Reductive deprotection of N-tritylaziridines

Full text of DOI:10.1021/jo0106243

7542
J . Org. Chem. 2001, 66, 7542-7546
Ch a r t 1
Red u ctive Dep r otection of
N-Tr ityla zir id in es
E. Vedejs,* A. Klapars, D. L. Warner, and A. H. Weiss
Department of Chemistry, University of Michigan,
Ann Arbor, Michigan 48109
edved@umich.edu
Received J une 18, 2001
Synthetic projects in our laboratory have encountered
the need to deprotect N-tritylaziridines related to 1, 3,
and 5 (Chart 1). There are several literature reports
describing conventional acid-catalyzed detritylation of
aziridinecarboxylates such as 1b.1,2 However, we could
not deprotect the more sensitive aziridines 3 and 5
without extensive aziridine ring cleavage using formic
acid/methanol1a or trifluoroacetic acid (TFA).1b After
considerable experimentation, good results were obtained
for a number of N-tritylaziridines using TFA together
with a reducing agent capable of trapping trityl cation
at 0 °C in dichloromethane or chloroform (Table 1). In
most of the substrates tested, the combination of TFA
and triethylsilane works well (method A).3,4 However, an
alternative procedure using methanesulfonic acid (MsOH)
in place of TFA is best for the sensitive aziridine 3a
(method B). A third procedure (method C; TFA/Me3N-
BH3)5 is generally faster than method A, is also broadly
applicable, and gives cleaner products in some cases.
Thus, cleavage of 55 using method A afforded the product
65 contaminated with Et3Si signals, apparently resulting
from silyl ether exchange (entry 13). This complication
is not a factor using method C (entry 14), and 6 was
obtained in good yield and purity. In other examples,
method C gave somewhat lower yields under the stan-
dard conditions, probably because TFA reacts rapidly
with Me3N-BH3 to release hydrogen. This competing
pathway deactivates the reagent and results in incom-
plete detritylation in some cases.
Ta ble 1. Dep r otection of N-Tr ityla zir id in esa
entry
substrate
method
time
product (yield, %)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
1a
1a
1b
1b
1c
1c
1c
3a
3a
3a
3b
3b
5
A
B
A
B
A
B
C
A
B
C
A
B
A
C
A
A
C
1 h
2a (85)
2a (<5)b
2b (56)c
2b (43)
2c (82)
2c (85)
2c (61)
13 (37)
4 (63)d
4 (<5)e
4 (19)
30 min
30 min
30 min
30 min
25 min
40 min
30 min
30 min
30 min
30 min
30 min
1 h
4 (40)
6 (86)f
6 (85)
5
7
19
21
5 min
30 min
30 min
5 min
8 (84)
20 (72)
22 (88)
a
All experiments in dichloromethane at 0 °C, 4 equiv each of
the acid and the reducing agent. Method A: TFA/Et3SiH. Method
B: methanesulfonic acid/Et3SiH. Method C: TFA/trimethylamine-
b
borane. Workup with diisopropylethylamine in all cases. The
alcohol from TBS cleavage was isolated (66%) with the trityl group
intact, along with recovered 1a (27%). c The crude product prior
to workup contains NMR signals of protonated 2b and triphenyl-
methane as the sole products. Material loss occurs due to partial
decomposition under the conditions of the standard workup
* To whom correspondence should be addressed. E-mail: edved@
umich.edu.
(1) (a) Mauger, A. B.; Stuart, O. A. Int. J . Pept. Protein Res. 1987,
30, 481. Kuyl-Yeheskiely, E.; Lodder, M.; van der Marel, G. A.; van
Boom, J . H. Tetrahedron Lett. 1992, 33, 3013. (b) Baldwin, J . E.;
Adlington, R. M.; Robinson, N. G. J . Chem. Soc., Chem. Commun. 1987,
153. (c) J ohnson, T. B.; Coward, J . K. J . Org. Chem. 1987, 52, 1771.
(d) Sato, K.; Kozikowski, A. P. Tetrahedron Lett. 1989, 30, 4073. (e)
Baldwin, J . E.; Spivey, A. C.; Schofield, C. J .; Sweeney, J . B.
Tetrahedron 1993, 49, 6309.
(2) Hydrogenolysis conditions have been used to cleave N-(phe-
nylfluorenyl)aziridines: Fernandez-Megia, E.; Montaos, M. A.; Sardina,
F. J . J . Org. Chem. 2000, 65, 6780.
(3) (a) O-Detritylation using silane cation trapping reagents: Pear-
son, D. A.; Blanchette, M.; Baker, M. L.; Guindon, C. A. Tetrahedron
Lett. 1989, 30, 2739. Mehta, A.; J aouhari, R.; Benson, T. J .; Douglas,
K. T. Tetrahedron Lett. 1992, 33, 5441. Ravikumar, V. T.; Krotz, A.
H.; Cole, D. L. Tetrahedron Lett. 1995, 36, 6587. (b) Reactivity issues
in cation trapping with silanes: Mayr, H.; Patz, M. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 938.
d
procedure. TsOH in place of MsOH gave a 73% yield of 4. e Ring
opening to 13 was the major pathway. f Product contaminated with
the Et3SiO analogue, e3% on mg scale, but as much as 20% on 5
g scale.
