Aromatic allylation via diazotization: Metal-free C-C bond formation

Article abstract of DOI:10.1021/jo0258103

A new method for the synthesis of allyl aromatic compounds not involving any metal-containing reagent or catalyst has been developed. Arylamines substituted with a large number of different substituents were converted via diazotizative deamination with tert-butyl nitrite in allyl bromide and acetonitrile to the corresponding allyl aromatic compounds. The allylation reaction was found to be suitable for larger scale synthesis due to short reaction times, a nonextractive workup, and robustness toward moisture, air, and type of solvent.

Full text of DOI:10.1021/jo0258103

Ar om a tic Allyla tion via Dia zotiza tion : Meta l-F r ee C-C Bon d
F or m a tion
Fredrik Ek,† Oskar Axelsson,‡ 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 April 11, 2002
A new method for the synthesis of allyl aromatic compounds not involving any metal-containing
reagent or catalyst has been developed. Arylamines substituted with a large number of different
substituents were converted via diazotizative deamination with tert-butyl nitrite in allyl bromide
and acetonitrile to the corresponding allyl aromatic compounds. The allylation reaction was found
to be suitable for larger scale synthesis due to short reaction times, a nonextractive workup, and
robustness toward moisture, air, and type of solvent.
SCHEME 1. Syn th etic Rou te tow a r d Ca n d id a te
Com p ou n d s Sta r tin g fr om
In tr od u ction
Allyl-3,5-d in itr oben zen ea
In our search for a new iodinated X-ray contrast agent,
we found a group of compounds (3) with promising
physiological and chemical properties (Scheme 1).1 The
substances were all synthesized from the same starting
material, allyl-3,5-dinitrobenzene (1h , Scheme 1). Initial
attempts to use the Stille cross-coupling reaction between
3,5-dinitroiodobenzene and allyltributyltin only gave 20%
of 1h after 7 days reaction time and workup involving
preparative HPLC.1 Furthermore, the sensitive nature
of the electron-deficient aromatic moiety toward nucleo-
philic and reductive reagents prevented the synthesis of
the diol 2 (Scheme 1) by other methods not involving 1h .2
To continue the project, we needed a new method suitable
for synthesizing 1h at a molar scale.
a
See ref 1 for the structure of R1 and R2.
have been developed and extensively used for the direct
allylation of aromatic compounds. Despite their versatil-
ity their use in organic synthesis is somewhat limited,
especially at a larger scale. Several of the methods use
highly toxic catalysts or reagents, which makes the
reactions environmentally problematic. In certain cases,
poor regioselectivity or lack of functional group tolerance
due to highly reactive reagents constitute major prob-
lems.
Free-radical allylations provide some of the mildest,
most general methods to introduce allyl groups into
functionalized molecules. Most research in this area has
been focused on allylation of sp3-hybridized carbons,13-18
and only a few papers have demonstrated radical ally-
lations of aromatic compounds.19-23 Migita et al. were the
Several methods such as Claisen rearrangement of
allyl phenyl ethers, cross-coupling catalyzed by transition
metals,3-9 allylation with reactive organometallic re-
agents,5,10 and electrophilic aromatic substitution11,12
† Lund University.
‡ Amersham Health R&D AB.
(1) 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.
(2) Dalla, V.; Cotelle, P.; Catteau, J . P. Tetrahedron Lett. 1997, 38,
1577-1580.
(3) J olly, P. W. In Comprehensive Organometallic Chemistry; Wilkin-
son, G., Stone, G., Eds.; Pergamon: Oxford, 1982; Vol. 8, pp 713-772.
(4) Trost, B. M.; Verhoeven, T. R. In Comprehensive Organometallic
Chemistry; Wilkinson, G., Stone, G., Eds.; Pergamon: Oxford, 1982;
Vol. 8, pp 799-938.
(5) Larock, R. C. Synthetic Organic Methodology: Comprehensive
Organic Transformations. A Guide to Functional Group Preparations;
1989.
(6) Lee, P. H.; Sung, S.-y.; Lee, K. Org. Lett. 2001, 3, 3201-3204.
(7) Mowery, M. E.; DeShong, P. J . Org. Chem. 1999, 64, 1684-1688.
(8) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457-83.
(9) Stille, J . K. Angew. Chem. 1986, 98, 504-19.
(10) Boudier, A.; Bromm, L. O.; Lotz, M.; Knochel, P. Angew. Chem.,
Int. Ed. 2000, 39, 4415-4435.
(11) Kodomari, M.; Nawa, S.; Miyoshi, T. J . Chem. Soc., Chem.
Commun. 1995, 1895-6.
(12) Lim, H. J .; Keum, G.; Kang, S. B.; Kim, Y.; Chung, B. Y.
Tetrahedron Lett. 1999, 40, 1547-1550.
(13) Curran, D. P., Porter, N. A., Giese, B., Eds. Stereochemistry of
Radical Reactions: Concepts, Guidelines, and Synthetic Applications,
1996.
(14) Giese, B. Radicals in Organic Synthesis: Formation of Carbon-
Carbon Bonds, 1986.
(15) Keck, G. E.; Enholm, E. J .; Yates, J . B.; Wiley, M. R. Tetrahe-
dron 1985, 41, 4079-94.
