, 1 H), 0.86 (d, J = 6.4 Hz, 3 H) ppm.
13C NMR (150 MHz, CDCl3): δ = 159.3, 137.6, 130.0, 129.3, 115.8,
113.8, 75.2, 73.0, 69.4, 55.3, 38.6, 36.9, 32.8, 15.1 ppm. HRMS:
calcd. for C16H24O3 264.1725; found 264.1725. [α]2D0 = +1.73 (c =
1.0, CHCl3). The enantiomeric excess (ee) was determined by using
(R)-Mosher’s analysis to be 85%.
(2R,3S)-2-Allyl-3-[2-(4-methoxybenzyloxy)ethyl]-2-methyloxirane
(7c): To a flask charged with diene 6c (2.46 g, 10 mmol, 1.0 equiv.)
was added dimethoxymethane (DMM, 100 mL), acetonitrile
(50 mL), buffer[12] (100 mL), 12 (157 mg), and Bu4N·HSO4 (50 mg,
catalytic), and the resulting mixture was cooled to 0 °C. A syringe
pump was fitted with two 60 mL syringes. One was charged with
K2CO3 (6.90 g) in water (60 mL), and the second was charged with
Oxone® (6.90 g) in water (60 mL). The K2CO3 and Oxone® solu-
tions were added to the vigorously stirred solution over 4 h, and
during this time to the reaction mixture were also added additional
amounts of 12 portionwise at the 1 h, 2 h, and 3 h timepoints
(157 mg per addition, 630 mg total for 4 additions, 2.50 mmol,
0.25 equiv.). After the additions of the base and Oxone® were com-
plete, the reaction was stirred for 15 min, and then hexanes
(200 mL) were added. The solution was transferred to a separatory
funnel, and the aqueous layer was extracted with hexanes
(4ϫ100 mL). The combined organic extracts were washed with
brine, dried with MgSO4, and filtered through a thin pad of packed
Celite. The solvent was removed under reduced pressure, and the
crude oil was purified by flash chromatography (20% EtOAc/hex-
ane) to yield monoepoxide 7c (1.93 g, 7.40 mmol, 74%) and bis-
(epoxide) 13 (305 mg, 1.10 mmol, 11%) as yellow oils; Rf = 0.17
(10% EtOAc/hexane). 1H NMR (400 MHz, CDCl3): δ = 7.24 (d, J
= 8.6 Hz, 2 H), 6.85 (d, J = 8.6 Hz, 2 H), 5.77 (ddt, J = 17.2, 10.2,
7.0 Hz, 1 H), 5.10–5.05 (m, 2 H), 4.43 (ABd, J = 11.7 Hz, 2 H),
3.59–3.56 (m, 2 H), 2.86 (dd, J = 7.4, 4.7 Hz, 1 H), 2.30 (dd, J =
7.0, 7.0 Hz, 1 H), 2.18 (dd, J = 7.0, 7.0 Hz, 1 H), 1.98–1.89 (m, 1
H), 1.77–1.68 (m, 1 H), 1.25 (s, 3 H) ppm. 13C NMR (100 MHz,
CDCl3): δ = 159.2, 133.5, 130.4, 129.3, 117.8, 113.8, 72.8, 67.3,
61.9, 60.1, 55.3, 37.9, 29.4, 22.1 ppm. HRMS: calcd. for C16H22O3
262.1569; found 262.1576.
{(2R,4R,5S)-5-[2-(4-Methoxybenzyloxy)ethyl]-4-methyltetrahydro-
furan-2-yl}methanol (16)
Procedure to Preactivate Co(nmp)2: To a flask charged with
Co(nmp)2 (15, 452 mg, 0.8 mmol, 0.1 equiv.) and iPrOH (100 mL)
was added tBuOOH (5.33 m solution, 0.2 mL, 1.08 mmol,
0.14 equiv.). The reaction was heated to 55 °C under oxygen for
1 h, and then the solvent was removed under reduced pressure. The
activated Co(nmp)2 was dried under high vacuum (0.1 Torr) for
5 min to ensure that any remaining peroxide was removed.