To obtain good yields of aziridines, it is important to
quench the crude products from methods A-C with a
tertiary amine base. Diisopropylethylamine was used for
this purpose as part of the standard procedure. The
reasons why quenching with base is necessary were found
to be substrate dependent, and were related to the details
of the workup and to the relative ease of aziridine ring
opening as discussed below.
The reasonably stable 1a was deprotected smoothly
using method A followed by workup with diisopropyl-
ethylamine or triethylamine (entry 1, 85% 2a isolated).
However, if the reaction was repeated without the
tertiary amine workup, and was quenched with ice-
water followed by rapid extraction with hexane, then 2a
was isolated in only 13% yield, together with detritylated
(4) Prior work on N-trityl cleavage using TFA/silane reagents:
Dinsmore, C. J .; MacTough, S. C.; Stokker, G. E.; Williams, T. M.
Chem. Abstr. 1998, 128, 140700. Yang, L.; Morriello, G. Tetrahedron
Lett. 1999, 40, 8197. Cleavage of the N-diphenylmethyl group with
hot TFA/EtSiH: Neumann, W. L.; Rogic, M. M.; Dunn, T. J . Tetrahe-
dron Lett. 1991, 32, 5865. Portner, J . R.; Wirschun, W. G.; Kuntz, K.
W.; Snapper, M. L.; Hoveyda, A. H. J . Am. Chem. Soc. 2000, 122, 2657.
(5) Vedejs, E.; Klapars, A.; Naidu, B. N.; Piotrowski, D. W.; Tucci,
F. C. J . Am. Chem. Soc. 2000, 122, 5401.
10.1021/jo0106243 CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/05/2001

Notes
J . Org. Chem., Vol. 66, No. 22, 2001 7543
Sch em e 1
products of aziridine ring cleavage according to NMR
assay. A similar experiment was then conducted in CDCl3
solution in an NMR tube according to method A, but with
no quench. Complete conversion was observed within 20
min at room temperature to the aziridinium salt 10 via
an intermediate presumed to be 9. Characteristic aziri-
dinium C-H protons were observed at 2.85, 2.93, and
3.47 ppm, together with 13C signals at 36.4 and 24.8 ppm
that are consistent with the structure 10. Furthermore,
an identical NMR spectrum was obtained by mixing
deprotected 2a with 3 equiv of TFA in CDCl3. The NMR
signals of 10 did not change over 6 h at room tempera-
ture, indicating that ring opening of the deprotected
aziridine to 11 or isomeric products does not take place
on this time scale. Furthermore, addition of triethylamine
to the solution of 10 obtained from 1a in the NMR
experiment resulted in conversion to 2a in 77% isolated
yield. Since the corresponding experiment quenched with
ice-water and no base gave only 13% 2a , the evidence
indicates that ring opening of 10 is facile in the presence
of water and TFA, but does not occur during the method
A procedure in dichloromethane or chloroform. Thus,
addition of the tertiary amine prevents aziridine ring
cleavage by neutralizing residual TFA, as well as TFA
released by hydrolysis of Et3SiOTFA. The latter byprod-
uct is derived from reaction of triethylsilane with the
trityl trifluoroacetate that results from acid-induced
N-trityl cleavage.
One additional experiment was performed starting
with 1a and TFA in CDCl3, but without triethylsilane.
Within 45 min at room temperature, the aziridine NMR
signals attributed to 9 had disappeared and were re-
placed by complex downfield signals consistent with
nucleophilic cleavage of the aziridine ring. Since the
analogous cleavage is not a major pathway with triethyl-
silane present, the evidence is consistent with a scenario
where the protonated N-tritylaziridine 9 is in equilibrium
with 2a , 10, and trityl trifluoroacetate (eq 1). If a
reducing agent such as Et3SiH is present, the equilibrium
is forced toward 2a and 10 as the trityl trifluoroacetate
is converted to triphenylmethane (eq 3). In the absence
of a hydride donor, nucleophilic cleavage pathways such
as eq 2 become competitive and products of aziridine
cleavage are formed prior to aqueous workup.
(Scheme 1). Structure 13 is the expected product from
initial aziridine ring opening and N-detritylation to 12,
followed by neutralization and trifluoroacetyl migration
from O to N.
In an attempt to prevent ring opening of 3a , the
experiment was repeated with the nonnucleophilic triflic
acid in place of TFA. The reaction did produce some of
the deprotected aziridine 4 (19%), but the tetracyclic
amine 15 was formed as the major product (58%).
