Synthesis of highly labile SG1-based alkoxyamines under photochemical conditions

Article abstract of DOI:10.1021/jo800422a

(Chemical Equation Presented) Highly labile SG1-based alkoxyamines were synthesized using the photodecomposition of both azo compounds and dithiocarbamates. The former method was a straightforward procedure to obtain the alkoxyamines, but a high [azo]/[nitroxide] ratio is needed. The latter method required only a stoichiometric amount of dithiocarbamate and allowed the recovery of the disulfide after irradiation. This enabled combination of the two methods in a process where only 0.75 equiv of azo compound is needed and where sulfurous compounds acted only as intermediates.

Full text of DOI:10.1021/jo800422a

Synthesis of Highly Labile SG1-Based
Alkoxyamines under Photochemical Conditions
of NMP, numerous works have proven that alkoxyamine-based
initiators could exert an efficient control, on the radical
3
,11
polymerization of various monomers,
on chain end func-
3
,12,13
3,13,14
,
†
‡
tionalization,
Recently, we showed
rate constant, the alkoxyamine BlocBuilder (based on the
-carboxy-1-methylethyl radical and the nitroxide N-tert-butyl-
and on macromolecular architecture.
Yohann Guillaneuf,* Jean-Luc Couturier,
15,16
,
†
†
†
that owing to a very high dissociation
Didier Gigmes,* Sylvain R. A. Marque, Paul Tordo, and
†
Denis Bertin
1
UMR-CNRS 6264 Laboratoire Chimie ProVence, UniVersite´
de ProVence, AV. Escadrille Normandie-Niemen, case 542,
Marseille 13397 Cedex 20, France, and ARKEMA, Centre de
Recherche de Rh oˆ ne Alpes, rue H. Moissan,
N-(1-diethylphosphono-2,2-dimethylpropyl) or SG1) was up to
now one of the most potent alkoxyamines reported in NMP.
Moreover, due to the carboxylic acid group being allowed to
17
18
work in aqueous media and to perform coupling reactions,
this alkoxyamine has been chosen by ARKEMA to be synthe-
6
9493 Pierre B e´ nite Cedex, France.
1
9
sized in a large scale. The industrial requirements for its
commercialization involve developing a new synthetic approach
taking into account special criteria such as large scale produc-
tion, economic viability, environmental friendliness, etc.
guillaneuf@srepir1.uniV-mrs.fr;
didier.gigmes@uniV-proVence.fr
ReceiVed March 3, 2008
SCHEME 1. SG1 and SG1-Based Alkoxyamines
Although many routes have been developed to synthesize
20
alkoxyamines, these compounds are mainly prepared through
two pathways: via nucleophilic substitution reaction or via
coupling reaction between alkyl and nitroxyl radicals. The nu-
cleophilic substitution of an alkyl halide by the hydroxylamine
anion is scarcely employed when SG1 nitroxide is used because
of the relative instability of the corresponding hydroxyaminate
anion. The coupling of radicals is therefore the most developed
and used approach. The coupling between alkyl and nitroxide
has been extensively used, and all common methods to prepare
alkyl radical have been employed: addition of a radical on an
Highly labile SG1-based alkoxyamines were synthesized
using the photodecomposition of both azo compounds and
dithiocarbamates. The former method was a straightforward
procedure to obtain the alkoxyamines, but a high [azo]/
[
nitroxide] ratio is needed. The latter method required only
21
a stoichiometric amount of dithiocarbamate and allowed the
recovery of the disulfide after irradiation. This enabled
combination of the two methods in a process where only
0
.75 equiv of azo compound is needed and where sulfurous
compounds acted only as intermediates.
(
8) Bertin, D.; Gigmes, D.; Marque, S. R. A.; Tordo, P. Tetrahedron 2005,
1, 8752–8761.
9) Wetter, C.; Jantos, K.; Woithe, K.; Studer, A. Org. Lett. 2003, 5, 2899–
902.
10) Leroi, C.; Bertin, D.; Dufils, P. E.; Gigmes, D.; Marque, S.; Tordo, P.;
6
(
2
Duringthepastfewyears,theradicalreactivityofalkoxyamines
has attracted a particular interest for its use in nitroxide mediated
(
Couturier, J. L.; Guerret, O.; Ciufolini, M. A. Org. Lett. 2003, 5, 4943–4945.
