Aromatic allylation via diazotization: Variation of the allylic moiety and a short route to a benzazepine derivative

Article abstract of DOI:10.1021/jo026784b

A continued study of the recently discovered diazotizative allylation (DiazAll) reaction of aniline derivatives is reported. Several allyl reagents, commonly used in radical allylation reactions, were evaluated, and some of these reagents resulted in ally

Full text of DOI:10.1021/jo026784b

Ar om a tic Allyla tion via Dia zotiza tion : Va r ia tion of th e Allylic
Moiety a n d a Sh or t Rou te to a Ben za zep in e Der iva tive
Fredrik Ek,† Lars-Go¨ran Wistrand,‡ and Torbjo¨rn Frejd*,†
Organic Chemistry 1, Department of Chemistry, Lund University, P.O. Box 124, S-221 00 Lund, Sweden,
and Amersham Health R&D AB, Medeon-Malmo¨, S-205 12 Malmo¨, Sweden
torbjorn.frejd@orgk1.lu.se
Received November 29, 2002
A continued study of the recently discovered diazotizative allylation (DiazAll) reaction of aniline
derivatives is reported. Several allyl reagents, commonly used in radical allylation reactions, were
evaluated, and some of these reagents resulted in allylation when used in the DiazAll reaction.
The best result was obtained with allyl bromide. Substituted allylic bromides gave the corresponding
allyl aromatic compounds in poor to excellent yields. In comparison with an established method
for aromatic allylation, the DiazAll reaction performed well and was superior when a more complex
allylic bromide was used. Finally, a new allylation-bromocyclization reaction was demonstrated
and used in the synthesis of a known inhibitor of phenylethanolamine N-methyltransferase (PNMT),
an enzyme involved in the biosynthesis of adrenaline.
In tr od u ction
derivatives are used, it is possible to access biologically
relevant substances, which will be demonstrated in this
paper.
Free-radical allylations provide some of the mildest
methods for the introduction of an allyl functionality in
organic compounds, although most of the reported ap-
plications involve alkyl radicals.1-3 Recently, we reported
a novel method for the synthesis of allyl aromatic
compounds via diazotization of arylamines with tert-butyl
nitrite in acetonitrile and allyl bromide.4,5 For simplicity,
we call the reaction DiazAll (diazotizative allylation). It
was found that a large number of different substituents
were tolerated due to the mild reaction conditions. A
convenient experimental procedure avoiding metal-
containing reagents and catalysts together with a dem-
onstrated insensitivity toward moisture, air, type of
solvent, and short reaction times made it useful for
multigram synthesis. The mechanism probably involves
a phenyl radical intermediate attacking allyl bromide as
seen in Scheme 1. In our previous paper, we investigated
structural changes in the aromatic moiety. We now
continue our study of the scope and limitation of the
DiazAll reaction and focus on the structural variation of
the allylic moiety. 3,5-Dinitroaniline was chosen as a
standard starting material because of the possibility to
use the corresponding allylated products in the synthesis
of iodinated X-ray contrast agents.6 When other aniline
Resu lts a n d Discu ssion
The experimental procedure of the DiazAll reaction is
very convenient and robust in comparison with most
other aromatic allylation reactions that usually require
sensitive reagents. However, dry reaction conditions and
inert atmosphere were employed in the present study,
but as reported previously, the reactions could be per-
formed without these precautions and still give almost
as good yields. The arylamine (solid) was added in
portions to an acetonitrile solution of tert-butyl nitrite
and the appropriate allylic component during approxi-
mately 20 min, if otherwise not stated, at a temperature
ensuring a gentle formation of nitrogen gas. WARNING!
A too r a p id a d d ition of th e a r yla m in e to th e r ea c-
tion m ixtu r e m a y r esu lt in a n u n con tr olled evolu -
tion of h ea t a n d ga s. A reaction time of 60 min was
employed although in several cases complete conversion
was achieved already ca. 5-10 min after the final
addition of the arylamines.
A number of different allylating reagents have been
reported in radical allylation reactions, the most common
reagents being allyltributyltin.1,3 However, the potential
toxicity of organotin compounds and the difficulties in
removal of tin byproducts have turned the focus toward
the development of alternative reagents such as sulfides,
sulfoxides, sulfones, silanes, and halides.3 In the ongoing
study of the DiazAll reaction, we were interested to see
which of these functionalities were best suited under the
present reaction conditions.
† Lund University.
‡ Amersham Health R&D AB.
(1) Keck, G. E.; Enholm, E. J .; Yates, J . B.; Wiley, M. R. Tetrahedron
1985, 41, 4079-94.
(2) Kosugi, M.; Kurino, K.; Takayama, K.; Migita, T. J . Organomet.
Chem. 1973, 56, C11-C13.
(3) Radicals in Organic Synthesis; Renaud, P., Sibi, M. P., Eds.;
Wiley-VCH: Weinheim, 2001; Vols. 1 and 2.
(4) Ek, F.; Axelsson, O.; Wistrand, L.-G.; Frejd, T. J . Org. Chem.
2002, 67, 6376-6381.
(5) Ek, F.; Wistrand, L. G. Preparation of allylic aromatic compounds
by reaction of aromatic amines with a nitrite and an allylic olefin. EP
1013636, 2000, 11 pp.
(6) Andersson, S.; Malmgren, H.; Golman, K.; Wistrand, L.-G.;
Almen, T.; Hanno, P. Water-soluble contrast media for X-ray imaging.
WO 0037115, 2000, 35 pp.
10.1021/jo026784b CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/05/2003
J . Org. Chem. 2003, 68, 1911-1918
1911

Ek et al.
SCHEME 1. Su ggested Mech a n ism of th e Dia zAll Rea ction
TABLE 1. In vestiga tion of Differ en t Lea vin g Gr ou p s in
th e Dia zAll Rea ction w ith 3,5-Din itr oa n ilin ea
for the failure could be that the highly nonpolar character
of these reagents may influence the polarity of the
reaction mixture thus preventing the initial, presumably,
ionic reaction between 1 and tert-butyl nitrite.
Since allyl bromide gave the best yield, we continued
to study the effects of structural variation in the allylic
moiety using allylic bromides as preferred reagents.
Thus, a number of different substituted allyl aromatic
compounds were synthesized in poor to excellent yields
(see Table 2). The resulting allyl derivatives and, in
particular, vinylic, allylic, and homoallylic bromides and
allyl acetates are obvious starting materials for the
synthesis of more complex organic molecules. In ionic
reactions involving highly reactive reagents there is often
a need for protection of sensitive functional groups.