Evidently, the N-tritylaziridinium triflate salt ring opens
faster than the trityl group is cleaved in this sensitive
system. The resulting benzylic carbocation 14 then
undergoes intramolecular Friedel-Crafts cyclization, as
in the closely analogous precedent from 16 to 17.7 The
structure of 15 was deduced by NMR comparisons with
17, and by the presence of 20 signals corresponding to
aromatic carbons in the 13C NMR spectrum.
The Friedel-Crafts cyclization could be minimized by
deprotecting 3a using method B (methanesulfonic acid/
triethylsilane). This procedure gave aziridine 4 in 63%
yield after the usual quenching with diisopropylethyl-
amine. Although this finding implies a simple acid-
induced detritylation, NMR monitoring of the reaction
revealed a more complicated sequence of events. Addition
of MsOH to a solution of 3a in CDCl3 containing Et3SiH
gave a bright yellow solution, suggesting the release of
trityl cation. The color faded within 5 min at room
temperature, and the NMR signals of triphenylmethane
appeared. The aziridine C-H signals at 2.81 and 2.5 ppm
disappeared, and new signals were observed at 6.42 and
6.13 ppm (0.5H each), consistent with the formation of a
diastereomer mixture of the ring-opened mesylate 18 as
the ammonium salt. When diisopropylethylamine was
added to this solution, the signals of 18 were replaced
by signals of the aziridine 4. In contrast to the analogous
intermediate 12 in the TFA experiment, the ring-opened
The implications of aziridine cleavage became more
clear when the deprotection of 3a 6 was investigated. In
contrast to the reactions of 1a or 1b, 3a gave no
deprotected aziridine product 4 using method A. Instead,
a ring-opened product, 13, was obtained in 37% yield (one
diastereomer isolated, stereochemistry not assigned)
(7) Hagishita, S.; Shiro, M.; Kuriyama, K. J . Chem. Soc., Perkin
Trans. 1 1984, 1655.
(6) Chiu, I.-C.; Kohn, H. J . Org. Chem. 1983, 48, 2857.

7544 J . Org. Chem., Vol. 66, No. 22, 2001
Notes
18 undergoes reclosure to 4 upon neutralization with the
tertiary amine, probably because the methanesulfonyl
group is slow to undergo the internal O to N migration
compared to trifluoroacetate.
ring opening since the final product after tertiary amine
quenching is still the same deprotected aziridine. This
is because workup with a tertiary amine induces reclo-
sure of the aminomesylate intermediates in the event
that the ring is cleaved. Indeed, we did not test all of the
substrates in Table 1 to prove whether the deprotection
occurs with aziridine ring cleavage. However, the efficient
detritylations using method A in entries 13 and 15
indicate that aziridine ring opening is not the dominant
reaction pathway. Otherwise, these experiments should
have resulted in the rearranged N-trifluoroacetyl alcohols
as major products, as in the reactions of 3a in the
presence of TFA. By this criterion, the detritylation of
1c using method A (entry 5) must also occur without
aziridine ring opening even though structure 1c re-
sembles 3 in substitution. Of course, the geometric
constraints that enforce orbital overlap between the
benzylic C-N bond and the aromatic π-system in 3 are
not present in 1c, so the direct detritylation competes
effectively with ring cleavage. The absence of cis/trans
isomerization during deprotection of 1c, 5, and 7 (NMR
assay) also argues against ring opening.
The reducing agent (triethylsilane or triethylamine-
borane) does not appear to be involved in the process that
breaks the N-trityl bond, but it plays an important role
by trapping the trityl cation. This prevents reversal of
the trityl cleavage step, and drives the equilibrium to
completion. The method has been applied to other sensi-
tive substrates in our laboratory, including examples in
the aziridinomitosene series.5 These results will be
reported later in the context of total synthesis.
The more labile N-monomethoxytrityl (N-mmt) sub-
strate 3b could be deprotected in low yield using method
A (entry 11). Method B gave a small improvement in this
case (entry 12, 40%). A detailed optimization was not
carried out because we found it more difficult to prepare
and purify 3b due to its tendency to undergo deprotection
and decomposition on silica gel. A prior report has
described similar difficulties with an N-mmt-aziridine,
although the protecting group survived an impressive
range of experimental procedures.8
Since detritylation of 3a using method B appears to
involve a ring-cleaved intermediate rather than the
starting aziridine, we briefly examined the deprotection
of two simple N-tritylamines 19 and 21 (entries 17 and
18) under the reductive detritylation conditions to dem-
onstrate generality. As expected, both 195 and 219 were
converted easily into the parent amines 20 and 22,
respectively. In the absence of potential complications
due to aziridine ring opening, the faster procedure using
method C (entry 17) is probably best since the short
reaction time minimizes potential problems with cleavage
of other protecting groups for these amines.10
Exp er im en ta l Section
N-Tritylaziridines 1b1d, 5,5 and 75 and N-tritylamines 195 and
219 were prepared according to procedures described in the
literature. The deprotected aziridines 2b,1d 2c,11 and 46 and the
amines 205 and 229a were identified by comparisons of spectro-
scopic data.
