Ring size and substituent steric effects in cyclic ketene-N,O/S-acetal trifluoroacetylations

Article abstract of DOI:10.1016/j.tetlet.2012.03.068

Trifluoroacetylations of cyclic ketene-N,O/S-acetals exhibit a ring size effect. The five-membered rings 2,4,4-trimethyl-2-oxazoline, 2-methyl-2-oxazoline, and 2-methyl-2-thiazoline, each form cyclic ketene-N,O/S-acetal intermediates which then react with trifluoroacetic anhydride to give β-monotrifluoroacetylation products. Conversely, the six-membered ring 2-methyl-2-oxazine gives its β,β- bistrifluoroacetylation product. In situ-generated five-membered ring N-Me cyclic ketene-N,O-acetals, 3,4,4-trimethyl-2-methylene-oxazolidine, and 3-methyl-2-methylene-oxazolidine, and six-membered ring 3-methyl-2-methylene-1, 3-oxazinane were each β,β-bistrifluoroacetylated. However, 3-methyl-2-methylene-oxazolidine also afforded a γ-lactam by an iodide-catalyzed rearrangement of its β,β-bistrifluoroacetylated derivative. In situ-generated 3-methyl-2-methylene-thiazolidine gives both β-mono- and β,β-bistrifluoroacetylation products and no lactam, in contrast to its N,O-analog 3-methyl-2-methyleneoxazolidine.

Full text of DOI:10.1016/j.tetlet.2012.03.068

Tetrahedron Letters 53 (2012) 2965–2970
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Ring size and substituent steric effects in cyclic ketene-N,O/S-acetal
trifluoroacetylations
⇑
⇑
Hondamuni I. De Silva, Yingquan Song , William P. Henry, Charles U. Pittman Jr.
Department of Chemistry, Mississippi State University, Mississippi State, MS 39762, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Trifluoroacetylations of cyclic ketene-N,O/S-acetals exhibit a ring size effect. The five-membered rings 2,4,4-
trimethyl-2-oxazoline, 2-methyl-2-oxazoline, and 2-methyl-2-thiazoline, each form cyclic ketene-N,O/S-
acetal intermediates which then react with trifluoroacetic anhydride to give b-monotrifluoroacetylation
products. Conversely, the six-membered ring 2-methyl-2-oxazine gives its b,b-bistrifluoroacetylation prod-
uct. In situ-generated five-membered ring N-Me cyclic ketene-N,O-acetals, 3,4,4-trimethyl-2-methylene-
oxazolidine, and 3-methyl-2-methylene-oxazolidine, and six-membered ring 3-methyl-2-methylene-1,
3-oxazinane were each b,b-bistrifluoroacetylated. However, 3-methyl-2-methylene-oxazolidine also affor-
Received 8 February 2012
Revised 9 March 2012
Accepted 20 March 2012
Available online 28 March 2012
Keywords:
Cyclic ketene acetal
Rearrangement
Ring size effect
Steric effect
ded a c-lactam by an iodide-catalyzed rearrangement of its b,b-bistrifluoroacetylated derivative. In situ-gen-
erated 3-methyl-2-methylene-thiazolidine gives both b-mono- and b,b-bistrifluoroacetylation products and
no lactam, in contrast to its N,O-analog 3-methyl-2-methyleneoxazolidine.
Ó 2012 Elsevier Ltd. All rights reserved.
Push–pull alkene
Aroylations of 2-methyl-1,3-oxazolines and thiazolines, and 2-
methyl-1,3-oxazines in the presence of triethylamine were shown
to proceed via in situ generated cyclic ketene-N,O/S-acetals and
generated a variety of interesting products.1 Cyclic ketene-N,O/S-
acetals (Scheme 1) are extremely nucleophilic combining the reac-
tivity of enamines and vinyl ethers or vinyl thioethers.2 Pittman’s
group has extensively explored chemistry of five- and six-mem-
bered ring cyclic ketene-N,O/S-acetals by reacting them with
mono-electrophiles,1b,3 such as acid chlorides and isocyanates,
and di-electrophiles,3a,4 such as 1,3-diacid chlorides, N-chlorocar-
bonyl isocyanate, and N-chlorosulfonyl isocyanate.