(11) (a) Guillaneuf, Y.; Gigmes, D.; Marque, S.; Astolfi, P.; Greci, L.; Tordo,
P.; Bertin, D. Macromolecules 2007, 40, 3108. (b) Couvreur, L.; Lefay, C.;
Belleney, J.; Charleux, B.; Guerret, O.; Magnet, S. Macromolecules 2003, 36,
1
–3
4
polymerization
(NMP) and in organic synthesis. Indeed,
numerous studies have shown that the thermal reversible
alkoxyamine dissociation is a highly efficient method for the
8
260. (c) Phan, T.; Bertin, D. Macromolecules 2008, 41, 1886–1895.
(12) Hill, N. L.; Braslau, R. J. Polym. Sci., Part A, Pol. Chem. 2007, 40,
848.
5
,6
clean generation of carbon-centered radicals. For instance,
this method has been successfully applied for metal-free
7
(13) Bernaerts, K. V.; Du Prez, F. E. Prog. Polym. Sci. 2006, 31, 671–722.
7,8
9,10
(14) Dufils, P.-E.; Chagneux, N.; Gigmes, D.; Trimaille, T.; Marque, S. R. A.;
cyclization and intermolecular radical addition.
In the field
Bertin, D. Polymer 2007, 48, 18, 5219–5225.
15) Chauvin, F.; Couturier, J.-L.; Dufils, P. E.; Gerard, P.; Gigmes, D.;
(
†
Universit e´ de Provence.
ARKEMA.
Guerret, O.; Guillaneuf, Y.; Marque, S. R. A.; Bertin, D.; Tordo, P. ACS Symp.
Ser. 2006, 944, 326.
‡
(
(
1) Solomon, D. H.; Rizzardo, E.; Cacioli, P. US Patent. 4,581,429, 1985.
2) Georges, M. K.; Veregin, R. P. N.; Kazmaier, P. M.; Hamer, G. K.
(16) Chauvin, F.; Dufils, P. E.; Gigmes, D.; Guillaneuf, Y.; Marque, S. R. A.;
Tordo, P.; Bertin, D. Macromolecules 2006, 39, 16, 5238–5250.
(17) Charleux, B.; Nicolas, J. Polymer 2007, 48, 20, 5813–5833.
(18) (a) Beaudoin, E.; Dufils, P. E.; Gigmes, D.; Marque, S.; Petit, C.; Tordo,
P.; Bertin, D. Polymer 2006, 47, 1, 98–106. (b) Vinas, J.; Chagneux, N.; Gigmes,
D.; Trimaille, T.; Favier, A.; Bertin, D. Polymer,submitted.
(19) Gigmes, D.; Bertin, D.; Guerret, O.; Marque, S. R. A.; Tordo, P.;
Chauvin, F.; Couturier, J.-L.; Dufils, P.-E. PCT WO 2004/014926, February 13,
2004.
Macromolecules 1993, 26, 2987–2988.
3) Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. ReV. 2001, 101, 12,
661–3688.
4) (a) Studer, A.; Schulte, T. Chem. Record 2005, 5, 27–35. (b) Wetter, C.;
(
3
(
Studer, A. Chem. Commun. 2004, 174–175. (c) Molawi, K.; Schulte, T.;
Siegenthaler, K. O.; Wetter, C.; Studer, A. Chem. Eur. J. 2005, 11, 2335–2350.
(
(
(
5) Studer, A. Chem. Soc. ReV. 2004, 33, 267–273.
6) Studer, A.; Amrein, S. Synthesis 2002, (7), 835–849.
7) Studer, A. Angew. Chem., Int. Ed. 2000, 39 (6), 1108–1111.
(20) Nesvadba, P. Chimia 2006, 60, 12, 832–840.
(21) Couturier, J.-L. ARKEMA, personal communication.
4
728 J. Org. Chem. 2008, 73, 4728–4731
10.1021/jo800422a CCC: $40.75  2008 American Chemical Society
Published on Web 05/22/2008

2
2
activated olefin, abstraction of a labile hydrogen atom on a
TABLE 1. Synthesis of Alkoxyamines Using Photodecomposition
of Azo Compounds
2
2
23
substrate by an alkoxyl radical, carbanion oxidation, use of
Jacobsen catalyst, thermal decomposition of azo compounds,
etc. A few years ago, Matyjaszewski applied the concept of
atom transfer radical polymerization (ATRP) to generate
readily alkyl radicals. The technique called atom transfer radical
addition (ATRA) affords high yield and purity. Nevertheless,
the use of transition-metal complexes such as copper complexes
is a major drawback for the scale-up of this technique due to
the production of toxic effluents.