P r ep a r a tion of Azir id in e 1a . A solution of N-trityl-O-(tert-
butyldimethylsilyl)-L-serinol9,12 (1.04 g, 2.32 mmol) and PPh3
(1.22 g, 4.65 mmol) in 14 mL of THF was cooled to 0 °C, and
diethyl azodicarboxylate (Aldrich; 0.73 mL, 4.65 mmol) was
added. The cooling bath was removed, and the reaction mixture
was allowed to warm to rt. After being stirred at rt for 18 h, the
yellow solution was poured into 50 mL of brine and extracted
with 3 × 30 mL of ether. The combined organic extract was dried
(MgSO4), and after removal of solvent (aspirator), the residue
was filtered through a 10 × 2.5 cm plug of silica, washing with
150 mL of 10:1 hexane/ether. After concentration by rotary
evaporation, the residue was purified by flash chromatography
on Whatman silica gel 60A (15 × 6 cm, 40:1 hexane/ether eluent,
20 mL fractions); fractions 15-22 gave 280 mg of the desired
aziridine as a white foam. Fractions 12-14 and 23-25 were
combined and, after concentration (rotary evaporation), repuri-
fied using flash chromatography on Whatman silica gel 60A (15
× 6 cm, 20 mL fractions, 400 mL of 99:1 hexane/ether and then
600 mL of 30:1 hexane/ether). Fractions 36-44 gave 362 mg of
Su m m a r y
The acid-induced reductive aziridine detritylation pro-
cedures can take place by two distinct pathways. Rela-
tively stable N-tritylaziridines such as 1a are deprotected
directly to the protonated aziridine, and quenching with
a tertiary amine neutralizes the salts to give the free
aziridine 2a . Solvolytically sensitive aziridines such as
3 are deprotected under similar conditions, but the choice
of acid is critical because the key intermediate is a ring-
opened ammonium salt. If deprotection of the sensitive
substrate is attempted with TFA as the acid source
followed by the tertiary amine, then the ring-opened
intermediate 12 rearranges to the N-trifluoroacetyl de-
rivative 13. With methanesulfonic acid as the proton
source (method B), the corresponding intermediate 18
recloses to the aziridine.
(11) Landor, S. R.; Sonola, O. O.; Tatchell, A. R. J . Chem. Soc.,
Perkin Trans. 1 1974, 1294. Kotera, K.; Okada, T.; Miyazaki, S.
Tetrahedron Lett. 1968, 9, 5677.
(12) Prepared by diisobutylaluminum hydride reduction of 21.
Characterization data: analytical TLC on K6F silica gel 60A, 5:1
hexane/ether, Rf ) 0.14; [R]D ) +14.1 (c ) 0.34, CH2Cl2); HRMS:
C28H37NO2Si, M + H 448.2660, error 3 ppm, base peak 243 amu; IR
(neat, cm-1) 1596, CdC; 400 MHz NMR (CDCl3, ppm) δ 7.59-7.53 (6
H, m) 7.30-7.24 (6 H, m) 7.22-7.16 (3 H, m) 3.42 (1 H, dd, J ) 10.6,
2.9 Hz) 3.33 (1 H, dd, J ) 9.9, 4.0 Hz) 3.05 (1 H, dd, J ) 9.9, 5.9 Hz)
2.89 (1 H, dd, J ) 10.6, 5.9 Hz) 2.78-2.72 (1 H, m) 2.35 (2 H, s) 0.85
(9 H, s) -0.03 (3 H, s) -0.04 (3 H, s); 13C NMR (100 MHz, CDCl3,
ppm) δ 146.8, 128.7, 127.9, 126.4, 70.8, 64.9, 63.8, 53.7, 25.8, 18.1, -5.5,
-5.5.
Under the conditions of method B, it is not necessary
to anticipate whether the ring-fused aziridines undergo
(8) Moran, E. J .; Tellew, J . E.; Zhao, Z.; Armstrong, R. W. J . Org.
Chem. 1993, 58, 7848.
(9) (a) Groneberg, R. D.; Regan, J . R.; Neuenschwander, K. W.;
Scotese, A. C. Chem. Abstr. 1995, 758641. (b) Vedejs, E.; Moss, W. O.
J . Am. Chem. Soc. 1993, 115, 7.
(10) Method B was found to cleave the OTBS group in 1a within 30
min.