2,4,4-Trimethyl-2-oxazoline 1 and 2-methyl-2-oxazoline 2, for
example, react differently with benzoyl chloride to give the ring-
retained bis-N,C-benzoylated compound 3 and the ring-opened
N,C,O-trisbenzoylated compound 4, respectively (Eqs. 1 and 2).1b,c
Compounds 3 and 4 are precursors to the ambident nucleophiles,
cyclic ketene-N,O-acetals 5 and 6, when debenzoylated by KOH
in CH3OH.1a,c Benzoylation of 2-methyl-2-thiazoline 7 (Eq. 3) fol-
lows the same reaction path as 1, rather than its N,O-analog 2, to
give the ring-retained bis-N,C-benzoylated compound 8.1b
E
X
E
X
β
E
R
R
R
N
X
-H
N
N
X = O, S
Scheme 1. The nucleophilicity of a cyclic ketene-N,O/S-acetal.
O
Ph
H
N
H
O
Ph
O
O
1) KOH,
PhCOCl
MeOH, rt
ð1Þ
ð2Þ
ð3Þ
N
O
Ph
O
HN
2) H+
Et3N, CH3CN,
REFLUX
1
3
5
O
Ph
H
O
Ph
H
O
O
1) KOH,
MeOH, rt
PhCOCl
N
O
Ph
O
HN
2) H+
Et3N, CH3CN,
REFLUX
Ph
N
2
O
6
4
Cl
O
S
H
N
Ph
O
PhCOCl
N
S
Ph
Et3N, CH3CN,
REFLUX
⇑
Corresponding author. Tel.: +1 662 325 7616; fax: +1 662 325 1618.
E-mail addresses: yingquan@gmail.com (Y. Song), cpittman@chemistry.mssta-
8
7
te.edu (C.U. Pittman).
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2012.03.068

2966
H. I. De Silva et al. / Tetrahedron Letters 53 (2012) 2965–2970
O
O
b-C-Benzoylated cyclic ketene-N,O/S-acetal 9 is the N-methyl-
CF3
ated analog of 5 (Eq. 4) and 10 is the N-methylated S-analog of 6
(Eq. 5). They were synthesized starting from their corresponding
cyclic ketene-N,O/S-acetals 11 and 12, respectively. Compounds
11 and 12, in turn, were synthesized from their iodide salts 13
and 14, respectively.1b
CF3
(CF3CO)2O
Et3N
O
O
N
O
ð6Þ
N
HN
O
F3C
THF, REFLUX
1
16
15
49%
not observed
O
β
CF3
PhCOCl
Et3N
Ph
I
NaH
THF
N
O
S
ð4Þ
ð5Þ
N
O
(CF3CO)2O
Et3N
O
N
O
THF, 0 o
C
N
O
O
THF, REFLUX
HN
O
11
13
14
9
CF3
CF3
O
2
17
31%
O
ð7Þ
ð8Þ
O
S
O
CF3
β
O
PhCOCl
Et3N
Ph
I
N
OH
NaH
THF
N
N
O
F3C
N
S
F3C
N
THF, 0 oC
or
18
19
OCOCF3
not formed
not observed
10
12
Trifluoroacetic anhydride is a stronger mono-electrophile than
CF3
O
benzoyl chloride. Trifluoroacetylated analogs of b-C-benzoylated
cyclic ketene-N,O/S-acetals 5, 6, 9 and 10 constitute strong push–
pull alkene systems derived from cyclic ketene-N,O/S-acetals. Con-
jugated push–pull alkenes are useful in nonlinear optical materi-
als.5 The above-mentioned chemistry was explored herein for the
synthesis of b-mono- or b,b-bis-trifluoroacetylated conjugated sys-
tems. This generated a variety of products.