However, in radical reactions, this can be avoided both
because O-H and N-H bonds are resistant toward
homolytical cleavage and the absence of the typical
nucleophilic behavior of anionic species. This was exem-
plified by the one-step synthesis of acrylic acid 2f (Table
2) without the need for protection of the carboxylic acid
moiety. Previously reported methods of the synthesis of
2-benzylacrylic acid derivatives involves at least two
synthetic steps.10,11 This type of compounds are important
in the synthesis of pseudopeptides of pharmacological
interest, e.g., inhibitors of neutral endopeptidase (NEP),
tumor necrosis factor (TNF), metalloprotease (MMP), and
angiotensin-converting enzymes (ACE).12-15
L
yield (%)
L
yield (%)
Brb
74
18
51
29
SMe
S(allyl)
SPh
28
34
61
25
Ic
SO2Ph
SO2CF3
SiMe3
a
A general procedure (internal standard) was used as described
b
in the Experimental Section. Selectivity: allyl-3,5-dinitroben-
zene/3,5-dinitrobromobenzene ) 25:1.4 c Selectivity: allyl-3,5-
dinitrobenzene/3,5-dinitroiodobenzene ) 1:4.3.
As seen in Table 1, allyl bromide gave the highest yield
of the allylated product, although both allyl phenyl
sulfide and allyl phenyl sulfone gave fair yields. Allyl
trifluoromethyl sulfones7 and allyltrimethylsilane3 have
been reported to give high yields in allylation of alkyl
radicals. However, using these reagents in this DiazAll
reaction gave only poor yields of 2a . Moreover, in ac-
cordance with an earlier paper,8 using allyl iodide in the
radical allylation reaction mainly resulted in iodination
and only 18% of 2a . This is probably due to a too
homolytically weak carbon-iodide bond which favors
halogen abstraction over addition to the double bond.
On the other hand, the use of allyl chloride did not result
in any allylation due to a too strong carbon-chlorine
bond toward homolytical cleavage. Surprisingly, allyl-
tributyltin did not give the allylated product despite
numerous reports of successful radical allylations using
this reagent.3,9 Only the starting material 1 was recov-
ered at the end of the experiments, although tempera-
ture, concentration, and type of solvent were varied.
Similarly, allyltris(trimethylsilyl)silane, allyltriisopro-
pylsilane, allyltriphenyltin, and allyldiphenylphosphine
were nonproductive. A large excess of the allylic reagents
is used the DiazAll reaction. Hence, one plausible reason
The yields dropped significantly when allyl bromides
with internal double bonds were used (Table 2, 2h and
2i). In these cases, the amount of the byproduct 3,5-
dinitrobromobenzene increased as well. There seems to
be ambiguous conclusions concerning the sensitivity of
(10) Issa, J .; Andrews, P. R.; Iskander, M. N.; Reiss, J . A. Synth.
Commun. 1995, 25, 1489-94.
(11) Hin, B.; Majer, P.; Tsukamoto, T. J . Org. Chem. 2002, 67, 7365-
7368.
(12) Robl, J . A. Preparation of N-formylhydroxylamine-containing
peptidyl compounds useful as ACE and NEP inhibitors. WO 9738705,
1997, 65 pp.
(13) Owen, D. A.; Montana, J . G.; Keily, J . F.; Watson, R. J .; Baxter,
A. D. Hydroxamic and carboxylic acid derivatives having MMP and
TNF inhibitory activity. WO 9805635, 1998, 88 pp.
(14) Acton, S. L.; Ocain, T. D.; Gould, A. E.; Dales, N. A.; Guan, B.;
Brown, J . A.; Patane, M.; Kadambi, V. J .; Solomon, M.; Stricker-
Krongrad, A. Preparation of amino acid derivatives for modulating
angiotensin converting enzyme-2 (ACE-2). WO 0239997, 2002, 395 pp.
(15) Barlaam, B. C.; Dowell, R. I.; Newcombe, N. J .; Tucker, H.;
Waterson, D. Preparation of arylpiperazines and arylpiperidines as
metalloproteinase inhibiting agents. WO 0162751, 2001, 35 pp.
(7) Xiang, J .; Evarts, J .; Rivkin, A.; Curran, D. P.; Fuchs, P. L.
Tetrahedron Lett. 1998, 39, 4163-4166.
(8) Migita, T.; Kosugi, M.; Takayama, K.; Nakagawa, Y. Tetrahedron
1973, 29, 51-5.
(9) Keck, G. E.; Byers, J . H. J . Org. Chem. 1985, 50, 5442-4.
1912 J . Org. Chem., Vol. 68, No. 5, 2003

Aromatic Allylation via Diazotization
TABLE 2. Resu lts of th e Allyla tion of 3,5-Din itr oa n ilin e on a 3 m m ol Sca le (DNP ) 3,5-Din itr op h en yl)
a
b
Temperature during the addition of 3,5-dinitroaniline/temperature after the addition. Isolated yields if otherwise not stated. c See
ref 4. 2.0 mmol scale. e 1.0 mmol scale. The yield was determined by 1H NMR spectroscopy using toluene as internal standard. 7% of the
d
corresponding cis isomer of 2j was also formed. The isolated yield of 2j was 35%.
radical reactions toward sterical hindrance. For instance,
Keck et al. found that crotyltributyltin/(t-BuO)2 did not
give the expected allylation of an alkyl bromide but
instead resulted in reductive dehalogenation.1 However,
Ryu et al. later demonstrated a successful allylation using
this particular reagent, although the yields were some-
what lower compared to allyltributyltin.16 Also Fuchs et
al. failed to use sterically hindered allylic triflones for
allylic substitution of C-H positions.7 The examples
shown (Table 2, 2h and 2i) are, to our knowledge, the
first cases of allylation of phenyl radicals using sterically
hindered allyl derivatives.
We also wanted to use 1-substituted allyl bromides in
the DiazAll reaction. However, they are often difficult to
synthesize due to the rapid isomerization to the more
stable 3-substituted allylic bromides as demonstrated by
the equilibrium at room temperature between 3-bro-
mobutene and crotyl bromide favoring the latter (15:85).17
Fortunately, it was possible to synthesize isomerically
pure allylic bromide 3 from butadiene monoxide.18,19
Using 3 in the DiazAll reaction together with 1 equiv of
1
pyridine gave 43% of 2j according to H NMR analysis
(Table 2). The yield of 2j dropped significantly if pyridine
was excluded.20 Compared to the nonsubstituted or
2-substituted allylic bromides a larger amount of 3,5-
dinitrobromobenzene was formed when 3 was employed
as the allylic component. Aryl radical attack on the
bromine atom of 3 would possibly be aided by anchimeric
(18) Hase, T. A.; Lahtinen, L.; Klemola, A. Acta Chem. Scand., Ser.
B 1977, B31, 501-4.