(16) Quiclet-Sire, B.; Zard, S. Z. J . Am. Chem. Soc. 1996, 118, 1209-
10.
(17) Sibi, M. P.; Chandramouli, S. V. Tetrahedron Lett. 1997, 38,
8929-8932.
(18) Sibi, M. P.; Porter, N. A. Acc. Chem. Res. 1999, 32, 163-171.
(19) Al Adel, I.; Adeoti Salami, B.; Levisalles, J .; Rudler, H. Bull.
Soc. Chim. Fr. 1976, 934-8.
10.1021/jo0258103 CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/15/2002
6376
J . Org. Chem. 2002, 67, 6376-6381

Aromatic Allylation via Diazotization
TABLE 1. Resu lts of Allyla tion w ith Allyl Br om id e
first to show that a phenyl radical, generated from
phenylazotriphenylmethane or N-nitrosoacetanilide, re-
acted with allyl sulfides or allyl bromide to afford
allylbenzene.20 Later, they also showed that aromatic
halides reacted with allyltributyltin, either via a thermal
reaction in the presence of AIBN or via a photoinduced
reaction, which gave allylbenzene.22 Allyl radicals, gener-
ated in situ from allyl iodide, have also been used for
direct allylation of aromatic compounds, although with-
out any regioisomeric control.24
selectivity
yieldb (%) 1a -p /5a -p c
entry
R1, R2
Ta (°C)
Since the radical reactions mentioned all involve
hazardous reagents or lack of regioselectivity, they were
unsuitable for our large-scale application. We therefore
decided to develop an alternative method suitable for our
purposes. Crucial information for the development of a
new allylation reaction was presented in a paper describ-
ing a modified procedure for the Meerwein reaction.25 In
this paper, phenyl radicals were generated in situ from
the corresponding arylamine, via deamination with an
alkyl nitrite, which in turn reacted with acrylonitrile (and
other conjugated olefins). Surprisingly, we could not find
any applications of this convenient procedure of phenyl
radical generation from an arylamine in combination
with allyl bromide, although, as mentioned, phenyl
radicals react with allyl bromide to give allyl aromatic
compounds. In this paper, we report a new method for
the synthesis of 1h also allowing the synthesis of a large
number of different allyl-substituted aromatics. We here
focus on structural variation of the aryl moiety and have
chosen arylamines preferably substituted with one or two
nitro groups because of their usefulness in the synthesis
of contrast agents and also because of the difficulties
often encountered in the preparation of this type of allyl
aromatic compounds utilizing existing methods.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
2-Cl, 4-NO2
2-CF3, 4-NO2
2-CN, 4-NO2
11-13/23
18-19/26
23-27/25
79 (1a )
85 (1b)
65 (1c)
46 (1d )
58 (1e)
40 (1f)
62 (1g)
25:1
35:1
17:1
7:1
14:1
10:1
14:1
25:1
5:1
12:1
6:1
3:1
2:1
2:1
2-COPh, 4-NO2 28-32/25
2-Br, 5-NO2
3-NO2, 4-OMe
3-CF3, 4-NO2
3,5-diNO2
40-45/45
35-40/40
13-15/25
11.5-13/22 81, 89d (1h )
3,5-diCl
3-COOEt, 5-NO2 10-12/25
4-NO2, H
30-35/30
48e (1i)
67 (1j)
55 (1k )
34 (1l)
19 (1m )
18f (1n )
0 (1o)
48-52/50
30-35/30
45-46/50
60-61/61
>60
4-Br, H
4-tert-butyl, H
4-OMe, H
2,6-diNO2
2,4-diNO2
60-61/60
49g (1p )
9:1
a
Temperature during the addition of the arylamines/temper-
b
ature after the addition. Isolated yields except for entry 14.
c Determined by HPLC and 1H NMR of the crude products.
Multigram scale synthesis. e 19% of 4-bromo-3,5-dichloroallyl-
d
benzene and 3% of 2,5-dibromo-1,4-dichlorobenzene were also
formed in the reaction. f Determined by 1H NMR using toluene
as internal standard. 1H NMR data were in agreement with
literature data.27 Extended reaction time (18 h) and additional
g
tert-butyl nitrite (8 equiv).
Subsequent removal of volatile compounds from the
reaction mixture at reduced pressure gave the crude
product, which was then purified by methods outlined
in the Experimental Section. A slightly modified proce-
dure was used for the synthesis of 1h on a multigram
scale (for details see the Experimental Section).
Resu lts a n d Discu ssion
The reactions presented in Table 1 were performed on
a 3 mmol scale. The arylamine (solid) was added in
portions to a dry and oxygen-free acetonitrile solution of
tert-butyl nitrite and allyl bromide during approximately
20 min, while the temperature was maintained as shown
in Table 1.26 Wa r n in g! A too r a p id a d d ition of th e
a r yla m in e to th e r ea ction 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 n itr ogen ga s.
A reaction time of 60 min was employed, although in
several cases complete conversion was achieved already
ca. 5 min after the final addition of the arylamines. An
extractive workup was not necessary, in contrast to the
methods using metal-containing reagents and catalysts.