Notes
J . Org. Chem., Vol. 66, No. 22, 2001 7545
the desired aziridine as a white foam (642 mg, 64% combined
yield): analytical TLC on K6F silica gel 60A, 20:1 hexane/ether,
Rf ) 0.21; [R]D ) +24.0 (c ) 1.22, CH2Cl2); no parent ion for
After 30 min at 0 °C, NEtiPr2 (2.0 mL, 11.5 mmol) was added.
The colorless solution was stirred at 0 °C for 30 min, poured
into 50 mL of ether, and washed with 2 × 20 mL of brine. The
combined organic extract was dried (Na2SO4) and concentrated
by rotary evaporation, and the residue was purified by flash
chromatography on silica gel (2 × 15 cm, 1:1 hexanes/acetone
eluent, 7 mL fractions). Fractions 11-25 gave 152 mg (63%) of
the aziridine 4.6
Meth od C. A solution of the aziridine 1c (155 mg, 0.343
mmol, 1 equiv) in 1.0 mL of anhydrous CH2Cl2 was cooled to 0
°C, and a solution of trimethylamine-borane (Aldrich; 32.6 mg,
0.446 mmol, 1.3 equiv) in dry CH2Cl2 (2 mL, including the
cannula and flask washings) was added via cannula. Trifluoro-
acetic acid (100 µL, 1.37 mmol, 4 equiv) was added dropwise
via syringe. The reaction flashed bright yellow and quickly
became colorless. Stirring was continued at 0 °C for 40 min, and
then methanesulfonic acid (31 µL, 0.48 mmol, 1.4 equiv) was
added dropwise to quench excess trimethylamine-borane (gas
evolution!). The addition of methanesulfonic acid is not required
for substrates that do not coelute with excess trimethylamine-
borane. After 5 min, NEtiPr2 (0.42 mL, 2.40 mmol, 7 equiv) was
added via syringe. The colorless solution was poured into 25 mL
of ether and washed with 25 mL of water and 25 mL of brine.
Purification as for method A gave 46 mg (61%) of aziridine 2c
as a white solid.
(2S,3R)-2-[5-[(3R) a n d (3S)-3-[(ter t-Bu tyld im eth ylsilyl)-
oxy]-5-(m et h oxyca r b on yl)-4-p en t yn yl]oxa zol-2-yl]-3-(io-
d om eth yl)a zir id in e (8). A solution of the aziridine 7 (420 mg,
0.562 mmol; prepared according to the method of ref 5) and
triethylsilane (0.36 mL, 2.25 mmol) in 10 mL of anhydrous CH2-
Cl2 was cooled to 0 °C, and trifluoroacetic acid (170 µL, 2.21
mmol) was added dropwise. The yellow color of the resulting
solution gradually faded over 5 min. After 30 min at 0 °C,
NEtiPr2 (0.48 mL, 2.75 mmol) was added. The colorless reaction
mixture was poured into 50 mL of brine and extracted with 2 ×
50 mL ether. The organic extract was dried (Na2SO4) and
concentrated by rotary evaporation, and the residue was purified
by flash chromatography on silica gel (2.5 × 15 cm, 3:1 hexane/
acetone eluent, 15 mL fractions). Fractions 8-13 gave 239 mg
(84%) of the desired aziridine (ca. 1:1 mixture of diastereomers)
as a tan oil: analytical TLC on silica gel 60 F254, EtOAc, Rf )
0.59; molecular ion (M + H+) calcd for C19H30IN2O4Si 505.10196,
found (FAB) m/e 505.1007, error 2 ppm; IR (neat, cm-1) 3234,
NsH; 2239, CtC; 1718, CdO. Preparation of the NMR sample:
ca. 2 mg of powdered 4 Å molecular sieves was suspended in a
solution of the aziridine 8 in an NMR tube: 500 MHz NMR
(CDCl3, ppm) δ 6.73 (1H, s) 4.56-4.50 (1H, m) 3.78 (3H, s) 3.53-
3.47 (1H, m) 3.41-3.35 (0.6H, m) 3.29-3.21 (1.4H, m) 2.98-
2.91 (0.6H, m) 2.86-2.79 (2H, m) 2.79-2.71 (0.4H, m) 2.10-
2.02 (2H, m) 1.77 (0.6H, dd, J ) 8.9, 8.9 Hz) 1.25-1.18 (0.4H,
m) 0.91 (9H, s) 0.16 (3H, s) 0.12 (3H, s); the NMR spectrum
displayed four sets of signals due to a 1:1 mixture of diastere-
omers and slow inversion at the aziridine nitrogen, a 3:2 ratio
of invertomers according to integration of the N-H signals at
1.77 and 1.25-1.18 ppm; 13C NMR (126 MHz, CDCl3, ppm) δ
159.13, 153.69, 152.72, 152.63, 152.06, 123.25, 122.84, 122.74,
87.74, 76.27, 61.34, 52.81, 40.87, 40.84, 39.11, 35.38, 35.31, 34.80,
34.78, 25.68, 21.04, 18.10, 4.11, 1.74, -4.54, -5.12.