(19) Petrov, A. A. J . Gen. Chem. (U.S.S.R.) 1941, 11, 991-5.
(20) Using 3 in the DiazAll reaction gave according to 1H NMR
analysis not only the expected products 2j, the corresponding cis iso-
mer of 2j, and 3,5-dinitrobromobenzene but also 3,5-dinitro-1-(1-
acetoxymethyl-2-propenyl)benzene in considerable amounts (ratio
1:0.15:1:1). We also noted that all of 3 had isomerized to 4-bromobut-
2-enyl acetate during the reaction. However, by adding 1 equiv of
pyridine to the reaction mixture before the addition of 1, the ratio was
changed in favor of 2j (ratio 2:0.3:1:0.3). The ratio between 2j and the
corresponding cis isomer of 2j was the same in both procedures.
Apparently, the added pyridine suppresses the rearrangement of 3 to
4-bromo-but-2-enyl acetate, which was confirmed by 1H NMR analysis
of the crude product.
(16) Ryu, I.; Kreimerman, S.; Niguma, T.; Minakata, S.; Komatsu,
M.; Luo, Z.; Curran, D. P. Tetrahedron Lett. 2001, 42, 947-950.
(17) Winstein, S.; Young, W. G. J . Am. Chem. Soc. 1936, 58,
104-7.
J . Org. Chem, Vol. 68, No. 5, 2003 1913

Ek et al.
SCHEME 2. Com p a r ison betw een th e Dia zAll
Rea ction a n d th e Ha lid e-Ma gn esiu m Exch a n ge
Rea ction
used, the yield of 10 never exceeded 35% even when a
number of parameters were varied (temperature, amount
of CuCN, concentration, and the use of slow addition of
the in situ generated arylmagnesium species to 8). The
product was always accompanied by the symmetrical bis-
arylated byproduct and other unidentified compounds.
When allyl bromide itself was used, the result was
comparable to those reported in the literature for similar
systems. We therefore decided to investigate the pos-
sibility of using the DiazAll reaction in the synthesis of
10. In contrast to the previous method, the yield in fact
increased somewhat when 8 was employed compared to
the case in which allyl bromide was used (from 72% to
82%). Although 10 equiv of 8 was added, five of these
could be recovered and reused. Comparing these two
methods clearly shows the potential of the DiazAll
reaction especially when using more complex allyl bro-
mides, and the good results may even justify the need of
an excess of the allylic bromides.
a
The yields were determined by 1H NMR spectroscopy using
toluene as internal standard.
The cyano group of aromatic nitriles is an interesting
handle for further derivatization, e.g., to generate het-
erocycles such as tetrazoles. This motivated the synthesis
of the tricylic fused tetrazole 13 belonging to a group of
compounds which have shown affinity for GABA-recep-
tors (Scheme 3).28 The tetrazole 11 was conveniently
synthesized in 84% yield from the corresponding com-
mercially available anthranilonitrile.29,30 We could imag-
ine that after an initial allylation of 11 with 8 the product
12 would be ready for cyclization, which might take place
either in situ or in a separate synthetic step. The
allylation took place indeed, but the final result was not
the expected compounds 12 or 13 but 15b. As indicated
in Scheme 3, 14b could react with bromine in a reaction
that resembles a bromolactonisation. We have earlier
suggested that bromine is formed during the DiazAll
reaction and the bromocyclization further supports this
hypothesis.4 The preliminary result with 8 was confirmed
with metallyl bromide and allyl bromide which gave 15%
of 15c and 22% of 15a , respectively (Scheme 3). Also the
anthranilic acid 16 gave the isocoumarin 17 in 39% yield
(Scheme 4). To obtain complete conversion of the aryl-
amine, it was sometimes necessary to add more than 2
equiv of tert-butyl nitrite to the reaction mixture. Al-
though the yields are poor, the allylation-bromocycliza-
tion constitutes a tandem reaction, which gives access
to useful substances from commercially available re-
agents in a one-pot procedure. For instance, there is now
a possibility to synthesize 19 in merely three steps, where
previously seven steps were required (Scheme 4).31 The
benzoazepine 19 is a potent inhibitor of phenylethano-
lamine N-methyltransferase (PNMT), an enzyme in-
volved in the biosynthesis of adrenaline.32 Moreover, with
the new tandem reaction it is possible to access new
analogues of 19, which could be difficult to synthesize
with the previously published methodology. Compound
assistance from the acetoxy group, thus decreasing the
carbon-bromine bond strength.
Alkyl radicals have been reported to react with tri-
phenylprop-2-ynylstannane to give terminal allens.21 In
a similar fashion, propargyl bromide was used in the
DiazAll reaction, which gave allene 2k , although the yield
was low (Table 2).
Some olefins could not be used in the DiazAll reaction
(Table 2). The dienyl bromide 4, which was synthesized
from penta-1,4-dien-3-ol using PBr3,22 did not give the
expected product. This is in contrast to the tributyltin
analogue of 4, which was reported to react with alkyl
radicals to give the dienyl products.23-25 One could
speculate that even if the diene is formed in the DiazAll
reaction this type of compounds are often quite reactive
and could continue to react with the bromine/bromine
radical formed in the reaction.
Unfortunately, sulfide 5 gave a large number of reac-
tion products which were not further analyzed. This
was somewhat unexpected since allyl phenyl sulfide
was successfully used in the DiazAll reaction as seen in
Table 1.
In a related project, we needed 10 in gram quantities
(Scheme 2). Several methods were considered, but we
decided to use the halide-magnesium exchange reaction,
which has proven to give good to excellent yields of allyl
aromatic compounds.26
The starting material 7 was synthesized utilizing a
modification of the method reported by Friedman et al.;
we replaced CCl4 with acetonitrile or some other water
miscible solvent.27
However, the result of the magnesium exchange reac-
tion was disappointing. Even though 5 equiv of 8 was
(21) Baldwin, J . E.; Adlington, R. M.; Basak, A. J . Chem. Soc., Chem.
Commun. 1984, 1284-5.
(28) Daya, S.; Kaye, P. T.; Mphahlele, M. J . Med. Sci. Res. 1996,
24, 137-41.
(29) Tetrazole 11 has been synthesized before but has not been
properly characterized. A synthetic procedure with characterization
data is found in the Supporting Information.
(30) Koguro, K.; Oga, T.; Mitsui, S.; Orita, R. Synthesis 1998, 910-
914.
(31) Grunewald, G. L.; Dahanukar, V. H. J . Heterocycl. Chem. 1994,
31, 1609-17.
(32) Grunewald, G. L.; Dahanukar, V. H.; Criscione, K. R. Bioorg.
Med. Chem. 2001, 9, 1957-1965.