2
4
25
•
alkoxyamine
solvent
isolated yield (%)
[Azo]/[NO ] ratio
1
2
3
4
acetonitrile
THF
H2O/EtOH
H2O/EtOH
93
75
60
70
2.8
3.0
2.85
2.8
2
6
2
7
not only to validate the photolytic synthetic route but also to
prepare functional highly labile alkoxyamines.
To overcome the above-mentioned drawback, this work
presents a convenient metal-free synthesis with high yield and
purity of tertiary SG1-based alkoxyamines, especially Bloc-
Builder. Since the tertiary SG1-based alkoxyamines exhibit
particularly low dissociation temperature (cleavage temperature
2
8
close to 30 °C), the alkyl radicals have to be generated at
room temperature or below. Except for the elegant work of
29
Braslau et al., only photochemistry could allow working using
these mild conditions. Usually, the alkoxyamines are stable
under UV irradiation and to induce homolytic cleavage,
sensitizers, or chromophore groups linked to the nitroxide moiety
3
0
have to be used.
The preparation of TEMPO-based alkoxyamines under pho-
tochemical conditions has been already reported. This method
is based on benzylic H-atom abstraction by photoinduced
3
1
alkoxyl radicals. The extent of this technique for the synthesis
of tertiary crowded SG1-based alkoxyamines was, however, not
Irradiations were performed at 10 °C in a 100 mL double-
jacketed reactor equipped at the top with a specific optic fiber
2
1
successful. Two other photochemical routes have been de-
veloped, a first one by Scaiano et al. in which the authors
used the photoinduced cleavage of the C-Br bond to generate
alkyl radicals. In a second one, Engel et al. have reported the
use of TEMPO as a spin-trap of alkyl radical generated by azo
compounds in order to study the cage effect but in this case the
alkoxyamines have not been isolated. In the same manner,
(
(
320-400 nm) linked to a Lumatec SuperLite SUV-DC-P lamp
3
1
-2
200 W high-pressure mercury lamp, 6 W ·cm at the end of
the fiber).
3
2
Following this procedure, we prepared (Table 1) the targeted
alkoxyamines in moderate to high yields. However, it can be
noted that such results needed a large excess of the azo
compound ([azo]0/[SG1]0 ratio ) 2.8-3.0). We investigated the
effect of the reactants concentration (with a constant [azo]0/
1
Rizzardo et al. have synthesized 4-cyano-4-(di-tert-butylami-
nyloxy)pentanol by photodecomposing the 4,4′-azobis(4-cyano-
n-pentanol) in presence of di-tert-butyl nitroxide DBNO. The
yield obtained in this experiment was quite low (40%).
[
1
SG1]0 ratio), and no influence was observed in the range of
.0-1.0 × 10 mol·L . These results confirmed that the
-3
-1
recombination between alkyl radicals occurred only in the cage
of solvent. Usually, the efficiency of azo compounds is close
to 0.8, that is a ratio of 0.625 should be sufficient to have a
quantitative yield. The high [azo]0/[SG1]0 ratio in our case
(efficiency of 0.18) could be due to the nonoptimal geometry
of irradiation, and mostly to the temperature in the medium.
The cage effect is indeed known to drastically decrease the
efficiency when decreasing temperature (0.9 to 0.45 when T
decreases from 98 to 60 °C). The increase of cage effect to
rationalize our result seems awfully large. We therefore
investigated the temperature dependence of the cage effect using
the modeling developed by Russell and Buback (see the
Supporting Information for more details) and found good
agreement (calculated 84%, experimental 82%) with our
experimental results.
Nevertheless, it can be noticed that the reaction yield is always
quantitative with respect to SG1 provided that there is enough
radicals to react with the nitroxide. The conditions used and
results are given in Table 1.
As the major side product, which was the dimer formed by
the recombination between two alkyl radicals, was inert for the
Attracted by these results, we investigated the preparation of
crowded tertiary SG1-based alkoxyamines 1-4 by irradiation
of commercially available azo compounds (azobisisobutyroni-
trile AIBN, 4,4′-azobiscyanopentanoic acid ACPA, ethyl 2,2′-
azobisisobutyrate DEAB, 2,2′-azobis(2-methylpropionamidine)
dihydrochloride V50). The purpose of these experiments was
3
3
(
22) Hawker, C. J.; Barclay, G. C.; Orellana, A.; Dao, J.; Devonport, W.