O
(CF3CO)2O
Et3N
CF3
O
N
S
HN
S
N
S
THF, REFLUX
F3C
7
20
21
not observed
44%
Trifluoroacetylation of the six-membered ring 2-methyl-2-oxa-
All reactions reported in this Letter used 2.2 to 2.6 equiv of tri-
fluoroacetic anhydride per 1.0 equiv of the heterocyclic reagent.
zine 22 (Eq. 9), in contrast to the five-membered ring analogs 1, 2
and 7, gave the b,b-bistrifluoroacetylation product 23 (Fig. 2)
rather than the b-monotrifluoroacetylation product 24. The unusu-
ally long C@C bond, 1.429(1) Å of 23 (vs 1.399(3) Å of 15), indicates
that it is more highly polarized and has reduced double bond char-
acter, allowing either fast rotation or libration of a bisected struc-
ture at the central bond. Fast rotation best explains why the two
COCF3 functions of 23 appear equivalent in its 13C NMR spectrum.
Only a 5.6(2)° (N1–C4–C3–C8) out-of-plane twist of the double
bond of 23 was present in the solid state. N–H to oxygen hydro-
gen-bonds exist in both 15 (2.289 Å) and 23 (2.000 Å).
Results
Trifluoroacetylations of 1, 2, 7 and 22
Trifluoroacetylations of 2,4,4-trimethyl-2-oxazoline 1, 2-
methyl-2-oxazoline 2 and 2-methyl-2-thiazoline 7 (Eqs. 6–8) gave
only the b-monotrifluoroacetylation products 15 (Fig. 1), 17 and
20, respectively, instead of the N,C-bistrifluoroacetylation products
16, 18 and 21, respectively. The ring-opened N,C,O-tristrifluoro-
acetylated compound 19 was also not formed when 2 reacted with
trifluoroacetic anhydride (Eq. 7), unlike its reaction with benzoyl
chloride which formed 4 (Eq. 2). These differences are intriguing.
O
O
O
(CF3CO)2O
Et3N
F3C
F3C
CF3
N
O
HN
O
HN
O
ð9Þ
THF, REFLUX
22
23
24
not observed
64%
Figure 1. The crystal structure of 3-(4,4-dimethyloxazolidin-2-ylidene)-1,1,1-
trifluoropropan-2-one 15. CCDC: 796901.
Figure 2. The crystal structure of 1,1,1,5,5,5-hexafluoro-3-[1,3] oxazinan-2-
ylidene-pentane-2,4-dione 23. CCDC: 796902.

H. I. De Silva et al. / Tetrahedron Letters 53 (2012) 2965–2970
2967
Figure 3. (a) The crystal structure of 1,1,1,5,5,5-hexafluoro-3-(3-methyl-oxazolidin-2-ylidene)-pentane-2,4-dione 30. (b) Torsional angle N1–C6–C3–C4 of 30. Only C3–C6
bond and atoms attached are shown. CCDC: 864636.
The crystal structures of 30 and 31 are shown in Figures 3 and 4,
respectively.
O
O
β
I
F3C
(CF3CO)2O
Et3N
CF3
N
O
N
O
ð10Þ
ð11Þ
N
O
THF, REFLUX
13
11
29
88%
β
(CF3CO)2O
Et3N
I
N
O
N
O
THF, REFLUX
27
25
O
N
O
Figure 4. The crystal structure of 1-methyl-3-(2,2,2-trifluoro-1,1-dihydroxy-
ethyl)-pyrrolidin-2-one 31. CCDC: 864637.
O
OH
F3C
CF3
+
CF3
OH
N
O
Trifluoroacetylations of five/six-membered cyclic ketene-N,O/S-
31
23%
30
24%
acetals generated in situ
Trifluoroacetylations of cyclic ketene-N,O/S-acetals 11, 12, 25
and 26, generated in situ by deprotonation of their corresponding
iodide salts 13, 14, 27 and 28 by Et3N, respectively, were also at-
tempted (Eqs. 10–13). Et3N was the base used previously for
cyclic ketene-N,O-acetal generation.3h 3,4,4-Trimethyl-2-methy-
lene-oxazolidine 11 gave b,b-bistrifluoroacetylation product 29
(Eq. 10). However, 3-methyl-2-methylene-oxazolidine 25 behaved
differently than 11 under the same reaction conditions. In addition
to b,b-bistrifluoroacetylation product 30, lactam 31 was also iso-
lated (Eq. 11). 25 forms only 30, which rearranges in the presence
of iodide to 31. After 5 h, only 30 and 31 are seen, but 30 is progres-
sively converted to 31 at longer times. 31 is observed as the only
product from 25 at long reaction times, but isolated yields are
low due to serious losses during an aqueous bicarbonate wash used
in the work up. 31 is acidic and water soluble. Trifluoroacetylation
of 3-methyl-2-methylene-thiazolidine 12 gave both b-monotriflu-
oroacetylated product 32 and b,b-bistrifluoroacetylated product
33 (Eq. 12). Six-membered ring, 3-methyl-2-methylene-1,3-oxaz-
inane 26, exclusively gave b,b-bistrifluoroacetylated product 34
(Eq. 13), in contrast to its five-membered ring analog 25 (Eq. 11).