As seen in Table 1, allylated products were isolated in
yields ranging from 0 to 85% on the basis of the
corresponding arylamines.28 In accordance with the typi-
cal behavior of radical reactions, a large number of
functionalities were tolerated.13,14 Electron-deficient ary-
lamines generally had a higher reactivity and gave higher
yield of the allylated products. A similar electronic effect
on reactivity has been reported in radical addition
reactions involving allyl, vinyl, and aryl substrates and
also in radical abstraction of hydrogen or halogen at-
oms.21,25,29,30 The common explanation for this behavior
is that electron-withdrawing groups convert the phenyl
radical (a probable intermediate in the mechanism of the
allylation reaction, which will be discussed later) from a
relatively nonpolar to an electrophilic radical.21,30-32
(20) Migita, T.; Kosugi, M.; Takayama, K.; Nakagawa, Y. Tetrahe-
dron 1973, 29, 51-5.
(21) Migita, T.; Takayama, K.; Abe, Y.; Kosugi, M. J . Chem. Soc.,
Perkin Trans. 2 1979, 1137-42.
(22) Migita, T.; Nagai, K.; Kosugi, M. Bull. Chem. Soc. J pn. 1983,
56, 2480-4.
(27) Pouchert, C. J .; Behnke, J . In The Aldrich Library of 13C and
1H FT NMR-spectra, 1st ed.; Aldrich: Milwaukee, WI, 1993; Vol. 2, p
208.
(28) Additional examples can be found in ref 47.
(29) Tilset, M.; Parker, V. D. Acta Chem. Scand., Ser. B 1982, B36,
123-4.
(30) Takayama, K.; Kosugi, M.; Migita, T. Chem. Lett. 1973, 193-
5.
(31) Tichy, S. E.; Thoen, K. K.; Price, J . M.; Ferra, J . J ., J r.; Petucci,
C. J .; Kenttaemaa, H. I. J . Org. Chem. 2001, 66, 2726-2733.
(32) Heidbrink, J . L.; Thoen, K. K.; Kenttaemaa, H. I. J . Org. Chem.
2000, 65, 645-651.
(23) Yashkina, L. V.; Kopylova, B. V. Izv. Akad. Nauk SSSR, Ser.
Khim. 1989, 957-9.
(24) Camaggi, C. M.; Leardini, R.; Zanirato, P. J . Org. Chem. 1977,
42, 1570-2.
(25) Doyle, M. P.; Siegfried, B.; Elliott, R. C.; Dellaria, J . F., J r. J .
Org. Chem. 1977, 42, 2431-6.
(26) All experiments made in order to find the optimal reaction
conditions for the allylation (including safety consideration) were
performed at 0.5 mmol scale. The yields and selectivity were deter-
mined by HPLC using 2-nitrobenzyl alcohol as an internal standard
(see the Experimental Section).
J . Org. Chem, Vol. 67, No. 18, 2002 6377

Ek et al.
Similar to the behavior of other radical addition
reactions, steric hindrance has an adverse effect on the
outcome of the allylation reaction.33-35 When 1o was
subjected to the standard conditions no evolution of
nitrogen gas could be detected. Even when the temper-
ature was increased to >60 °C, the corresponding ally-
lated product could not be isolated. Also, 4p was more
unreactive than 4h . According to HPLC analysis, only
20% of 4p was consumed after 1 h when the allylation
reaction was performed at 60 °C. However, extended
reaction time (18 h) and addition of extra tert-butyl nitrite
(8 equiv was added in portions during 4 h) gave almost
complete conversion of 4p , although the isolated yield of
1p was only 49%. Similarly, the steric bulk of the benzoyl
substituent in 4d is probably the reason for the moderate
yield of the allylated product 1d .
periments indicated that the phenyl radical attacks both
allylic components indiscriminately and that only attack
on the bromide leads to allylation.
In our standard method, a large amount (15 equiv) of
allyl bromide was used. The yield of the desired allylation
product decreased considerably at lower than 7.5 equiv
of allyl bromide. However, most of the excess of allyl
bromide could be recovered by distillation directly from
the reaction mixture; the distilled allyl bromide contained
acetonitrile and traces of tert-butyl nitrite. Thus, ap-
proximately 10-11 equiv of allyl bromide could be
isolated and reused after the reaction, indicating that
4-5 equiv was consumed. We also noted that while the
yield of the allylation product was quite dependent on
the amount of allyl bromide the yield of the side product
5h was only marginally effected ((1%).
Electron-withdrawing substituents not only facilitated
the allylation reaction but also made the corresponding
phenyl radical more selective for the addition to allyl
bromide relative to abstraction of a bromine atom, the
most important side reaction. As seen in Table 1, the
more electron-deficient arylamines generally gave a
higher ratio of allylation versus bromination. Addition-
ally, steric hindrance appeared to increase the tendency
of bromination as exemplified in entries 4 and 16 (Table
1).
Finally, a method for the large-scale (0.41 mol) syn-
thesis of 1h was developed. A smaller amount of aceto-
nitrile was used in order to limit the total volume of the
reaction mixture. The addition time of the arylamine had
to be prolonged because of the exothermicity of the
reaction. An increased yield compared to the same
reaction at a 3 mmol scale was noticed. Filtration of the
crude product dissolved in toluene through two pads of
neutral alumina gave a product that was pure enough
to be crystallized from isooctane. Thus, chromatography,
which would not be acceptable in a larger scale procedure,
was avoided.
