2
F. Xie et al. / Tetrahedron Letters xxx (2016) xxx–xxx
Table 1
Screening of the reaction conditions.
OH
H
N
N
Me
NFSI/Sm(OTf)3
solvent, MWI
Me
O
Me
Me
1
2
Entry
Solvent
NFSI loading (mol%)
Lewis acid loading (mol%)
Temp (°C)
Time (min)
Yielda (%)
1
2
3
4
5
6
7
8
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
THF
0
1
0
1
1
1
1
5
5
5
1
1
1
1
100
100
100
100
100
100
100
100
100
100
100
100
100
10
10
10
10
10
10
10
10
10
150
30
30
30
11
30
95
93
77
13
96
80
47
93b
10
80
7
15
25
15
10
5
15
10
5
15
15
15
15
9
10
11
12
13
DCE
toluene
Note:
a
Run on a 0.2 mmol scale and isolated yield.
Run with heating.
b
Increasing of the Sm(OTf)3 loading afforded no dramatic improve-
ment in yield (Table 1, entries 7–9). The reaction could be per-
formed with heating by prolonging reaction times (Table 1, entry
10). In screening solventssuch as tetrahydrofuran (THF), dichlor-
oethane (DCE) and toluene, it was observed that DCE could also
act as an effective solvent, while THF and toluene were not the
optimal choices (Table 1, entries 11–13).
When the combinations of those Lewis acids with NFSI were
employed, acceptable yields were achieved. The blank experiments
by using metal chloride afforded only trace amount of products or
no products at all (Table 3, entries 5–11). These results demon-
strated that in the assistance of Lewis acids, even those with lower
catalytic activities the NFSI has much better activity than NFSI
alone in catalyzing Beckmann rearrangement.
With the optimized conditions in hand, a library of oxime com-
pounds was synthesized with structural and electronic modifica-
tions to examine the scope of the NFSI/Sm(OTf)3 system in
catalyzing Beckmann rearrangement. 1-(Aryl)ethanone oximes
1a–1e were prepared firstly and subjected to the optimal condi-
tions. All the tested substrates gave the N-(aryl)acetamide prod-
ucts 2a–2e in moderate to excellent yields regardless of the aryl
rings bearing electron-rich or electron-deficient substitutions
(Table 2, entries 1–5). The symmetrical bis(aryl)methanone oxime
substrates 1f–1i were then examined under the standard condi-
tions and corresponding N-(aryl)benzamides products 2f–2i were
obtained in excellent yields (Table 2, entries 6–9). The substrates
of unsymmetrical bis(aryl)methanone oximes 1j–1k gave the elec-
tron-rich phenyl ring migration N-(aryl)benzamide products 2j–2k
as the major products, accompanied with about 30% electron-defi-
cient phenyl ring migration N-(aryl)benzamide products (Table 2,
entries 10–11).
It is worth mentioning that the reaction could also proceed
smoothly under neat conditions without losing any efficiency
(Scheme 1). And the reaction temperature could be decreased to
80 °C or 60 °C by prolonging reaction time accompanied by slightly
lower yield. The reaction could be performed on a gram scale by
heating for relatively longer reaction time.
In order to gain insights into the new catalytic system, we
turned towards the evaluation of some available Lewis acid with
the combination of NFSI in catalyzing Beckmann rearrangement.
Firstly, several rare earth metal triflates were investigated. It was
shown that they could catalyze the transformation in about 10%
yield in the absence of NFSI, consistent with the result of samar-
ium(III) triflate without addition of NFSI. However, when a mixture
of 15 mol% NFSI and 1 mol% rare earth metal triflates was used, the
yield was improved significantly (Table 3, entries 1–4). Some non-
active Lewis acids with chloride as couterions were also examined.
To identify the real catalytic species, several control experi-
ments were performed. The possibility of the formation of more
active Sm[N(SO2Ph)2]3 by counterion exchange was excluded
14
by synthesizing the Sm[N(SO2Ph)2]3 using literature reported
method.15 The freshly prepared catalyst showed uncomparable
activity with the combination of NFSI and Sm(OTf)3 (Table 4, entry
1). In contrast, the combination of NFSI and Sm[N(SO2Ph)2]3 pro-
vided reasonable yield (Table 4, entry 2). Additionally, to study
the influence of electrophilic fluorine on the catalysis, N-methyl-
benzenesulfonimide (NMSI) was prepared. The combination of
NMSI and Sm(OTf)3 showed much weaker catalytic activity than
NFSI/Sm(OTf)3 and the combination of NMSI and CaCl2 showed
no activity at all, which indicated that the fluorine played an essen-
tial role in catalyzing the reactions (Table 4, entries 3–4). To shed
lights on the catalytic effects of the electrophilic fluorine, two com-
mercially available electrophilic fluorine reagents were examined.
As expected, they were able to catalyze the Beckmann rearrange-
ment in 73% and 83% yields respectively (Table 4, entries 5–6).
However, the combination of Sm(OTf)3 and Selectflour afforded
similar yields to Selectfluro itself, which indicated that the elec-
trophilicity of the fluorine of Selectfluor was not further increased
by Sm(OTf)3. It also indicated that the Lewis acids significantly
increased the electrophilicity of the fluorine of NFSI to the degree
that is comparable to that of the Selectflour fluorine (Table 4, entry
7).2d,16
From those above findings, a mechanistic proposal on the inter-
action between NFSI and Lewis acids was presented in Scheme 2.
The bidental coordination of Lewis acids with the sulfonyl groups
on NFSI in a six-membered ring pattern17 would decrease the
electron density on nitrogen of the counteranion, which would
enhance the electrophilicity of fluorine by redistributing the elec-
tron density. In order to determine the electronic nature of the
fluorine, DFT calculations were performed and the atomic charges