Rea ction betw een th e Azir id in e 3a a n d Meth a n esu lfon -
ic Acid Mon itor ed by NMR. A solution of the aziridine 3a
(48 mg, 0.129 mmol) in 0.6 mL of anhydrous CDCl3 was cooled
to 0 °C, and methanesulfonic acid (32 µL, 0.493 mmol) was
added. The resulting dark yellow solution was immediately
transferred via cannula into a nitrogen-flushed NMR tube
capped with a rubber septum. After 5 min at rt, the 300 MHz
NMR spectrum displayed a major set of the following selected
signals assigned to the intermediate 18, a ca. 1:1 mixture of
diastereomers: δ 6.42 (0.5H, d, J ) 5.2 Hz) 6.13 (0.5H, d, J )
5.5 Hz) 4.28-4.16 (1H, br m), 3.55 (0.5H, dd, J ) 16.5, 8.2 Hz)
3.44-3.29 (1H, m) 3.28 (1.5H, s) 3.22 (0.5H, dd, J ) 16.5, 7.2
Hz) 3.14 (1.5H, s) ppm. The solution gradually turned dark green
after 5 h at rt although the 300 MHz NMR spectrum changed
only a little.
C
28H35NOSi, M + H+ 430.2574, error 2 ppm, base peak 234 amu;
IR (neat, cm-1) 1596, CdC; 400 MHz NMR (CDCl3, ppm) δ 7.51-
7.48 (6 H, m) 7.29-7.17 (9 H, m) 3.97 (1 H, dd, J ) 10.6, 5.0 Hz)
3.65 (1 H, dd, J ) 10.6, 5.9 Hz) 1.68 (1 H, d, J ) 2.9 Hz) 1.44-
1.39 (1 H, m) 1.13 (1 H, d, J ) 6.2 Hz) 0.87 (9 H, s) 0.04 (3 H, s)
0.02 (3 H, s); 13C NMR (100 MHz, CDCl3, ppm) δ 144.6, 129.5,
127.4, 126.5, 73.6, 65.9, 34.4, 25.9, 25.4, 18.3, -5.2.
P r ep a r a tion of Azir id in e 1c. The title compound was
prepared according to the procedure for 3a in 92% yield from
2c:11 analytical TLC on K6F silica gel 60A, 1:1 hexane/ether, Rf
) 0.80; pure material was obtained by crystallization from ether/
hexane, mp 149-151 °C, colorless, semitransparent; no parent
ion for C34H29N, M + Na 474.2216, error 4 ppm, base peak 208
amu; IR (neat, cm-1) 1598, CdC; 300 MHz NMR (CDCl3, ppm)
δ 7.60-7.00 (23 H, m) 6.61-6.57 (2 H, m) 3.07 (1 H, dd, J )
14.3, 3.8 Hz) 2.76 (1 H, dd, J ) 14.3, 9.2 Hz) 2.42 (1 H, d, J )
6.3 Hz) 1.82-1.72 (1 H, m); 13C NMR (100 MHz, CDCl3, ppm) δ
144.4, 139.2, 137.6, 129.7, 128.7, 128.3, 128.0, 128.0, 127.4, 126.7,
126.7, 125.8, 75.8, 41.0, 39.3, 32.6.
N-Tr itylin d a n o[1,2-b]a zir id in e (3a ). To a solution of in-
dano[1,2-b]aziridine (4)6 (689 mg, 5.25 mmol) and NEtiPr2 (1.8
mL, 10.3 mmol) in 10 mL of anhydrous CH2Cl2 was added a
solution of triphenylmethyl chloride (2.22 g, 7.96 mmol) in
anhydrous CH2Cl2 (10 mL, including the cannula and flask
washings) dropwise via cannula. After 30 min at rt, the pale
tan solution was poured into ether and washed with water
followed by brine. The organic phase was dried (Na2SO4) and
concentrated by rotary evaporation, and the residue was purified
by flash chromatography on silica gel (2.5 × 20 cm, 10:1 hexanes/
ether eluent, 20 mL fractions). Fractions 5-10 gave 1.83 g (93%)
of the product as white crystals: analytical TLC on silica gel 60
F254, 10:1 hexane/EtOAc, Rf ) 0.44; pure material was obtained
by crystallization from hexanes, fine white crystals, mp 157-
158 °C; molecular ion (M + H+) calcd for C28H24N 374.19088,
found (DCI, NH3) m/e 374.1922, error 4 ppm; IR (neat, cm-1
)
1594, CdC; 500 MHz NMR (CDCl3, ppm) δ 7.48 (6H, d, J ) 7.3
Hz) 7.38 (1H, d, J ) 6.6 Hz) 7.28-7.25 (6H, m) 7.24-7.17 (6H,
m) 3.31 (1H, d, J ) 17.1 Hz) 2.99 (1H, dd, J ) 17.1, 4.6 Hz) 2.81
(1H, d, J ) 4.6 Hz) 2.50 (1H, dd, J ) 4.6, 4.6 Hz); 13C NMR (126
MHz, CDCl3, ppm) δ 145.6, 145.1, 143.6, 129.4, 127.5, 127.0,
126.7, 125.9, 125.6, 124.2, 74.4, 43.1, 38.7, 35.5.