β
(CF3CO)2O
Et3N
I
N
S
N
S
THF, REFLUX
14
12
O
N
O
O
S
ð12Þ
F3C
F3C
CF3
+
S
N
32
33
60%
10%
O
O
O
β
(CF3CO)2O
Et3N
I
F3C
CF3
N
O
N
O
THF, REFLUX
N
ð13Þ
26
28
34
76%

2968
H. I. De Silva et al. / Tetrahedron Letters 53 (2012) 2965–2970
Only one 13C NMR resonance was observed for the two CF3
O
O
O
groups and two C@O carbons in all b,b-bistrifluoroacetylation
products 29, 30, 33 and 34. This indicates that the exocyclic C@C
bonds of 29, 30, 33 and 34 are highly polarized and have signifi-
cantly reduced double bond character allowing either fast rotation
or bisected structure libration at the exocyclic C@C bond. Elonga-
tion of exocyclic C@C bond (1.443(2) Å) and a 51.5(2)° (N1–C6–
C3–C4) out-of-plane twist in 30 (Fig. 3) confirmed its reduced bond
order, and a large degree of charge separation. This double bond
twist results from its reduced bond order, the steric repulsion be-
tween b-carbon substituents with N-methyl group and its stable
zwitterionic nature. We earlier reported the synthesis of twisted
push–pull alkenes from N,N-dimethyl cyclic ketene-N,N-acetals.3g
Lactam 31 has an OH to oxygen hydrogen-bond (1.99 Å) in the so-
lid state (Fig. 4) and its 1H NMR spectrum exhibits two distinct OH
protons (Supplementary data).
O
CF3
+
O
CF3
F3C
N
O
N
O
F3C
N
O
+
N
2
18
38
37a
(CF3CO)2O
Et3N
Scheme 3. An alternative N-detrifluoroacetylation mechanism for 18 by 2-
methyl-2-oxazoline 2.
O
O
O
O
H
H
N
O
CF3
(CF3CO)2O
O
F3C
CF3
(CF3CO)2O
O
N
O
F3C
N
F3C
Et3N, THF
REFLUX
Et3N, THF
REFLUX
22
39
40a
Discussion
H
O
O
O
O
O
CF3
O
F3C
O
F3C
CF3
Trifluoroacetylations of 2,4,4-trimethyl-2-oxazoline 1, 2-
methyl-2-oxazoline 2 and 2-methyl-2-thiazoline 7 involve the ini-
tial formation of N,C-bistrifluoroacetylation intermediates 16, 18
and 21, respectively (Eqs. 6–8). This also occurs in their benzoyla-
tions. However, 16, 18 and 21 are never observed because the
labile N-trifluoroacetyl function is lost. Scheme 2 illustrates the
mechanism using 18 as an example.1b,c The strongly electrophilic
trifluoroacetyl moiety on the ring nitrogen of 16, 18 and 21
is highly susceptible to nucleophilic attack. The trifluoroacetyl
function on the b-carbons of 16, 18 and 21 helps to stabilize the
incipient conjugated anion (37a–37c in Scheme 2) formed upon
the loss of the N-trifluoroacetyl function. The nucleophile respon-
sible for this N-detrifluoroacetylation to ambident anion 37 could
be the trifluoroacetate anion, or, more likely, water, which is used
in the aqueous work up. This detrifluoroacetylation pathway finds
analogy in methanol N-debenzoylation of the b-phenyl N,C-
bisbenzoylated cyclic ketene-N,O-acetal.1a It is also possible that
2 itself acts as a nucleophile to cause N-detrifluoroacetylation of
18 to form 37a and 38 (Scheme 3).6
F3C
O
F3C
CF3
- H
O
CF3
N-detrifluoro-
acetylation
H2O
N
N
O
HN
O
40b
41
23
Scheme 4. The six-membered ring N,C-bistrifluoroacetylated cyclic ketene-N,O-
acetal 39 undergoes a third trifluoroacetylation at its b-carbon during formation of
23.