Macromolecules 1996, 29, 16, 5245–5254.
(
23) Wetter, C.; Jantos, K.; Woithe, K.; Studer, A. Org. Lett. 2003, 5, 2899–
2
902.
3
4
(
24) Dao, J.; Benoit, D.; Hawker, C. J. J. Polym. Sci., Part A, Pol. Chem.
1
998, 362161–2167.
(
(
(
25) Wang, D.; Whu, Z. Macromolecules 1998, 31, 6727–6729.
26) Matyjaszewski, K.; Xia, J. H. Chem. ReV. 2001, 101, 9, 2921–90.
27) (a) Matyjaszewski, K.; Greszta, D. US Patent. 5,910,549, 1999. (b)
Woodworth, B. E.; Zhang, X.; Metzner, Z.; Gaynor, S. G.; Matyjaszewski, K.
Macromolecules 1998, 31, 5955.
(28) Beaudoin, E.; Bertin, D.; Gigmes, D.; Marque, S. R. A.; Siri, D.; Tordo,
P. Eur. J. Org. Chem. 2006, 7, 1755–1768.
(
29) (a) Braslau, R.; Burrill, L. C.; Siano, M.; Naik, N.; Howden, R. K.;
Mahal, L. K. Macromolecules 1997, 30 (21), 6445–6450. (b) Braslau, R.;
Tsimelzon, A.; Gewandter, J. Org. Lett. 2004, 6 (13), 2233–2235.
(
30) (a) Scaiano, J. C.; Connoly, T. J.; Mohtat, N.; Pliva, C. N. Can. J. Chem.
1
2
997, 75, 92–97. (b) Hu, S.; Malpert, J. H.; Yang, X.; Neckers, D. C. Polymer
000, 41, 445–452.
(31) Connolly, T. J.; Bardovi, M. V.; Mohtat, N.; Scaiano, J. C. Tetrahedron
(33) Moad, G.; Solomon, D. H., The Chemistry of Free Radical Polymeri-
zation; Pergamon: Oxford, 1995; p 63.
(34) Buback, M.; Huckestein, B.; Kuchta, F.-D.; Russell, G.; Schmid, E.
Macromol. Chem. Phys. 1994, 195, 2117–2140.
Lett. 1996, 37 (28), 4919–4922.
32) Billera, C. F.; Dunn, T. B.; Barry, D. A.; Engel, P. S. J. Org. Chem.
998, 63, 9763–9768.
(
1
J. Org. Chem. Vol. 73, No. 12, 2008 4729

SCHEME 2. One-Pot Synthesis of the BlocBuilder
Alkoxyamine
SCHEME 3. Synthesis of BlocBuilder by the
Photodecomposition of Dithiocarbamate
used in polymerization where dithiocarbamate could act as
iniferters and control the polymerization of styrene, acrylate,
38
and methacrylate derivatives. The spin-trapping technique has
39
been used to understand the mechanism of decomposition and
the results showed that if the alkyl R group linked to the sulfur
is stable enough, the dissociation occurred exclusively between
the R and dithiocarbamyl group. In this system, we avoid the
irreversible termination in the solvent cage, since the nondif-
fusion of the alkyl radical only led to the reformation of the
dithiocarbamate.
polymerization, no further purification was required. However,
if necessary, the dimer can be easily removed by column
chromatography. In addition, alkoxyamine 3 and 4 are very
interesting because purification can be also made by selective
precipitation in an appropriate solvent (dichloromethane for the
alkoxyamine 3, acidic water for alkoxyamine 4).
After showing that the photodecomposition of azo was a
efficient route to prepare SG1-based alkoxyamines, we decided
to use this strategy to prepare the alkoxyamine BlocBuilder.
Since the azo precursor corresponding to this alkoxyamine is
not commercially available, the 2,2′-azobisisobutyric acid
sodium salt was prepared by hydrolysis of azoesters (i.e.,
DEAB). The neutralization of this product to prepare 2,2′-
azobisisobutyric acid only conducted to a strong exothermic
degradation. The sodium salt exhibits a higher stability: an
aqueous DSC measurement showed no thermal dissociation
before the thermal degradation which occurred at 210 °C (see
the Supporting Information for details).