(22) Prevost, C.; Miginiac, P.; Miginiac-Groizeleau, L. Bull. Soc.
Chim. Fr. 1964, 2485-92.
(23) Landais, Y.; Planchenault, D. Tetrahedron 1995, 51, 12097-
108.
(24) Kraus, G. A.; Siclovan, T. M.; Watson, B. Synlett 1995, 201-2.
(25) Kraus, G. A.; Andersh, B.; Su, Q.; Shi, J . Tetrahedron Lett. 1993,
34, 1741-4.
(26) Boudier, A.; Bromm, L. O.; Lotz, M.; Knochel, P. Angew. Chem.,
Int. Ed. 2000, 39, 4415-4435.
(27) Friedman, L.; Chlebowski, J . J . Org. Chem. 1968, 33, 1636-8.
1914 J . Org. Chem., Vol. 68, No. 5, 2003

Aromatic Allylation via Diazotization
SCHEME 3. A New Allyla tion -Br om ocycliza tion Rea ction
SCHEME 4. A th r ee-step syn th esis of 19 a n d 16
a
The yield was determined by 1H-NMR spectroscopy using toluene as internal standard.
19 has also been utilized as starting material in the
synthesis of substances which inhibit the neuronal iso-
form of nitric oxide synthase and thereby may be useful
in the treatment of a number of diseases or conditions,
e.g., stroke (hypoxia), ischaemia, and pain.33
aromatic system. The major strength of the reaction is
the convenient experimental procedure, the tolerance of
functionalities, and the possibility to access allyl aromatic
compounds directly from commercially available aryl-
amines, circumventing aryl halides or arylboronic acids,
common starting materials in other allylation reactions.
The DiazAll reaction is under continuous development
in our laboratories.
In conclusion, allyl bromide was found the be the
reagent of choice in the DiazAll reaction although both
allyl phenyl sulfide and allyl phenyl sulfone gave accept-
able yields. More surprising was the failure of allyltribu-
tyltin, which is otherwise extensively used in radical
allylation reactions. Several functionalities, which could
cause problems in other methods, were tolerated. Some
of the compounds generated, e.g., the aromatic allyl
bromide, allyl acetates, and vinyl bromide, would be
useful starting materials for further transformations into
more complex molecules. Even when sterically hindered
allyl bromides were employed in the DiazAll reaction the
allylated products could be isolated, although the yields
were poor. These results confirm the observation made
in our previous study that steric hindrance is badly
tolerated. In comparison with an established method for
the synthesis of allylated aromatic compounds, the Di-
azAll reaction performed well and was even superior
when a more complex allylic bromide was used, both
regarding the yield but also because of the more conve-
nient experimental procedure. Finally, a new tandem
allylation-bromolactonization reaction was demonstrated,
and one of the products was used in the synthesis of a
known inhibitor of PNMT.
Exp er im en ta l Section
Gen er a l Meth od s. HPLC analyses were performed on a
HiChrom column (Kromasil 100-5C18, 150 × 4.6 mm); elu-
ent: CH3CN (HPLC grade)/H2O (0.1% TFA); flow rate 1 mL/
min. Preparative HPLC was performed with a HiChrom
column (Kromasil 100-10C18, 250 × 20 mm); eluent: CH3CN
(HPLC grade)/H2O; flow rate 20 mL/min. NMR spectra were
recorded on a 400 MHz instrument using CDCl3, DMSO-d6 or
acetone-d6 as internal standard. Elemental analyses were
made by A. Kolbe, Mikroanalytisches Laboratorium, Germany.
Chromatographic separations were performed on Matrex
Amicon normal-phase silica gel 60 (0.035-0.070 mm). Thin-
layer chromatography was performed on Merck precoated TLC
plates with silica gel 60 F-254, 0.25 mm. After eluation, the
TLC plates were visualized with UV light and sprayed with a
solution of KMnO4 (10 g), K2CO3 (50 g), NaOH (20 mL, 5%),
and H2O (900 mL) followed by heating. Chemicals were
reagent grade except for 2,3-dibromopropene, which was of
technical grade (80% purity). All reagents were used as
received if not otherwise noted. Propargyl bromide (97%
purity), 3-bromocyclohexene, 3-bromo-2-methylpropene, and
allyl bromide were distilled, and 5-nitroanthranilic acid (95%
purity) was recrystallized from MeOH prior to use. i-PrMgBr
was freshly prepared from 2-bromopropane and magnesium.
3-Bromo-2-bromomethylpropene,34 3-bromo-2-acetoxymethyl-
The DiazAll reaction is a useful complement to existing
methodology to introduce an allyl functionality into an
(33) Macdonald, J .; Matz, J .; Shakespeare, W. Preparation of
N-tetrahydroisoquinolylthiophenecarboximidamides and analogues as
neuronal nitric oxide synthase inhibitors. WO 9717344, 1997, 86 pp.
(34) Axenrod, T.; Das, K. K.; Yazdekhasti, H.; Dave, P. R.; Stern,
A. G. Org. Prep. Proced. Int. 1997, 29, 358-360.
J . Org. Chem, Vol. 68, No. 5, 2003 1915

Ek et al.
propene,35 2-bromo-but-3-enyl acetate,18,19 3-bromo-2-phenyl-
propene,36 5-bromo-penta-1,3-diene,22 1-phenyl-2-phenylsul-
fanylbut-3-en-1-ol,37 3-trifluoromethanesulfonylpropene,38 and
allyltris(trimethylsilyl)silane39 were synthesized according to
literature procedures. Solvents were of p.a. quality except for
the acetonitrile used in analytical and preparative HPLC,
which was of HPLC grade. The acetonitrile used in the
reactions was stored over molecular sieves (4 Å).
Gen er a l P r oced u r e for th e Allyla tion . The arylamine
(3.0 mmol) was added during 20 min to a solution of tert-butyl
nitrite (535 µL, 4.5 mmol) and the appropriate allylic reagent
(45.0 mmol) in dry CH3CN (3.0 mL if not otherwise stated)
under an argon atmosphere while maintaining the specified
temperature. At the end of the addition of the arylamine, extra
tert-butyl nitrite (180 µL, 1.5 mmol) was added. The reaction
mixture was then stirred at the specified temperature for 1 h.
Acetonitrile, tert-butyl nitrite, and the remaining allylic
reagent (depending on the boiling point) were distilled off at
reduced pressure (10 mmHg).