Changes in the reaction conditions such as concentra-
tion, moisture levels, and temperature only moderately
affected the yield (66-79%) and the selectivity (20-25:
1).26 The reaction proceeded even in allyl bromide as
solvent with minor changes in the outcome (yield 70%,
selectivity 15:1). Unexpectedly, the replacement of tert-
butyl nitrite with isoamyl nitrite lowered the yield with
11%. A possible explanation could be the greater ten-
dency of homolytic cleavage of the secondary carbon-
hydrogen bonds in the isoamyl unit.36 Further, to see if
the phenyl radical was able to select between allylic
substrates with and without a suitable terminal leaving
group, equal amounts allyl bromide and allyl acetate
were employed. (OAc is considered to be a poor leaving
group in radical reactions due to the strong C-O bond).20
The yield dropped from 74% to 45%. When only allyl
acetate was used in the reaction the allyl product could
not be detected. Instead, a multitude of nonidentified
compounds was formed according to HPLC. These ex-
In agreement with the general behavior of radical
reactions, we found only small variations in the product
distribution when changing the solvent. Acetonitrile
(74%, 25:1) and nitromethane (75%, 20:1) were found to
be the best solvents, although both acetone and dimethox-
yethane could be applied with minor alteration of the
yield and selectivity (69%, 20:1). DMSO and tetrachlo-
romethane gave somewhat lower yield (55-60%) and
showed a poorer selectivity (10:1). THF gave the same
yield of the allylated product as acetone but resulted in
the lowest ratio of allylation versus bromination (8:1).
The mechanism of the aprotic diazotization of ary-
lamines is not yet fully understood. However, we believe
that the mechanism for the Gomberg-Bachmann (GB)
reaction, proposed by Ru¨chhardt et al.,37 is probably also
applicable to aprotic diazotization of arylamines, includ-
ing the allylation reaction (Scheme 2). We have been able
to isolate a reactive precipitate from the reaction mixture
formed during the addition of 4h . It slowly dissolved
when the addition of 4h was completed.38 Analysis of the
isolated polar precipitate (Wa r n in g! Th e p r ecip ita te
violen t ly d ecom p osed u p on h ea t in g), with TOF
HRMS (APCI-) (acetonitrile) and 1H NMR (DMSO-d6)
suggested that the material was similar to 10, thus
supporting the GB mechanism. This material is likely
to be in a pH-dependent equilibrium with the corre-
sponding diazonium salt 9 and the diazotate 11.39,40
Furthermore, addition of the precipitate to allyl bromide
in acetonitrile resulted in the evolution of nitrogen gas
and formation of 1h and the byproducts normally found
in the allylation reaction. Similar to the mechanism
postulated for the GB reaction, 9 and 11 then form the
diazotic anhydride (12) which, in turn, decomposes and
gives nitrogen gas, phenyl radical (14), and the long-lived
diazotyl radical (13).41,42 The diazotyl radical probably
abstracts a hydrogen atom or an electron from the
reaction mixture, thus regenerating the diazotic acid/
diazotate system (9-11). There is also a possibility that
(37) Ruechardt, C.; Merz, E. Tetrahedron Lett. 1964, 2431-6.
(38) A precipitate was also formed with other arylamines but were
not analyzed as thoroughly as the one from 4h .
(39) Machackova, O.; Sterba, V. Collect. Czech. Chem. Commun.
1972, 37, 3313-27.
(33) Colombani, D. Prog. Polym. Sci. 1999, 24, 425-480.
(34) Mella, M.; Fagnoni, M.; Albini, A. J . Org. Chem. 1994, 59,
5614-22.
(35) Zarkadis, A. K.; Neumann, W. P.; Duennebacke, D.; Penenory,
A.; Stapel, R.; Stewen, U. Chem. Ber. 1993, 126, 1179-85.
(36) Kopinke, F. D.; Zimmermann, G.; Anders, K. J . Org. Chem.
1989, 54, 3571-6.
(40) Kuokkanen, T. Finn. Chem. Lett. 1987, 14, 184-92.
(41) Binsch, G.; Ruechardt, C. J . Am. Chem. Soc. 1966, 88, 173-4.
(42) Binsch, G.; Merz, E.; Ruechardt, C. Chem. Ber. 1967, 100, 247-
58.
6378 J . Org. Chem., Vol. 67, No. 18, 2002

Aromatic Allylation via Diazotization
SCHEME 2. Su ggested Mech a n ism of th e
Allyla tion Rea ction
SCHEME 3. Alter n a tive Mech a n ism for th e
F or m a tion of P h en yl Ra d ica ls
which was indeed found in the reaction mixture in
substantial amounts (a control using authentic material
was made).44
In an alternative mechanism suggested for aprotic
diazotization of arylamines with alkyl nitrite, a diazotic
ester (16) has been suggested as the precursor of phenyl
radicals (Scheme 3).25,45 However, the assumed less polar
character of 16 (in our case substituted with two nitro
groups) compared to the polar nature of the precipitate
together with the fact that the diazotic ester was not
1
detected by MS nor was the H NMR analysis consistent
with the diazotic ester as the precipitate makes this
alternative mechanism less probable.