The first step of this sequence (Scheme 3) consists of the
synthesis of the corresponding dithiocarbamate, i.e., the S-(1-
methyl-1-carboxyethyl)-N,N-diethyl dithiocarbamate 6. Three
different methods are available in the literature to prepare this
compound: the nucleophilic substitution between the dithiocar-
40
bamate sodium salt and the 2-bromoisobutyric acid, the radical
4
1
addition of an azo compound and a disulfide followed by an
hydrolysis and the haloform reaction between acetone, chloro-
4
2
form, and the dithiocarbamate sodium salt.
These three methods have been used (see the Supporting
Information), and we chose the nucleophilic substitution for the
easily purification process regardless of the moderate yield.
The photolysis of 6 in the presence of SG1 was carried out
3
1
in ethanol. We observed by P NMR only the peak of the
targeted alkoxyamine. It can be noticed that the side product
issued from the cross recombination between SG1 and the
dithiocarbamyl radical 7 was not detected. We investigated the
effect of the [dithiocarbamate]0/[nitroxide]0 ratio and we ob-
served that a ratio of 1.1 is sufficient to obtain a quantitative
yield in BlocBuilder. This result shows that no alkyl self-
termination occurred in this system compared to the photolysis
of azo compounds. This procedure was extended to secondary
dithiocarbamate, i.e., S-(1-carboxyethyl)-N,N-diethyl dithiocar-
bamate with success even if the alkoxyamine has not been
A two- or three-step (depending of the starting materials
AIBN or DEAB, see Scheme 2), one-pot synthesis has been
3
5
developed to produce the BlocBuilder. The first step, which
is not compulsory, was the preparation of azoesters from AIBN
3
6
according to the procedure of Gritsenko. During this step, an
iminoether derivative was reacted with water and can be isolated
or kept to be used for the next step. The second step consisted
in the hydrolysis of the azoester in a H2O/EtOH mixture by
adding 3 equiv of NaOH. The nitroxide in solution in EtOH
was added, and then the mixture was photolyzed until the the
medium was bleached. The alkoxyamine BlocBuilder was then
purified very easily by precipitation in cold acidic water (93%
yield by the two-step process).
3
1
isolated ( P NMR spectrum, see the Supporting Information
for details).
The purification procedure in this method is as follows: after
the aqueous acidic precipitation, we obtained a mixture of
BlocBuilder and tetraethylthiuram disulfide (TEDS) coming
from the dimerization of dithiocarbamyl radicals. In order to
remove the disulfide, a NaOH solution was added to the medium
until pH 10 to solubilize the alkoxyamine, and then the mixture
was filtrated off to remove the side product as a yellow powder
The photodecomposition of azo compounds enabled the
preparation in one pot of the desired alkoxyamine with high
yield and purity in a metal-free process. Nevertheless, from an
industrial point of view, the inherent loss of radicals due to the
cage effect during the decomposition of azo compound at low
or moderate temperature involved to use a large excess of azo
initiator compared to the nitroxide.
(
more than 80% recovery). A further acidification/precipitation
step enabled us to obtain the desired alkoxyamine (95% yield)
4
3
as a white powder after filtration. Elemental analysis showed
To develop an alternative but sustainable metal-free synthesis
of BlocBuilder, we studied the photodecomposition of dithio
compounds and in particular dithiocarbamate compounds,
known to undergo homolytic cleavage of the sulfur-carbon
(
(
38) Otsu, T. J. Polym. Sci Part A: Polym. Chem. 2000, 38, 12, 2121–2136.
39) Doi, T.; Matsumoto, A.; Otsu, T. J. Polym. Sci Part A: Polym. Chem.
1994, 32, 2241–2249.
37
(40) Ishizu, K.; Khan, R. A.; Ohta, Y.; Furo, M. J. Polym. Sci Part A: Polym.
Chem. 2004, 42, 76–82.
bond under UV irradiation. This property has been extensively
(41) Bouhadir, G.; Legrand, N.; Quiclet-Sire, B.; Zard, S. Z. Tetrahedron
(
35) Guillaneuf, Y.; Couturier, J.-L.; Guerret, O.; Gigmes, D. WO 2005113566,
004.
36) Gritsenko, F. R.; Spirin, Y. L.; Kochetova, G. I.; Kochetov, D. P. SoV.
Prog. Chem. 1977, 43, 68–71.
37) Otsu, T. J. Polym. Sci. 1956, 21, 559.