With In ter n a l Sta n d a r d . 3,5-Dinitroaniline (92 mg, 0.5
mmol) was added to a solution of tert-butyl nitrite (119 µL,
1.0 mmol) and the allylic reagent (see Table 1) in CH3CN (0.5
mL) during 10 min while maintaining the temperature of the
reaction mixture between 11.5 and 14 °C. The reaction mixture
was then stirred at 22 °C for 1 h followed by addition of
2-nitrobenzyl alcohol (50 mg, 0.33 mmol) as an internal
standard. The yield and product distribution were determined
by HPLC using a calibration curve.40
3,5-Din itr o-1-(2-acetoxym eth yl-2-pr open yl)ben zen e (2e).
The reaction was performed according to the general procedure
on a 2.0 mmol scale. The remaining allyl bromide was removed
by distillation at reduced pressure. Column chromatography
(heptane-EtOAc 5:1) followed by crystallization (heptane-
ether 2:1) gave 420 mg (75%) of the title compound as yellow
crystals: mp 48-49 °C; 1H NMR (CDCl3, 400 MHz) δ 8.91, (t,
1H, J ) 2.1 Hz), 8.41, (d, 2H, J ) 2.1 Hz), 5.33, (s, 1H), 5.05,
(s, 1H), 4.51, (s, 2H), 3.64, (s, 2H), 2.06, (s, 3H); 13C NMR
(CDCl3, 100 MHz) δ 170.6, 149.0, 143.4, 141.0, 129.5, 117.7,
117.4, 66.1, 39.6, 21.0. Anal. Calcd for C12H12N2O6: C, 51.43;
H, 4.32; N, 10.00. Found: C, 51.52; H, 4.26; N, 9.92.
2-(3,5-Din itr oben zyl)a cr ylic Acid (2f). The reaction was
performed according to the general procedure except that three
times the amount of CH3CN was used as solvent. The remain-
ing crystalline residue was washed with heptane (5 × 50 mL)
to remove nonreacted 2-(bromomethyl)acrylic acid (2-(bromo-
methyl)acrylic acid crystallized from the combined heptane
phases). Column chromatography (heptane-EtOAc-MeOH-
HOAc 5:4:1:0.04) of the remaining crude product followed by
crystallization from toluene gave 450 mg (60%) of the title
compound as large yellow needles: mp 149-151 °C; 1H NMR
(DMSO-d6, 400 MHz) δ 12.72, (br s, 1H), 8.69, (t, 1H, J ) 2.0
Hz), 8.50, (br d, 2H, J ) 1.7 Hz), 6.22, (br s, 1H), 5.84, (br s,
1H), 3.86, (s, 2H); 13C NMR (DMSO-d6, 100 MHz) δ 167.2,
147.8, 144.0, 138.7, 129.1, 127.8, 116.5, 36.4. Anal. Calcd for
C
10H8N2O6: C, 47.63; H, 3.20; N, 11.11. Found: C, 47.74; H,
3.16; N, 10.97.
3,5-Din itr o-1-(2-m eth yl-2-p r op en yl)ben zen e (2b). The
reaction was performed according to the general procedure.
Column chromatography (heptane-EtOAc 23:2) gave 670 mg
of a pale yellow oil, which crystallized upon standing overnight
in the refrigerator. Recrystallization (heptane-EtOAc 49:1)
gave 500 mg (75%) of the title compound as pale yellow
crystals: mp 46-47 °C; 1H NMR (CDCl3, 400 MHz) δ 8.92, (t,
1H, J ) 2.1 Hz), 8.41, (d, 2H, J ) 2.1 Hz), 5.01, (s, 1H), 4.83,
(s, 1H), 3.56, (s, 2H), 1.74, (s, 3H); 13C NMR (CDCl3, 100 MHz)
δ 149.0, 144.8, 142.5, 129.5, 117.4, 115.3, 44.4, 22.4. Anal.
Calcd for C10H10N2O4: C, 54.05; H, 4.54; N, 12.61. Found: C,
54.19; H, 4.66; N, 12.48.
3,5-Din itr o-1-(2-br om o-2-p r op en yl)ben zen e (2c). The
reaction was performed according to the general procedure.
The remaining allyl bromide was removed by distillation at
reduced pressure. Column chromatography (heptane-EtOAc
10:1) gave 800 mg (93%) of the title compound as a pale yellow
oil: 1H NMR (CDCl3, 400 MHz) δ 8.98, (t, 1H, J ) 2.1 Hz),
8.47, (d, 2H, J ) 2.1 Hz), 5.89, (m, 1H), 5.71, (d, 1H, J ) 2.1
Hz), 4.00, (s, 2H); 13C NMR (CDCl3, 100 MHz) δ 149.05, 142.1,
129.6, 129.3, 121.4, 118.2, 47.3. Anal. Calcd for C9H7BrN2O4:
C, 37.66; H, 2.46; N, 9.76. Found: C, 37.75; H, 2.38; N, 9.78.
3,5-Din itr o-1-(2-br om om eth yl-2-pr open yl)ben zen e (2d).
The reaction was performed according to the general proce-
dure. The remaining allyl bromide was removed by distillation
at reduced pressure (0.7 mmHg, 30 °C, 5.3 g, 55% recovery).
Column chromatography (heptane-EtOAc 9:1) followed by
crystallization (heptane-EtOAc 9:1) gave 645 mg (71%) of the
title compound as pale yellow crystals: mp 73-74 °C; 1H NMR
(CDCl3, 400 MHz) δ 8.96, (t, 1H, J ) 2.1 Hz), 8.45, (d, 2H, J
) 2.1 Hz), 5.44, (s, 1H), 5.03, (s, 1H), 3.93, (s, 2H), 3.81, (s,
2H); 13C NMR (CDCl3, 100 MHz) δ 149.1, 143.4, 142.8, 129.7,
119.5, 117.9, 39.9, 35.5. Anal. Calcd for C10H9BrN2O4: C, 39.89;
H, 3.01; N, 9.30. Found: C, 39.83; H, 2.93; N, 9.18.
3,5-Din itr o-1-(2-p h en yl-2-p r op en yl)ben zen e (2g). The
reaction was performed according to the general procedure
except that twice the amount of CH3CN was used as solvent.
The remaining allyl bromide was removed by distillation at
reduced pressure (0.3 mmHg, 50-55 °C, 5.9 g, 66% recovery).
Column chromatography (heptane-EtOAc 49:1 to 23:2) fol-
lowed by crystallization (heptane-EtOAc 49:1) gave 495 mg
(58%) of the title compound as pale yellow crystals: mp
1
130-131 °C; H NMR (CDCl3, 400 MHz) δ 8.87, (t, 1H, J )
2.1 Hz), 8.42, (d, 2H, J ) 2.1 Hz), 7.42-7.39, (m, 2H), 7.35-
7.25, (m, 3H), 5.65, (s, 1H), 5.22, (d, 1H, J ) 0.8 Hz), 4.09, (s,
2H); 13C NMR (CDCl3, 100 MHz) δ 148.9, 144.9, 144.6, 139.3,
129.4, 129.2, 128.7, 126.5, 126.5, 117.5, 117.3, 41.7. Anal. Calcd
for C15H12N2O4: C, 63.38; H, 4.25; N, 9.85. Found: C, 63.23;
H, 4.24; N, 9.76.