Cyclization: The preactivated Co(nmp)2 (15, 0.8 mmol, 0.1 equiv.)
was diluted with iPrOH (100 mL), to this mixture was added
alcohol 8 (2.06 g, 7.8 mmol, 1 equiv.). The reaction was heated to
55 °C under oxygen for exactly 1 h and then allowed to cool to
room temp. The solvent was removed under reduced pressure and
then under high vacuum (0.1 Torr) to remove all trace amounts of
iPrOH. The crude mixture was diluted with EtOAc (40 mL) and
filtered through a thin pad of silica (Ͻ1 cm) over packed Celite to
remove the catalyst. The pad was washed with EtOAc (400 mL),
and the filtrate was concentrated under reduced pressure to give
THF-alcohol 16 (2.05 g, 7.34 mmol, 94%) as a yellow oil that was
used without further purification. (The product rapidly decom-
poses, and the decomposition product gives characteristically broad
peaks at δ = 3.65 and 3.45 ppm. The presence of the decomposition
product also results in the loss of the fine-splitting patterns, and
the peaks are reported as multiplets.) 1H NMR (600 MHz, CDCl3):
δ = 7.25 (d, J = 8.6 Hz, 2 H), 6.87 (d, J = 8.6 Hz, 2 H), 4.43 (d, J
= 2.0 Hz, 2 H), 4.06 (ddt, J = 9.4, 6.2, 3.1 Hz, 1 H), 3.79 (s, 3 H),
3.62–3.48 (m, 4 H), 2.09–2.03 (m, 1 H), 1.94–1.85 (m, 2 H), 1.73–
1.65 (m, 1 H), 1.37–1.29 (m, 1 H), 1.01 (d, J = 6.6 Hz, 3 H) ppm.
13C NMR (150 MHz, CDCl3): δ = 159.1, 130.6, 129.2, 113.7, 82.4,
(Benzyloxy)difluoroborane–Diethyl Ether (11): To a round-bot-
tomed flask charged with TMSOBn (1.90 g, 10.5 mmol,
1.05 equiv.) in diethyl ether (100 mL) and fitted with a rubber sep-
tum and a balloon filled with argon (20.5 gauge needle) was added
BF3·OEt2 (1.26 mL, 10 mmol, 1.0 equiv.). To facilitate the evapora-
tion of the TMSF, the septum was pierced with another 20.5 gauge
needle. The argon balloon was replaced as necessary, and the solu-
tion evaporated to a viscous oil in about 1 h. The argon flow was
maintained for an additional 10 min.[13] To the residue was added
an additional portion of diethyl ether (8 mL) to give an approxi-
mately 1.0 m solution of BF2OBn·OEt2 (11).[14] It may be necessary
to repeat the evaporation process.[1 4] If well sealed, the
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O21331
/KAP1
Date: 10-12-12 16:52:24
Pages: 6
Regio- and Stereoselective Opening of Epoxides
Wang, N. A. Morra, H. Zhao, J. S. T. Gorman, V. Lynch, R.
McDonald, J. F. Reichwein, B. L. Pagenkopf, Can. J. Chem.
2009, 87, 328–334.
78.3, 72.6, 67.4, 65.2, 55.3, 40.1, 36.6, 34.3, 16.4 ppm. HRMS:
calcd. for C16H24O4 280.1675; found 280.1667.
[4] a) K. Jarowicki, P. J. Kocienski, L. Qun, Org. Synth. 2002, 79,
11–13; b) K. Jarowicki, P. J. Kocienski, Synlett 2005, 167–169.
[5] R. O. Hutchins, I. M. Taffer, W. Burgoyne, J. Org. Chem. 1981,
46, 5214–5215.
Supporting Information (see footnote on the first page of this arti-
cle): Copies of NMR spectra.
[6] The other Lewis acids screened are MgBr2, ZnCl2, TiCl4,
SnCl4, Yb(OTf)3, and Sc(OTf)3.
Acknowledgments
[7] N. A. Morra, B. L. Pagenkopf, Synthesis 2008, 511–514.
[8] The 19F NMR peak shift can be correlated to the strength of
the Lewis acid. The order of the strongest to weakest is
BF2OTf·OEt2 (–146.9 ppm), BF2OMs·OEt2 (–148.3 ppm), and
BF2OBn·OEt2 (–151.5 ppm). Trifluorotoluene (–63.9 ppm) was
used as an internal standard.
We thank the University of Western Ontario and the National Sci-
ences and Engineering Research Council of Canada (NSERC) for
their financial assistance. N. M. thanks NSERC for a graduate fel-
lowship (CGS-D3).