whereas its five-membered ring analog resisted b-C-acylation.7
Likewise, we suggest that the six-membered N,C-bistrifluoroacety-
lation intermediate 39 undergoes a third trifluoroacetylation at its
b-carbon, forming intermediate 40a (Scheme 4). Subsequent depro-
tonation of either 40a, or its enol form 40b, gives 41. Highly reactive
41 is easily N-detrifluoroacetylated by water during the work up to
form 23. By contrast, the exocyclic C@C bond of the five-membered
ring N,C-bistrifluoroacetylationproduct 18 (Scheme2) resists a third
trifluoroacetylation.
Trifluoroacetylation of 2 did not give the ring-opened N,C,O-tri-
strifluoroacetylated compound 19 (Eq. 7). This is in contrast to its
benzoylation to 4 (Eq. 2).1c Presumably, the b-keto oxygen of the
N,C-bistrifluoroacetylation product 18 is less nucleophilic
(Scheme 5) than its corresponding N,C-bisbenzoylation analog,
due to the strong electron withdrawal from ring nitrogen by COCF3
The b,b-bistrifluoroacetylation of 2-methyl-2-oxazine 22 versus
the b-monotrifluoroacetylations of 2,4,4-trimethyl-2-oxazoline 1
and 2-methyl-2-oxazoline 2 highlights the b-carbon’s nucleophilic-
ity difference between six-membered cyclic ketene-N,O-acetals and
their five-membered analogs. An N-acylated six-membered ring
enaminoester intermediate was reported to undergo b-C-acylation,
O
O
O
(CF3CO)2O
Et3N
H
N
(CF3CO)2O
Et3N
O
CF3
O
O
O
CF3
CF3
N
O
F3C
N
O
O
CF
(CF3CO)2O
Et3N
3 (CF3CO)2O
Et3N
F3C
O
THF, REFLUX
O
THF, REFLUX
2
35
N
O
H2O
F3C
N
O
THF
REFLUX
THF
REFLUX
18
F3C
N
2
O
O
X
18
42
H
H
O
CF3
CF3
CF3
CF3
O
O
O
F3C
H2O
O
Less nucleophilic than
its benzoyl analog.
O-acylation doesn't occur
F3C
O
OH2
N
N
O
O
O
CF3
36
37a
O
CF3
CF3
O
F3C
N
O
H
N
H
O
H
N
O
CF3
O
CF3
18
+
F3C
H
N
OH
O
HN
O
37b
19
37c
17
OCOCF3
Scheme 2. N-Detrifluoroacetylation of 18 by water.
Scheme 5. Why the trisacylated, ring-opened compound 19 cannot be obtained.

H. I. De Silva et al. / Tetrahedron Letters 53 (2012) 2965–2970
2969
function. Therefore, O-trifluoroacetylation and subsequent ring
opening by attack of the counter ion, CF3COOꢀ, do not occur. If O-
trifluoroacetylation did occur to give 42, then CF3COOꢀ is too weak
a nucleophile to attack C-5 of 42 to give the ring opening (Scheme 5).
Reactions of the iodide salts 13, 27, 14 and 28 (Eqs. 10–13) with
(CF3CO)2O and Et3N, are similar to reported reactions of (CF3CO)2O
with N-methylated iodide salt of 2-methylbenzothiazole, in the
presence of pyridine.8 Eqs. 10–13 involve in situ generation of cyc-
lic ketene-N,O/S-acetals 11, 25, 12 and 26, respectively. b,b-Bistri-
fluoroacetylations of 11, 25, 12 and 26 occur when excess
(CF3CO)2O is present versus the cyclic ketene-N,O/S-acetal being
generated in situ. Therefore, these reactions did not stop at the
monotrifluoroacetylation step. These b,b-bistrifluoroacetylations
find analogy in our previous work on b,b-bis(N-arylamido) cyclic
ketene-N,O-acetal synthesis.3h The reaction of 12 with (CF3CO)2O
(Eq. 12) also gave b-monotrifluoroacetylation product 32, because
the b-carbon of 32 is less nucleophilic than its oxygen analog. This
may be due to the lower electron donating capacity of sulfur versus
oxygen which slows further acylation of 32 to 33.