3,5-Din itr o-1-(1-br om om eth yl-2-pr open yl)ben zen e (2h ).
The reaction was performed according to the general procedure
except that three times the amount of CH3CN was used as
solvent. Column chromatography (heptane-EtOAc 9:1) gave
250 mg (28%) of the title compound as pale yellow crystals:
1
mp 59-60 °C; H NMR (CDCl3, 400 MHz) δ 8.98, (t, 1H, J )
2.1 Hz), 8.41, (dd, 2H, J ) 2.1, 0.5 Hz), 6.08-5.98, (ddd, 1H,
J ) 17.2, 10.4, 7.3 Hz), 5.39, (ddd, 1H, J ) 10.4, 1.0, 0.7 Hz),
5.28, (ddd, 1H, J ) 17.2, 1.3, 0.7), 4.01, (br q, 1H, J ) 7.2 Hz),
3.81-3.76, (dd, 1H, J ) 10.5, 5.5 Hz), 3.70-3.65, (dd, 1H, J )
10.4, 8.6 Hz); 13C NMR (CDCl3, 100 MHz) δ 149.1, 146.0, 136.4,
128.7, 120.0, 118.2, 51.2, 35.0. Anal. Calcd for C10H9BrN2O4:
C, 39.89; H, 3.01; N, 9.30. Found: C, 40.03; H, 2.95; N, 9.25.
3-(3,5-Din itr op h en yl)cycloh exen e (2i). The reaction was
performed according to the general procedure. The remaining
allyl bromide was removed by distillation at reduced pressure.
Column chromatography (cyclohexane-t-BuOMe 15:1) fol-
lowed by crystallization (heptane-EtOAc 49:1) gave 210 mg
(28%) of the title compound as pale yellow crystals: mp
1
90-91 °C; H NMR (CDCl3, 400 MHz) δ 8.90, (t, 1H, J ) 2.1
(35) Magnusson, G.; Lindqvist, F. J . Chem. Soc., Chem. Commun.
1990, 1080-1.
Hz), 8.43, (d, 2H, J ) 2.1 Hz), 6.13-6.08, (m, 1H), 5.73-5.69,
(m, 1H), 3.68, (m, 1H), 2.20-2.12, (m, 3H), 1.75-1.57, (m, 3H);
13C NMR (CDCl3, 100 MHz) δ 151.71, 149,0 131.9, 128.5, 127.2,
117.2, 41.9, 32.7, 25.1, 20.9. Anal. Calcd for C12H12N2O4: C,
58.06; H, 4.87; N, 11.29. Found: C, 58.14; H, 4.96; N, 11.36.
(36) Pines, H.; Alul, H.; Kolobielski, M. J . Org. Chem. 1957, 22,
1113-14.
(37) Ridley, D. D.; Smal, M. A. Aust. J . Chem. 1980, 33, 1345-55.
(38) Hendrickson, J . B.; J udelson, D. A.; Chancellor, T. Synthesis
1984, 320-2.
tr a n s-4-(3,5-Din itr op h en yl)bu t-2-en yl Aceta te (2j). The
reaction was performed according to the general procedure on
a 1.0 mmol scale. Pyridine (79 µL, 1 mmol) was added to the
reaction mixture before the addition of 3,5-dinitroaniline. The
(39) Chatgilialoglu, C.; Ballestri, M.; Vecchi, D. Tetrahedron Lett.
1996, 37, 6383-6386.
(40) Tutorial found in the software of Varian Star Chromatography
workstation version 4.51.
1916 J . Org. Chem., Vol. 68, No. 5, 2003

Aromatic Allylation via Diazotization
remaining allyl bromide was removed by distillation at reduced
141.3, 136.3, 135.7, 132.2, 120.9, 118.6, 116.5, 115.2, 38.1, 35.7;
HRMS (FAB+) calcd for C11H10Br2N [(M + H)+] 313.9181,
found 313.9174. Anal. Calcd for C11H9Br2N: C, 41.94; H, 2.88;
N, 4.45. Found: C, 42.04; H, 2.90; N, 4.37.
pressure. The yield of 2j was determined to be 43% according
1
to H NMR spectroscopy using toluene as internal standard.
The toluene was then removed by distillation at reduced
pressure. Flash chromatography (heptane-EtOAc 4:1) fol-
lowed by crystallization (heptane-EtOAc 4:1) of the remaining
yellow oil gave 100 mg (35%) of the title compound as pale
9-Nitr o-5-br om om eth yl-5,6-d ih yd r otetr a zolo[5,1-a ]iso-
qu in olin e (15a ). The reaction was performed according to the
general procedure using 11 (103 mg, 0.50 mmol) as starting
material and 10 equiv of allyl bromide in dry CH3CN (1 mL).
Column chromatography (heptane-EtOAc 1:1) gave 34 mg
(22%) of the title compound as pale yellow crystals (90-95%
1
yellow crystals: mp 57-58 °C; H NMR (CDCl3, 400 MHz) δ
8.91, (t, 1H, J ) 2.1 Hz), 8.38, (d, 2H, J ) 2.1 Hz), 5.96-5.87,
(dtt, 1H, J ) 15.4, 6.7, 1.3 Hz), 5.80-5.72, (dtt, 1H, J ) 15.4,
6.0, 1.4 Hz), 4.60, (dd, 2H, J ) 6.0, 1.1 Hz), 3.64, (d, 2H, J )
6.7 Hz), 2.08, (s, 3H); 13C NMR (CDCl3, 100 MHz) δ 170.9,
148.8, 144.3, 130.6, 129.0, 128.8, 117.3, 64.3, 38.2, 21.1. Anal.
Calcd for C12H12N2O6: C, 51.43; H, 4.32; N, 10.00. Found: C,
51.49; H, 4.37; N, 9.98.