In conclusion, a new method for the synthesis of allyl
aromatic compounds has been developed. Arylamines
substituted with a large number of different substituents
gave the corresponding allyl aromatic compounds with
complete regioselectivity. Electron-withdrawing substit-
uents were necessary to obtain good yields. The allylation
reaction was relatively insensitive toward moisture, air,
and type of solvent. The robustness of the reaction
combined with nonextractive workup, reaction temper-
ature above 0 °C and short reaction times makes it
suitable for larger scale synthesis. This allylation proce-
dure constitutes a carbon-carbon bond forming reaction
without the need for metal-containing reagents and
catalysts, thus avoiding the problem of residual toxic
metals in compounds for biological testing. Furthermore,
the large number of commercially available aniline
derivatives makes it possible to synthesize a broad range
of allyl aromatic compounds.
the diazotyl radical (13) may form a dimer, which then
could decompose into nitrogen, oxygen, and two phenyl
radicals.
The phenyl radical (14) reacts with the allyl bromide,
either in a concerted or a stepwise mechanism (via 15),
resulting in the formation of an allyl aromatic compound
and a bromine atom. The bromine atom then continues
to react with the excess of the allyl bromide resulting in
partial polymerization. In fact, purification of the crude
product left a fraction that contained the oligomerized
allyl bromide. Analysis of the 1H NMR spectrum and
mass spectrum of this fraction revealed a complex
mixture of compounds, substituted with up to three
bromine atoms.
The phenyl radical 14 may also abstract a bromine
atom from allyl bromide forming the corresponding
bromobenzene derivative (5h ). As mentioned, the bro-
mine abstraction often constitutes the major side reaction
in this process. The halogen abstraction, which is a
unique behavior of radicals, strongly indicates that
phenyl radicals are indeed formed in the allylation
reaction.43 An alternative bromine atom source would be
Br2, which could be formed by dimerization of bromine
atoms. This would result in, e.g., 1,2,3-tribromopropane,
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);
eluent: 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: CH3-
CN (HPLC grade)/H2O; flow rate 20 mL/min. NMR spectra
were recorded at 400 MHz using CDCl3 or acetone-d6 as
internal standard. Mass spectra were recorded in the following
modes: EI(+) (70 eV) using both direct inlet and inlet via a
gas chromatograph equipped with a CP-SIL 8CB column,
FAB(+) and APCI(-) (-25 V, source temperature 120 °C and
probe temperature of 500 °C). Chromatographic separations
were performed on Matrex Amicon normal phase silica gel 60
(0.035-0.070 mm). TLC 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. Ethyl 3-amino-5-nitrobenzoate was synthesized
according to a literature procedure, mp 159-161 °C (lit. mp
(44) We appreciate the suggestion by the reviewer.
(45) Doyle, M. P.; Dellaria, J . F., J r.; Siegfried, B.; Bishop, S. W. J .
Org. Chem. 1977, 42, 3494-8.
(43) Wassmundt, F. W.; Kiesman, W. F. J . Org. Chem. 1997, 62,
8304-8308.
J . Org. Chem, Vol. 67, No. 18, 2002 6379

Ek et al.
158-160 °C).46 tert-Butyl nitrite and allyl bromide were
distilled prior to use. Reagents and solvents used in the large-
scale synthesis of allyl-3,5-dinitrobenzene were used as re-
ceived. 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 presented in Table
1 was distilled under argon atmosphere from CaH2 and then
degassed for 30 min with argon prior to use. In retrospect, this
was probably unnecessary. Spectral data for the allyl aromatic
compounds were in agreement with spectral data published
if not otherwise stated.47
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 allyl bromide (3.9 mL, 45.0
mmol) in dry and degassed CH3CN (3 mL) under argon
atmosphere while maintaining the specified temperature
(Table 1). At the end of the addition of arylamine, extra tert-
butyl nitrite (180 µL, 1.5 mmol) was added. The reaction
mixture was then stirred at a temperature specified in Table
1 for 1 h. The volatile material in the reaction mixture was
then removed at reduced pressure.
117.8, 37.9; IR (neat) 1670, 1526, 1348 cm-1; HRMS (FAB+)
calcd for C16H14NO3 (M + H) 268.0974, found 268.0964.
Allyl-2-br om o-5-n itr oben zen e (1e). Column chromatog-
raphy (heptane-ethyl acetate 49:1) gave 510 mg of a pale
yellow oil. The oil was dissolved in pentane, and the solution
was cooled to -20 °C upon which 1e precipitated as white
crystals. Recrystallization (pentane, room temperature to -20
°C) gave 420 mg (58%) of the title compound as white
crystals: mp 43-44 °C; 1H NMR (CDCl3, 400 MHz) δ 8.10 (d,
1H, J ) 2.7 Hz), 7.95 (dd, 1H, J ) 8.7, 2.7 Hz), 7.74 (d, 1H, J
) 8.7 Hz), 6.03-5.92 (m, 1H), 5.26-5.13 (m, 2H), 3.60 (m, 2H,
J ) 6.5 Hz); 13C NMR (CDCl3, 100 MHz) δ 147.3, 141.6, 133.7,
133.7, 131.9, 124.9, 122.5, 118.4, 40.1; IR (neat) 1526, 1346
cm-1; HRMS (EI+) calcd for C9H8BrNO2 (M) 240.9738, found
240.9739.