31 is ultimately formed. However, a rearrangement from 49 to 31,
via 50 and 48, cannot be completely ruled out. The NMR spectrum
of 31 in CDCl3 shows 16–17% of enol 47 present, but this enol is not
the form observed in the crystalline solid state.
Based on this mechanism, an iodide-catalyzed transformation
of 30 to 31 was independently conducted (Eq. 14). This reaction
generated 31 on work up. Compound 30 was completely converted
to 31 based on TLC, 1H and 13C NMR analyses. However, 31 is acidic
and water soluble. Hence, an aqueous sodium bicarbonate washing
during the work up removed 31 and led to the low isolated yield
(37%) shown in Eq. 14.
CF3
O
O
OH
CF3
O
CF3
OH
Bu4NI
N
ð14Þ
N
O
THF, REFLUX
20h
31
30
37%
Attempts to detrifluoroacetylate 34 as a synthetic route to b-
The formation of lactam 31 (Eq. 11) finds analogy in the earlier
iodide-catalyzed rearrangement of b,b-bis(N-arylamido) cyclic ke-
tene-N,O-acetals.3h A plausible mechanism for the formation of
31 is proposed (Scheme 6). The b,b-disubstituted cyclic ketene-
N,O-acetal 30 undergoes a nucleophilic attack by iodide at C5,
forming ring-opened enolate intermediate 43. Conversely, the
two methyl groups at C4 of compound 29 (Eq. 10) sterically retard
attack by iodide at C5.3h The ambident nucleophile 43 undergoes
an intramolecular SN2 attack at C5 forming lactam 44. Subsequent
detrifluoroacetylation of 44 by water during aqueous work up
gives 31 via 45–48. A COCF3 moiety of 44 is more susceptible to at-
tack by water compared to that of monotrifluoroacetylation inter-
mediate 49. This is due to the lack of conjugation with an exocyclic
C@C in 44 as is present in 49. Therefore, the hydrated keto product
monotrifluoroacetylated cyclic ketene-N,O-acetal 51, using propyl-
amine, methanolic KOH, or phenylhydrazine, were unsuccessful
(Eq. 15). This reaction does not occur because cleavage of carbonyl
carbon to b-carbon bond would initially have to form the high en-
ergy vinyl anion, 52.
O
O
O
O
..
F3C
CF3
CF3
CF3
Nu
N
O
N
O
N
O
Nu = PrNH2,
methanolic KOH
or PhNHNH2
34
51
52
ð15Þ
In summary, a ring size effect has been found in trifluoroacety-
lations of 2,4,4-trimethyl-2-oxazoline 1, 2-methyl-2-oxazoline 2,
2-methyl-2-thiazoline 7 and 2-methyl-2-oxazine 22. b,b-Bistriflu-
oroacetylated N-methyl cyclic ketene-N,O/S-acetals were synthe-
sized via trifluoroacetylation of in situ generated N-methyl cyclic
ketene-N,O/S-acetals. An iodide-catalyzed rearrangement of b,b-
bistrifluoroacetylated five-membered cyclic ketene-N,O-acetal 30
to its lactam 31 was discovered.
O
O
N
O
O
H
N
CF3
F3C
CF3
(CF3CO)2O
Et3N
I
N
O
O
THF, REFLUX
27
49
30
I
I
Acknowledgments
O
O
The authors acknowledge the educational and general funds of
Mississippi State University for partial financial support of this
work. We would like to thank Hua Guo for helping obtain NMR
spectra. Guozhong Ye and Sabornie Chatterjee helped to interpret
NMR spectra and grow single crystals for X-ray crystallography.