1
purity according to HPLC and H NMR analyses). An analyti-
cally pure sample was prepared by preparative HPLC (CH3-
1
CN-H2O 60:40): mp 143-144 °C; H NMR (acetone-d6, 400
MHz) δ 8.77, (dd, 1H, J ) 2.4, 0.4 Hz), 8.42, (dd, 1H, J ) 8.5,
2.4 Hz), 7.89, (br d, 1H, J ) 8.0 Hz), 5.51, (m, 1H), 4.22, (dd,
1H, J ) 11.4, 5.2 Hz), 4.13, (dd, 11.4, 4.2 Hz), 3.97-3.88, (ddt,
1H, J ) 17.3, 6.6, 0.5 Hz), 3.80-3.72, (dd, 1H, J ) 17.3, 6.9
Hz); 13C NMR (acetone-d6, 100 MHz) δ 150.7, 148.5, 142.0,
131.5, 127.1, 123.3, 120.5, 55.8, 33.9, 33.2. Anal. Calcd for
1-Allen yl-3,5-d in itr oben zen e (2k ). The reaction was per-
formed according to the general procedure. Column chroma-
tography (heptane-EtOAc 23:2) followed by crystallization
(heptane-EtOAc 23:2) gave 145 mg (23%) of the title com-
pound as pale yellow crystals: mp 122-124 °C; 1H NMR
(CDCl3, 400 MHz) δ 8.85, (t, 1H, J ) 2.1 Hz), 8.44, (d, 2H, J
) 2.1 Hz), 6.34, (t, 1H, J ) 6.7 Hz), 5.43, (d, 2H, J ) 6.6 Hz);
13C NMR (CDCl3, 100 MHz) δ 211.3, 149.3, 139.3, 126.5, 117.0,
92.4, 81.9. Anal. Calcd for C9H6N2O4: C, 52.43; H, 2.93; N,
13.59. Found: C, 52.46; H, 2.85; N, 13.49.
5-Br om o-2-iod oben zon itr ile (7). 2-Amino-5-bromoben-
zonitrile (296 mg, 1.5 mmol) was added during 10 min to tert-
butyl nitrite (0.36 mL, 2.3 mmol) and iodine (1.1 g, 4.5 mmol)
in dry CH3CN (2 mL) under argon atmosphere while main-
taining the temperature between 30 and 35 °C. Stirring was
continued for 60 min at 23 °C. Subsequent addition of
saturated aqueous Na2SO3 (20 mL) gave a precipitate which
was collected by filtration. Recrystallization from heptane gave
397 mg (86%) of the title compound as off-white crystals: mp
123-124 °C (lit.41 mp 113-114 °C42).
C
10H8BrN5O2: C, 38.73; H, 2.60; N, 22.58. Found: C, 38.88;
H, 2.53; N, 22.48.
9-Nitr o-5,5-bis(br om om eth yl)-5,6-dih ydr otetr a zolo[5,1-
a ]isoqu in olin e (15b). The reaction was performed according
to the general procedure using 11 (103 mg, 0.50 mmol) as
starting material and 10 equiv of 3-bromo-2-bromomethylpro-
pene in dry CH3CN (1 mL). Column chromatography (hep-
tane-EtOAc 1:1) gave 41 mg (20%) of the title compound as
pale yellow crystals (90-95% purity according to HPLC and
1H NMR analyses). An analytically pure sample was prepared
by preparative HPLC (CH3CN-H2O 65:35): mp 179-180 °C;
1H NMR (acetone-d6, 400 MHz) δ 8.79, (dd, 1H, J ) 2.4, 0.5
Hz), 8.43, (dd, 1H, J ) 8.5, 2.4 Hz), 7.91, (br d, 1H, J ) 8.5,
0.5 Hz), 4.34, (d, 2H, J ) 11.4 Hz), 4.25, (d, 11.4 Hz), 4.03, (br
s, 2H); 13C NMR (acetone-d6, 100 MHz) δ 150.8, 148.7, 141.2,
131.7, 127.5, 122.6, 120.8, 63.8, 37.6, 36.7. Anal. Calcd for
2-Allyl-5-br om oben zon itr ile (9). The reaction was per-
formed according to the general procedure using 10 (590 mg,
3.0 mmol) as starting material. The temperature during the
addition was 30-35 °C, and the temperature after the addition
was kept at 26 °C. Column chromatography (heptane-EtOAc
49:1) gave 524 mg of 13 as a clear oil containing 8% of 2,5-
dibromobenzonitrile. This corresponds to a 72% yield of 13.
An analytically pure sample was obtained by preparative
HPLC (CH3CN-H2O 65:35): 1H NMR (CDCl3, 400 MHz) δ
7.74, (d, 1H, J ) 2.1 Hz), 7.64, (dd, 1H, J ) 8.4, 2.1 Hz), 7.22,
(dd, 1H, J ) 8.4, 0.5 Hz), 5,96-5.86, (m, 1H), 5.19-5.09, (m,
2H), 3.56, (dt, 2H, J ) 6.6, 1.2 Hz); 13C NMR (CDCl3, 100 MHz)
δ 143.0, 136.2, 135.3, 134.4, 131.5, 120.2, 118.1, 116.6, 114.5,
38.2. Anal. Calcd for C10H8BrN: C, 54.08; H, 3.63; N, 6.31.
Found: C, 54.06; H, 3.62; N, 6.38.
5-Br om o-2-(2-b r om om et h yl-2-p r op en yl)b en zon it r ile
(10). The reaction was performed according to the general
procedure starting from 10 (99 mg, 0.50 mmol) and using 10
equiv of 3-bromo-2-bromomethylpropene in CH3CN (1 mL).
The addition temperature and the temperature after addition
were 30-35 and 24 °C, respectively. The remaining allylic
bromide was removed by distillation at reduced pressure (0.7
mmHg, 30 °C). The yield of the title compound was determined
to be 82% according to 1H NMR analysis using toluene as
internal standard. The toluene was then removed by distilla-
tion at reduced pressure. Flash chromatography (heptane-
EtOAc 95:5) followed by recrystallization (heptane-diethyl
ether 10:1) provided the title compound (100 mg, 62%, Rf )
0.31)43 as white crystals: mp 45-46 °C; 1H NMR (CDCl3, 400
MHz) δ 7.78, (d, 1H, J ) 2.1 Hz), 7.68, (dd, 1H, J ) 8.3, 2.1
Hz), 7.29, (d, 1H, J ) 8.3 Hz), 5.34, (br s, 1H), 4.91, (br s, 1H),
3.92, (s, 2H), 3.73, (s, 2H); 13C NMR (CDCl3, 100 MHz) δ 142.4,
C
11H9Br2N5O2: C, 32.78; H, 2.25; N, 17.38. Found: C, 32.89;
H, 2.18; N, 17.26.