Allyl-4-m eth oxy-3-n itr oben zen e (1f). The crude product
was dissolved in 20 mL of ethyl acetate-heptane (1:2), and
the solution was filtered through a pad of silica. The solvent
was removed at reduced pressure. Column chromatography
(heptane-ethyl acetate 97:3) of the residue gave 230 mg (40%)
of 1f as a yellow oil: 1H NMR (CDCl3, 400 MHz) δ 7.44 (d,
1H, J ) 2.7 Hz), 7.26 (d, 1H, J ) 8.6 Hz), 7.09 (dd, 1H, J )
8.6, 2.7 Hz), 6.01-5.89 (m, 1H), 5.11-5.01 (m, 2H), 3.85 (s,
3H), 3.61 (m, 2H, J ) 6.4 Hz); 13C NMR (CDCl3, 100 MHz) δ
158.4, 149.6, 135.5, 132.8, 126.8, 119.8, 116.7, 109.2, 55.8, 36.4;
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 allyl bromide in 0.5 mL of the appropriate
solvent 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) as an internal standard. The
yield and product distribution were determined by HPLC.
IR (neat) 1533, 1352, 1252 cm-1; HRMS (EI+) calcd for C10H11
NO3 (M) 193.0739, found 193.0743.
-
Allyl-4-n itr o-3-tr iflu or om eth ylben zen e (1g). Column
chromatography (heptane-ethyl acetate 97:3) gave 430 mg
(62%) of 1g as a pale yellow oil: 1H NMR (CDCl3, 400 MHz) δ
7.86 (d, 1H, J ) 8.3 Hz), 7.65 (d, 1H, J ) 1.2 Hz), 7.55 (dd,
1H, J ) 8.3, 1.4 Hz), 6.00-5.89 (m, 1H), 5.26-5.14 (m, 2H),
3.54 (d, 2H, J ) 6.7 Hz); 13C NMR (CDCl3, 100 MHz) δ 146.1,
134.6, 132.9, 128.1 (q, J ) 5.0 Hz), 125.4, 124.0, 123.7, 123.4,
Allyl-2-ch lor o-4-n itr oben zen e (1a ). Column chromatog-
raphy (heptane-ethyl acetate 97:3) followed by preparative
HPLC (gradient solution: 50:50 to 70:30 for 30 min and then
95:5 (CH3CN/H2O)) gave 467 mg (79%) of 1a as a pale yellow
oil: 1H NMR (CDCl3, 400 MHz) δ 8.25 (d, 1H, J ) 2.4 Hz),
8.07 (dd, 1H, J ) 8.5, 2.3 Hz), 7.42 (d, 1H, J ) 8.5 Hz), 6.01-
120.7, 118.4, 39.6; IR (neat) 1541, 1360, 1315, 1144 cm-1
;
HRMS (EI+) calcd for C10H8F3NO2 (M) 231.0507, found
231.0499.
5.89 (m, 1H), 5.23-5.10 (m, 2H), 3.59 (d, 2H, J ) 6.5 Hz); 13
C
NMR (CDCl3, 100 MHz) δ 147.0, 145.4, 134.9, 133.6, 130.8,
124.6, 121.8, 118.1, 37.6; IR (neat) 1522, 1350 cm-1; HRMS
(EI+) calcd for C9H8ClNO2 (M) 197.0244, found 197.0247.
Allyl-3,5-d in itr oben zen e (1h ). Column chromatography
(heptane-ethyl acetate 23:2) gave 505 mg (81%) of 1h as a
yellow oil: 1H NMR (CDCl3, 400 MHz) δ 8.87 (t, 1H, J ) 2.1
Hz), 8.40 (d, 2H, J ) 2.1 Hz), 6.04-5.94 (m, 1H), 5.30-5.20
(m, 2H), 3.65 (d, 2H, J ) 6.7 Hz); 13C NMR (CDCl3, 100 MHz)
δ 148.6, 144.6, 134.0, 128.9, 119.2, 116.9, 39.5; IR (neat) 1541,
1344 cm-1; HRMS (EI+) calcd for C9H8N2O4 (M) 208.0484,
found 208.0486.
Allyl-4-n itr o-2-tr iflu or om eth ylben zen e (1b). Column
chromatography (heptane-ethyl acetate 49:1) gave 587 mg
(85%) of 1b as a pale yellow oil: 1H NMR (CDCl3, 400 MHz)
δ 8.52 (d, 1H, J ) 2.4 Hz), 8.34 (dd, 1H, J ) 8.5, 2.4 Hz), 7.58
(d, 1H, J ) 8.5 Hz), 6.00-5.88 (m, 1H), 5.24-5.12 (m, 2H),
3.67 (d, 2H, J ) 6.5 Hz); 13C NMR (CDCl3, 100 MHz) δ 146.4,
146.1, 134.4, 132.6, 130.0 (q, J ) 32.2 Hz), 127.2, 126.4, 124.5,
121.8, 121.7 (q, J ) 6.0 Hz), 118.3, 36.4 (m); IR (neat) 1533,
1356, 1312, 1132 cm-1; HRMS (EI+) calcd for C10H8F3NO2 (M)
231.0507, found 231.0515.