O
H
N
OH
F3C
CF3
O
31
O
CF3
CF3
OH
O
O
N
N
OH
H
43
50
I
I
Supplementary data
H2O
Supplementary data (complete experimental details and the
characterizations of all the compounds) associated with this article
can be found, in the online version, at http://dx.doi.org/10.1016/
j.tetlet.2012.03.068.
O
O
CF3
N
N
CF3
CF3
H
44
O
48
References and notes
H2O
1. (a) Tohda, Y.; Kawashima, T.; Ariga, M.; Akiyama, R.; Shudoh, H.; Mori, Y. Bull.
Chem. Soc. Jpn. 1984, 57, 2329; (b) Zhou, A., Ph.D. dissertation, 2004, Mississippi
State University.; (c) Song, Y.; De Silva, H. I.; Henry, W. P.; Ye, G.; Chatterjee, S.;
Pittman, C. U., Jr. Tetrahedron Lett. 2011, 52, 4507.
2. (a) McElvain, S. M.; Curry, M. J. J. Am. Chem. Soc. 1948, 70, 3781; (b) McElvain, S.
M. Chem. Rev. 1949, 45, 453. and reference cited therein.
3. (a) Zhou, A.; Pittman, C. U., Jr. Synthesis 2006, 37; (b) Zhou, A.; Pittman, C. U., Jr.
Tetrahedron Lett. 2004, 45, 8899; (c) Zhou, A.; Cao, L.; Li, H.; Liu, Z.; Pittman, C. U.,
Jr. Synlett. 2006, 201; (d) Zhou, A.; Cao, L.; Li, H.; Liu, Z.; Cho, H.; Henry, W. P.;
O
O
O
O
O
OH
+ H
CF3
N
OH2
CF3
CF3
N
CF3
N
45
46
O
47
Scheme 6. Two possible routes for formation of 31 from 30 or 49.

2970
H. I. De Silva et al. / Tetrahedron Letters 53 (2012) 2965–2970
Pittman, C. U. Tetrahedron 2006, 62, 4188; (e) Ye, G.; Chen, C.; Chatterjee, S.;
5. (a) Williams, D. J. Angew. Chem., Int. Ed. Engl. 1984, 23, 690; (b) Kanis, D. R.;
Ratner, M. A.; Marks, T. J. Chem. Rev. 1994, 94, 195.
6. Methyl-2-oxazoline is a stronger nucleophile than iodide, see: (a) Saegusa, T.;
Ikeda, H. Macromolecules 1973, 6, 808; (b) Hrkach, J. S.; Matyjaszewski, K.
Macromolecules 1992, 25, 2070.
Collier, W. E.; Zhou, A.; Song, Y.; Beard, D. J.; Pittman, C. U., Jr. Synthesis 2010,
141; (f) Ye, G.; Chatterjee, S.; Zhou, A.; Barker, B. L.; Chen, C.; Song, Y.; Pittman, C.
U., Jr. Synthesis 2010, 1209; (g) Ye, G.; Henry, W. P.; Chen, C.; Zhou, A.; Pittman,
C. U., Jr. Tetrahedron Lett. 2009, 50, 2135; (h) Song, Y.; Henry, W. P.; De Silva, H. I.;
Ye, G.; Pittman, C. U., Jr. Tetrahedron Lett. 2011, 52, 853; (i) Chatterjee, S.; Ye, G.;
Song, Y.; Barker, B. L.; Pittman, C. U., Jr. Synthesis 2010, 19, 3384.
7. Brunerie, P.; Célérier, J. P.; Petit, H.; Lhommet, G. J. Heterocycl. Chem. 1986, 23,
1183.
4. (a) Zhou, A.; Pittman, C. U., Jr. Tetrahedron Lett. 2005, 46, 2045; (b) Ye, G.; Zhou,
A.; Henry, W. P.; Song, Y.; Chatterjee, S.; Beard, D. J.; Pittman, C. U., Jr. J. Org.
Chem. 2008, 73, 5170; (c) Chatterjee, S.; Ye, G.; Pittman, C. U., Jr. Tetrahedron Lett.
2010, 51, 1139.
8. Bailey, A. S.; Ellis, J. H.; Harvey, J. M.; Hilton, A. N.; Peach, J. M. J. Chem. Soc. Perkin
Trans. 1 1983, 795.