9-Nitr o-5-br om om eth yl-5-m eth yl-5,6-d ih yd r otetr a zolo-
[5,1-a ]isoqu in olin e (15c). The reaction was performed ac-
cording to the general procedure using 11 (103 mg, 0.50 mmol)
as starting material and 10 equiv of methallyl bromide in dry
CH3CN (1 mL). Column chromatography (heptane-EtOAc 1:1)
gave 25 mg (15%) of the title compound as pale yellow crystals
(90% purity according to HPLC and 1H NMR analyses). An
analytically pure sample was prepared by preparative HPLC
1
(CH3CN-H2O 60:40): mp 121-123 °C; H NMR (acetone-d6,
400 MHz) δ 8.78, (dd, 1H, J ) 2.4, 0.4 Hz), 8.43, (dd, 1H, J )
8.4, 2.4 Hz), 7.88, (dd, 1H, J ) 8.4, 0.4 Hz), 4.10, (d, 1H, J )
11.3 Hz), 4.01, (d, 11.3 Hz), 3.90, (d, 1H, J ) 17.2 Hz), 3.75,
(d, 1H, J ) 17.2 Hz), 2.03,(s, 3H); 13C NMR (acetone-d6, 100
MHz) δ 150.2, 148.6, 141.7, 131.6, 127.2, 123.2, 120.6, 61.7,
39.7, 39.1, 24.7. Anal. Calcd for C11H10BrN5O2: C, 40.76; H,
3.11; N, 21.61. Found: C, 40.89; H, 3.12; N, 21.46.
3-Br om om eth yl-7-n itr o-2,3-d ih yd r oisocou m a r in (17).
The reaction was performed according to the general procedure
using 16 (910 mg, 5.0 mmol) as starting material and 10 equiv
of allyl bromide in dry CH3CN (10 mL). The yield of 17 was
1
determined to be 39% according to H NMR using toluene as
internal standard. The toluene was then removed by distilla-
tion at reduced pressure. Column chromatography (heptane-
EtOAc 3:1) followed by crystallization (heptane-EtOAc 3:1)
gave 460 mg (32%) of the title compound as pale yellow
crystals: mp 92-93 °C; 1H NMR (CDCl3, 400 MHz) δ 8.95, (d,
1H, J ) 2.4 Hz), 8.42, (dd, 1H, J ) 8.4, 2.4 Hz), 7.53, (d, 1H,
J ) 8.4 Hz), 4.80, (m, 1H), 3.72, (dd, 1H, J ) 11.0, 4.3 Hz),
3.63, (dd, 1H, J ) 11.0, 6.8 Hz), 3.31, (d, 2H, J ) 7.0 Hz); 13
C
NMR (CDCl3, 100 MHz) δ 162.3, 148.1, 144.5, 129.4, 128.5,
(41) Gray, G. W.; Lacey, D.; Hird, M.; Toyne, K. J . Laterally cyano-
and fluoro-substituted terphenyls and liquid-crystal mixtures contain-
ing them. WO 8903821, 1989, 53 pp.
(42) Melting point of the crude product. NMR data can be found in
the Supporting Information.
(43) As seen in the Experimental Section, a large difference was
detected between the yield using internal standard and the isolated
yield. This is probably due to deterioration of the product during
column chromatography.
J . Org. Chem, Vol. 68, No. 5, 2003 1917

Ek et al.
126.2, 125.8, 76.6, 32.1, 31.8. Anal. Calcd for C10H8BrNO4: C,
41.98; H, 2.82; N, 4.90. Found: C, 42.08; H, 2.86; N, 4.79.
4-Hyd r oxy-8-n itr o-2,3,4,5-tetr a h yd r oben zo[c]a zep in -1-
on e (18). Isocoumarin 17 (145 mg, 0.5 mmol), dissolved in dry
MeOH (5 mL), was added dropwise to dry MeOH (10 mL)
saturated with NH3(g) during 15 min at 0 °C. The pale yellow
solution was then stirred for 30 min at 0 °C and 51 h at
23-25 °C. Removal of the MeOH-NH3 at reduced pressure
gave a yellow residue that was partly dissolved again in MeOH
(20 mL) and NH4OH (2 mL, 25%). Silica was added to the
mixture, and the volatile material was removed at reduced
pressure. Column chromatography (CHCl3-MeOH-NH4OH
(25%) 9:1:0.1) of the yellow residue followed by rinsing the
obtained pale yellow solid with heptane-EtOAc (1:1) gave 105
mg (95%) of the title compound as a white solid: mp 228-229
cooled in an ice bath. Careful addition of methanol destroyed
the remaining borane, and the volatile material in the reaction
mixture was removed at reduced pressure. The remaining pale
yellow oil was dissolved in MeOH, and then HCl (3 mL, 6 M)
was added to this solution. The solution was refluxed for 2 h,
and the MeOH was removed at reduced pressure. The remain-
ing aqueous phase was made basic (pH >10) with NaOH (2
M) and extracted with CH2Cl2 (5 × 10 mL). The combined
organic phases were dried over anhydrous potassium carbon-
ate. Removal of CH2Cl2 at reduced pressure gave 46 mg of 19
as a yellow solid that was dissolved in acetone. Subsequent
addition of diethyl ether precipitated 40 mg (77%) of the
title compound as a pale yellow solid: mp 150-151 °C (lit.31
147-149 °C).
1
Ack n ow led gm en t. We thank Amersham Health
R&D AB, the Swedish Research Council, and The Royal
Physiographic Society in Lund for financial support. The
Dead Sea Bromine Group is gratefully acknowledged
for a free sample of tribromoneopentyl alcohol, and we
thank Mr. J an Glans for help with preparative HPLC.
°C; H NMR (DMSO-d6, 400 MHz) δ 8.40, (m, 1H), 8.25, (m,
2H), 7.55, (m, 1H), 5.21, (d, 1H, J ) 4.2 Hz), 4.18, (m, 1H),
3.12-3.00, (m, 2H), 2.81-2.68, (m, 2H); 13C NMR (DMSO-d6,
100 MHz) δ 169.4, 146.6, 143.9, 137.0, 131.3, 124.9, 123.1, 71.2,
46.0.44 Anal. Calcd for C10H10N2O4: C, 54.05; H, 4.54; N, 12.61.
Found: C, 53.89; H, 4.68; N, 12.55.
4-H yd r oxy-8-n it r o-2,3,4,5-t e t r a h yd r o-1H -b e n zo[c]-
a zep in e (19). Solid 18 was added in portions to BH3‚THF (1
mL, 1 M) in THF (5 mL) at 0 °C under argon atmosphere.
The clear solution was refluxed under argon for 24 h and then
Su p p or tin g In for m a tion Ava ila ble: The synthesis of 9
and 10 using the magnesium exchange reaction and the
synthesis of 11. NMR data for 18 in CD3OD and 7 in CDCl3.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(44) One of the signals in the 13C-spectrum of 18 is coinciding with
the solvent signals (DMSO-d6). NMR data (CD3OD) of 18 can be found
in the Supporting Information.
J O026784B
1918 J . Org. Chem., Vol. 68, No. 5, 2003