Allyl-2-cya n o-4-n itr oben zen e (1c). Column chromatog-
raphy (heptane-ethyl acetate 23:2) gave 367 mg (65%) of 1c
as a pale yellow oil: 1H NMR (CDCl3, 400 MHz) δ 8.51 (d, 1H,
J ) 2.4 Hz), 8.39 (dd, 1H, J ) 8.6, 2.4 Hz), 7.58 (dd, 1H, J )
8.6, 0.5 Hz), 6.00-5.90 (m, 1H), 5.29-5.17 (m, 2H), 3.74 (d,
2H, J ) 6.6 Hz); 13C NMR (CDCl3, 100 MHz) δ 150.9, 146.4,
133.1, 131.0, 128.0, 127.5, 119.2, 115.8, 114.0, 38.5; IR (neat)
2232, 1530, 1354 cm-1; HRMS (EI+) calcd for C10H8N2O2 (M)
188.0585, found 188.0585.
Allyl-3,5-d ich lor oben zen e (1i). Column chromatography
(pentane) followed by preparative HPLC (gradient solution:
50:50 to 80:20 for 30 min and then 95:5 (CH3CN/H2O) gave
267 mg (48%) of 1i as a colorless oil: 1H NMR (CDCl3, 400
MHz) δ 7.21 (t, 1H, J ) 1.9 Hz), 7.08 (t, 2H, J ) 1.9 Hz), 5.95-
5.84 (m, 1H), 5.16-5.08 (m, 2H), 3.34 (d, 2H, J ) 6.7 Hz); 13
C
NMR (CDCl3, 100 MHz) δ 143.4, 135.6, 134.8, 127.1, 126.4,
117.3, 39.5; IR (neat) 1587, 1568, 1431, 922, 849, 797 cm-1
;
HRMS (EI+) calcd for C9H8Cl2 (M) 186.0003, found 186.0003.
Eth yl 3-Allyl-5-n itr oben zoa te (1j). Column chromatog-
raphy (heptane-ethyl acetate 47:3) gave 470 mg (67%) of 1j
as a pale yellow oil, which crystallized upon standing overnight
in the refrigerator; melting below +20 °C: 1H NMR (CDCl3,
400 MHz) δ 8.70 (m, 1H), 8.24 (m, 1H), 8.20 (m, 1H), 6.03-
5.91 (m, 1H), 5.24-5.13 (m, 2H), 4.44 (q, 2H, J ) 7.1 Hz), 3.56
(d, 2H, J ) 6.5 Hz), 1.44 (t, 3H, J ) 7.1 Hz); 13C NMR (CDCl3,
100 MHz) δ 164.6, 148.4, 142.7, 135.5, 135.1, 132.2, 127.3,
122.5, 118.0, 61.9, 39.5, 14.3; IR (neat) 1728, 1537, 1352, 1283,
1200 cm-1; HRMS (EI+) calcd for C12H13NO4 (M) 235.0845,
found 235.0846.
2-Allyl-5-n itr oben zop h en on e (1d ). Column chromatog-
raphy (heptane-ethyl acetate 23:2) gave 370 mg (46%) of 1d
as a pale yellow oil, which crystallized upon standing overnight
1
in the refrigerator: mp 50-52 °C; H NMR (acetone-d6, 400
MHz) δ 8.35 (dd, 1H, J ) 8.5, 2.5 Hz), 8.19 (d, 1H, J ) 2.4
Hz), 7.85-7.81 (m, 2H), 7.74-7.68 (m, 2H), 7.59-7.53 (m, 2H),
5.95-5.84 (m, 1H), 5.04-4.97 (m, 2H), 3.55 (m, 2H, J ) 6.6
Hz); 13C NMR (acetone-d6, 100 MHz) δ 196.2, 147.1, 147.0,
140.7, 137.7, 136.5, 134.8, 132.7, 130.9, 129.7, 125.7, 124.0,
4-Allyln itr oben zen e (1k ). Column chromatography (hep-
tane-ethyl acetate 23:2) followed by preparative HPLC (gradi-
ent solution: 50:50 for 10 min then 60:40 for 20 min (CH3CN/
H2O) gave 267 mg (55%) of 1k as a pale yellow oil: 1H NMR
(CDCl3, 400 MHz) δ 8.14 (m, 2H, J ) 8.8 Hz), 7.34 (m, 2H, J
) 8.8 Hz), 5.99-5.89 (m, 1H), 5.18-5.09 (m, 2H), 3.49 (d, 2H,
J ) 6.7 Hz); 13C NMR (CDCl3, 100 MHz) δ 147.8, 146.6, 135.5,
(46) Curtis, R. J . Prakt. Chem. 1907, 76, 255.
(47) Ek, F.; Wistrand, L. G. Preparation of allylic aromatic com-
pounds by reaction of aromatic amines with a nitrite and an allylic
olefin. EP 1013636, 2000, 11 pp.
6380 J . Org. Chem., Vol. 67, No. 18, 2002

Aromatic Allylation via Diazotization
129.4, 123.7, 117.4, 39.9; IR (neat) 1518, 1348 cm-1; HRMS
(EI+) calcd for C9H9NO2 (M) 163.0633, found 163.0641.