N-Met h oxyt r it ylin d a n o[1,2-b]a zir id in e (3b ). The title
compound was prepared according to the procedure for 3a in
74% yield from 4: Analytical TLC on K6F silica gel 60A, 5:1
hexane/ether, Rf ) 0.26 (plate pretreated with NEt3); FAB
HRMS for C29H25NO, M + Na 426.1838, error 1 ppm, base peak
273 amu; IR (neat, cm-1) 1605, CdC; 400 MHz NMR (CDCl3,
ppm) δ 7.51-7.46 (4 H, m) 7.40-7.33 (3 H, m) 7.29-7.24 (5 H,
m) 7.23-7.16 (4 H, m) 6.81 (2 H, d, J ) 9.2 Hz) 3.79 (3 H, s)
3.31 (1 H, d, J ) 17.2 Hz) 3.00 (1 H, dd, J ) 17.2, 4.6 Hz) 2.80
(1 H, d, J ) 4.6 Hz) 2.49 (1 H, dd, J ) 4.6, 4.6 Hz); 13C NMR
(100 MHz, CDCl3, ppm) δ 158.2, 145.6, 143.6, 130.8, 129.2, 127.5,
127.0, 126.6, 126.6, 125.9, 125.7, 124.2, 112.7, 74.0, 63.5, 55.2,
43.1, 38.7, 35.5.
Rep r esen ta tive P r oced u r es. Meth od A. A solution of the
aziridine 1c (110 mg, 0.243 mmol, 1 equiv) in 5.0 mL of
anhydrous CH2Cl2 was cooled to 0 °C, and triethylsilane (155
µL, 0.973 mmol, 4 equiv) was added followed by trifluoroacetic
acid (74 µL, 0.973 mmol, 4 equiv). The reaction flashed yellow
upon addition of the acid, and became colorless within 2 min.
Stirring was continued at 0 °C for 30 min, and then NEtiPr2
(210 µL, 1.21 mmol, 5 equiv) was added via syringe. After 10
min of stirring, the colorless solution was poured into 25 mL of
ether and washed with brine. The combined organic extract was
dried (MgSO4) and concentrated by rotary evaporation, and the
residue was purified by preparative TLC (20 × 10 × 0.2 cm, 1:1
hexanes/ether eluent) to give 41 mg (82%) of aziridine 2c11 as a
white solid.
Dep r otection of th e Azir id in e 3a w ith Tr ieth ylsila n e
a n d Meth a n esu lfon ic Acid (Meth od B). A solution of the
aziridine 3a (682 mg, 1.83 mmol) in 10 mL of CH2Cl2 was cooled
to 0 °C, and triethylsilane (1.2 mL, 7.51 mmol) was added
followed by methanesulfonic acid (0.49 mL, 7.55 mmol). The
yellow color of the resulting solution gradually faded over 5 min.
Rea ction betw een th e Azir id in e 3a , Tr ieth ylsila n e, a n d
Meth a n esu lfon ic Acid Mon itor ed by NMR. A solution of the
aziridine 3a (55 mg, 0.147 mmol) in 0.6 mL of anhydrous CDCl3
was cooled to 0 °C, and triethylsilane (48 µL, 0.300 mmol) was

7546 J . Org. Chem., Vol. 66, No. 22, 2001
Notes
added followed by methanesulfonic acid (38 µL, 0.586 mmol).
The resulting yellow solution was immediately transferred via
cannula into a nitrogen-flushed NMR tube capped with a rubber
septum. The yellow color faded over 1 min at rt. After 10 min at
rt, a 300 MHz NMR spectrum of the colorless solution displayed
the same characteristic set of signals as reported for the previous
experiment as well as a signal corresponding to triphenyl-
methane, δ 5.54 (1H, s) ppm.
(107 mg, 0.286 mmol) and triethylsilane (180 µL, 1.13 mmol) in
2 mL of CH2Cl2 was cooled to 0 °C, and trifluoromethanesulfonic
acid (52 µL, 0.588 mmol) was added dropwise, causing brief
flashes (ca. 5 s) of a transient yellow color. After the solution
was stirred for 30 min at 0 °C, NEtiPr2 (150 µL, 0.861 mmol)
was added. The cooling bath was removed, and the reaction
mixture was allowed to warm to rt. After 2 h, the resulting light
green solution was poured into brine and extracted with ether.