[9] Z.-X. Wang, Y. Tu, M. Frohn, J.-R. Zhang, Y. Shi, J. Am.
Chem. Soc. 1997, 119, 11224–11235.
[1] a) N. A. Morra, B. L. Pagenkopf, Org. Lett. 2011, 13, 572–575;
b) for C-18–C-34, see: S. Roy, C. D. Spilling, Org. Lett. 2010,
12, 5326–5329; c) for C-1–C-5, see: M. P. Paudyal, N. P. Rath,
C. D. Spilling, Org. Lett. 2010, 12, 2954–2957; d) for C-1–C-9,
see: L. Ferrié, B. Figadére, Org. Lett. 2010, 12, 4976–4979; e)
for C-7–C-20, see: S. Mahapatra, R. G. Carter, Org. Biomol.
Chem. 2009, 7, 4582–4585; f) for C-18–C-29, see: A. Arm-
strong, C. Pyrkotis, Tetrahedron Lett. 2009, 50, 3325–3328; g)
for C-1–C-9, see: R. H. Bates, J. B. Shotwell, W. R. Roush, Org.
Lett. 2008, 10, 4343–4346; h) for C-19–C-34, see: D. K. Mo-
hapatra, H. Rahaman, M. S. Chorghade, M. K. Gurjar, Synlett
2007, 567–570; i) for C-11–C-29, see: J. B. Shotwell, W. R.
Roush, Org. Lett. 2004, 6, 3865–3868; j) for C-1–C-10 and C-
17–C-29, see: T. Kubota, M. Tsuda, J. Kobayashi, Tetrahedron
2003, 59, 1613–1625; k) for C-1–C-9, see: H. Ishiyama, M. Ishi-
bashi, J. Kobayashi, Chem. Pharm. Bull. 1996, 44, 1819–1822.
[2] For the synthesis of amphidinolide F, which was recently re-
ported, see: S. Mahaparta, R. G. Carter, Angew. Chem. Int. Ed.
2012, 32, 7948–7951.
[10] M.-X. Zhao, Y. Shi, J. Org. Chem. 2006, 71, 5377–5379.
[11] The catalyst can be deactivated by a competing Baeyer–Villiger
oxidation, which is disfavored at a higher pH.
[12] The buffer is a 0.05 m solution of Na2B2O7·H2O in 1ϫ10–4
m
aqueous Na2EDTA.
[13] This setup was found to be optimal for production of
BF2OBn·OEt2 without the formation of BF(OBn)2·OEt2. Al-
ternative procedures were examined including heating, reduced
pressure, and nitrogen flow through a Schlenk line, but they
were less effective.
[14] If the solution contains significant amounts of BF3·OEt2, the
evaporation procedure can be repeated as needed until the 19F
NMR shows only product. The concentration of the solution
can be more accurately determined by adding an internal stan-
dard (e.g., trifluorotoluene) and comparing peak integration
using No-D (no-deuterium) 19F NMR spectroscopy. For fur-
ther details, see: T. R. Hoye, B. M. Eklov, M. Voloshin, J. Am.
Chem. Soc. 2004, 126, 2567–2570.
[3] a) C. Palmer, N. A. Morra, A. C. Stevens, B. Bajtos, B. P. Ma-
chin, B. L. Pagenkopf, Org. Lett. 2009, 11, 5614–5617; b) J.
Received: October 8, 2012
Published Online: ᭿
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O21331
/KAP1
Date: 10-12-12 16:52:24
Pages: 6
N. A. Morra, B. L. Pagenkopf
FULL PAPER
Studies Toward Amphidinolides
N. A. Morra, B. L. Pagenkopf* ......... 1–6
BF2OBn·OEt2: A Lewis Acid, Its Use in a
Regio- and Stereoselective Opening of Tri-
substituted Epoxides, and Its Application
towards Amphidinolide C and F
The new Lewis acid BF2OBn·OEt2 was
generated by an anionic redistribution be-
tween BF3·OEt2 and benzyloxytrimethyl-
silane and used in a regio- and stereoselec-
tive epoxide ring-opening reaction. The
utility of this reaction was demonstrated by
the conversion of commercially available
2,3-dihydrofuran into the C-1–C-9 frag-
ment of amphidinolide C and F in five
steps and 49% overall yield.
Keywords: Natural products / Asymmetric
catalysis / Regioselectivity / Cyclization /
Lewis acids / Epoxidation
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0