4-Allylbr om oben zen e (1l). Column chromatography (pen-
tane) gave 200 mg (34%) of 1l as a colorless oil: 1H NMR
(CDCl3, 400 MHz) δ 7.42 (m, 2H, J ) 8.3 Hz), 7.07 (m, 2H, J
) 8.3 Hz), 6.00-5.88 (m, 1H), 5.11-5.04 (m, 2H), 3.34 (d, 2H,
J ) 6.7 Hz); 13C NMR (CDCl3, 100 MHz) δ 139.0, 136.8, 131.5,
130.4, 119.7, 116.3, 39.6; HRMS (EI+) calcd for C9H9Br (M)
195.9888, found 195.9887. 1H NMR data were in agreement
with literature data.48
CN were distilled off from the reaction mixture at reduced
pressure, and toluene (500 mL) was added to the orange-brown
residue. The resulting mixture was filtered twice through
alumina pads (10 × 10 cm), and the pads were washed with a
total of 1.5 L of toluene. The toluene was distilled off at reduced
pressure, and isooctane (200 mL) was added to the remaining
pale yellow residue. The mixture was stirred at 60 °C for 0.5
h (to extract the partly polymerized allyl bromide and 1,2,3-
tribromopropane from the product) and then cooled to ap-
proximately -50 °C. As soon as a white precipitate (partly
polymerized allyl bromide and 1,2,3-tribromopropane) started
to form, the solvent was decanted (including the precipitate)
and to the remaining yellow oil another portion of isooctane
(200 mL, 20 °C) was added. The oily residue in isooctane was
then stirred at -50 °C with a spatula until crystals formed.
In some cases, it was necessary to decant the isooctane phase
and add fresh isooctane before the oil started to crystallize.
The crystals (remaining in the flask) were then washed twice
with isooctane (precooled to -50 °C) at -50 °C. Residual
isooctane was removed at reduced pressure to give 78.9 g of
1h as a yellow oil containing less than 4% of 5h according to
1H NMR analysis. This corresponds to 89% yield of allyl-3,5-
dinitrobenzene. In an alternative procedure, the crystals (after
washing with isooctane at -50 °C) were collected by filtration
at -50 °C and then washed once with isooctane (precooled to
-50 °C). The latter method gave the same yield of slightly
purer 1h . Spectral data were identical to those of the reference
sample from the small scale synthesis.
4-Allyl-ter t-bu tylben zen e (1m ). Column chromatography
(heptane) gave 100 mg (19%) of 1m as a colorless oil. 1H NMR
data were in agreement with literature data.49
Allyl-2,4-d in itr oben zen e (1p ). In addition to the general
procedure for the allylation, the reaction time was extended
18 h (60 °C reaction temperature). During the first 4 h of this
additional reaction time, extra tert-butyl nitrite (2.86 mL, 24
mmol) was added in portions to the reaction mixture. Column
chromatography (heptane-ethyl acetate 23:2) followed by
preparative HPLC (gradient solution: 60:40 to 70:30 for 20
min and then 95:5 (CH3CN/H2O) of the remaining residue gave
305 mg (49%) of 1p as a yellow oil: 1H NMR (CDCl3, 400 MHz)
δ 8.77 (d, 1H, J ) 2.4 Hz), 8.38 (dd, 1H, J ) 8.5, 2.4 Hz), 7.62
(d, 1H, J ) 8.5 Hz), 6.00-5.90 (m, 1H), 5.24-5.12 (m, 2H),
3.80 (d, 2H, J ) 6.5 Hz); 13C NMR (CDCl3, 100 MHz) δ 149.1,
146.6, 142.0, 133.3, 133.2, 127.1, 120.3, 118.9, 36.9; IR (neat)
1526, 1347 cm-1; HRMS (EI+) calcd for C9H7N2O4 (M - 1)
207.0406, found 207.0407.
La r ge-Sca le Syn th esis of Allyl-3,5-d in itr oben zen e (1h ).
Neat 4h (75 g, 0.41 mol) was added in portions to a solution
of tert-butyl nitrite (84.6 mL, 0.71 mol) and allyl bromide (530
mL, 6.15 mol) in CH3CN (25 mL), keeping the temperature
between 11 and 15 °C. Before the addition of the final 25% of
4h , more tert-butyl nitrite (21 mL, 0.18 mol) was added. The
reaction mixture was then stirred at room temperature (23-
25 °C) for 1 h. Excess tert-butyl nitrite, allyl bromide, and CH3-
Ack n ow led gm en t. We thank Amersham Health
R&D AB for financial support, Dr. Peter Michelsen and
Einar Nilsson for obtaining mass spectral data, and J an
Glans for help with the preparative HPLC.
Su p p or tin g In for m a tion Ava ila ble: 13C NMR spectra for
all compounds analyzed by HRMS. This material is available
free of charge via the Internet at http://pubs.acs.org.
(48) J ones, L. B.; Foster, J . P. J . Org. Chem. 1970, 35, 1777-81.
(49) Fouquet, E.; Pereyre, M.; Rodriguez, A. L. J . Org. Chem. 1997,
62, 5242-5243.
J O0258103
J . Org. Chem, Vol. 67, No. 18, 2002 6381