The combined organic extract was dried (Na2SO4) and concen-
trated by rotary evaporation, and the residue was purified by
flash chromatography on silica gel (1.5 × 15 cm, 2:1 hexane/
acetone eluent, 5 mL fractions). Fractions 11-15 gave 7.0 mg
(19%) of the deprotected aziridine 4 as a colorless oil. Fractions
3-7 were concentrated by rotary evaporation, and the residue
was purified by another flash chromatography on silica gel (1.5
× 15 cm, 20:1 hexane/ether eluent, 5 mL fractions). Fractions
7-10 gave 62 mg (58%) of 15 as a pale pink oil: analytical TLC
on silica gel 60 F254, 10:1 hexane/EtOAc, Rf ) 0.38; molecular
ion (M + H+) calcd for C28H24N 374.19088, found (CI, NH3) m/e
374.1894, error 4 ppm; IR (neat, cm-1) 3327, NsH; 1598, CdC;
500 MHz NMR (CDCl3, ppm) δ 7.53 (1H, d, J ) 7.1 Hz) 7.37-
7.26 (6H, m) 7.23-7.17 (4H, m) 7.14-7.09 (4H, m) 7.01-6.97
(2H, m) 6.75 (1H, dd, J ) 7.8, 1.1 Hz) 4.07 (1H, d, J ) 4.6 Hz)
3.71 (1H, dd, J ) 5.6, 4.6 Hz) 3.20 (1H, dd, J ) 16.1, 5.6 Hz)
2.73 (1H, d, J ) 16.1 Hz) 2.13 (1H, s); 13C NMR (126 MHz,
CDCl3, ppm) δ 148.6, 146.1, 144.4, 141.7, 140.4, 135.8, 130.4,
130.3, 129.1, 128.8, 127.9, 127.7, 126.8, 126.7, 126.5, 126.4, 126.2,
125.5, 125.1, 124.5, 67.6, 53.7, 47.9, 40.2.
1-Hyd r oxy-2-(tr iflu or oa ceta m id o)in d a n (13). A solution
of the aziridine 3a (171 mg, 0.458 mmol) in 5 mL of anhydrous
CH2Cl2 was cooled to 0 °C, and triethylsilane (0.30 mL, 1.88
mmol) was added followed by trifluoroacetic acid (0.14 mL, 1.82
mmol). The yellow color of the resulting solution gradually faded
over 5 min. After 30 min at 0 °C, NEtiPr2 (0.40 mL, 2.30 mmol)
was added. The colorless reaction mixture was stirred at 0 °C
for 1 min, poured into 20 mL of saturated aqueous NaHCO3,
and extracted with CH2Cl2. The combined organic extract was
dried (Na2SO4) and concentrated by rotary evaporation, and the
residue was purified by flash chromatography on silica gel (1.5
× 15 cm, 3:1 hexanes/acetone eluent, 10 mL fractions). Fractions
7-15 were concentrated, and the residue was purified by
preparative TLC on silica gel (20 × 20 × 0.1 cm, 2:1 hexanes/
ethyl acetate eluent) to give 42 mg (37%) of the product as white
crystals: analytical TLC on silica gel 60 F254, 1:1 hexane/EtOAc,
Rf ) 0.62; pure material was obtained by crystallization from
CH2Cl2, fine white crystals, mp 111-112 °C; molecular ion (M
+ NH4+) calcd for C11H14F3N2O2 263.10074, found (CI, NH3) m/e
263.1007, error 0.2 ppm; IR (neat, cm-1) 3482, NsH; 3308, Os
H; 1698, CdO; 500 MHz NMR (CDCl3, ppm) δ 7.41 (1H, d, J )
7.3 Hz) 7.33 (1H, ddd, J ) 7.3, 7.3, 1.2 Hz) 7.29-7.22 (3H, m)
5.07 (1H, dd, J ) 4.8, 4.8 Hz) 4.59-4.53 (1H, m) 3.34 (1H, dd, J
) 16.0, 7.5 Hz) 2.94 (1H, dd, J ) 16.0, 7.5 Hz) 2.46 (1H, d, J )
4.8 Hz); 13C NMR (126 MHz, CDCl3, ppm) δ 157.4 (q, J ) 37.1
Hz), 141.2, 140.5, 129.8, 127.6, 125.4, 125.1, 115.8 (q, J ) 287.9
Hz), 74.0, 53.2, 36.3; 19F NMR (376 MHz, CDCl3, ppm) δ -76.3.
(6a S*,11bS*)-5,5-Dip h en yl-6,6a ,7,11b-tetr a h yd r o-5H-in -
d en o[2,1-c]isoqu in olin e (15). A solution of the aziridine 3a
Ack n ow led gm en t. This work was supported by the
National Institutes of Health (Grant CA17918).
Su p p or tin g In for m a tion Ava ila ble: NMR spectra of the
new compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
J O0106243