Lett. 1999, 40, 277–280. (b) Thang, S. H.; Chong, Y. K.; Mayadunne, R. T. A.;
Moad, G.; Rizzardo, E. Tetrahedron Lett. 1999, 40, 2435–2438.
(42) LaiJ. T. US Patent. 0120101, 2003.
(43) Guillaneuf, Y.; Couturier, J.-L.; Guerret, O.; Gigmes, D.; Bertin, D.
WO2006111637, 2005.
2
(
(
4
730 J. Org. Chem. Vol. 73, No. 12, 2008

SCHEME 4. New Synthetic Procedure Used To Prepare the
BlocBuilder
SuperLite SUV-DC-P lamp (200 W high-pressure mercury lamp,
6 W·cm- at the end of the fiber) at 365 nm for 4.5 h at 10 °C.
The solvent was removed to yield a pale yellow oil, which was
purified by silica gel column chromatography using pentane/diethyl
ether 1:1 as eluent. The product (2.48 g, 6.1 mmol, yield 75%)
2
31
was obtained as a pale yellow oil. P NMR (CDCl
3
, 121.49 MHz):
1
δ 25.48 (s). H NMR (CDCl
3
, 300.13 MHz): δ 1.00 (s, 9 H); 1.09
1
(
s, 9 H); 1.21 (m, 9 H); 1.50 (s, 3 H); 1.60 (s, 3 H); 3.13 (d, JH-P
1
)
25.88 Hz, 1 H); 3.90-4.10 (m, 4 H); 4.26 (t, JH-H ) 6.8 Hz,
1
3
3
2
H). C NMR (CDCl
3
, 75.47 MHz): δ 13.6 (CH
3
); 15.7 ( JC-P
)
3
6
.59 Hz)-16.1 ( JC-P ) 6.04 Hz) (CH
3
); 22.0-27.6 (CH
3
); 27.7
3
2
(CH
3
); 29.5 ( JC-P ) 5.49 Hz) (CH
3
); 35.5 ( JC-P ) 6.04 Hz) (C
); 60.2 (CH
q
);
no trace of sulfur in the final product. One extra advantage of
this method lies on the recovery of the tetraethylthiuram
disulfide. This product could indeed be recycled as a value-
added product but could also be one of the reactants used to
prepare the dithiocarbamate 6 by the radical addition method
5
8.0 (2JC-P ) 7.69 Hz) (CH
2
2
); 61.2 ( JC-P ) 6.04 Hz)
2
1
(
CH
2
); 61.7 (C
q
); 68.7 ( JC-P ) 137.22 Hz) (CH); 83.2 (C
q
); 174.6
+
(C
q
). ESI-HRMS: calcd for C19
10.2663.
-Methyl-2-[N-tert-butyl-N-(1-diethoxyphosphoryl-2,2-dime-
thylpropyl)aminoxy]propionic Acid (BlocBuilder) Alkoxyamine.
A N -bubbled solution of S-(1-methyl-1-carboxyethyl)-N,N-diethyl
H40NO
6
P [M + H] 410.2666, found
4
2
4
1
(
92% yield, see the Supporting Information). From an
economic point of view, we developed an experimental proce-
dure to prepare the BlocBuilder using the recycled disulfide and
combining the photodecomposition of DEAB and dithiocar-
bamate 6 (Scheme 4).
As the intermediate 6 is thermally stable, this procedure
allowed us to avoid the low efficiency of azo compounds
2
-2
dithiocarbamate 6 (4.8 g, 2.0 × 10 mol) and SG1 (6.28 g, 1.82
-2
× 10 mol) in 350 mL of ethanol was irradiated at 365 nm using
the previously described optic fiber for 4.5 h at 10 °C. After the
solution bleached, the resulting mixture was precipitated in
acid-water (200 mL + 3 mL of 33% HCl solution), filtered, and
dried under vacuum. This pale yellow powder was then diluted in
(
thermal or photo) decomposition at low temperature. Indeed,
2
00 mL of water at pH 9. The yellow precipitate is removed by
the first step has been carried out in refluxing ethyl acetate.
Interestingly, the dithiocarbamate acts only as an intermediate
and does not appear in the overall reaction scheme.
filtration (TEDS, > 80% recovery), and the filtrate was acidified
by addition of HCl. After filtration and drying under vacuum, the
product (6.58 g, 17.2 mmol, yield 95%) was obtained as a white
In conclusion, two alternative routes to the ATRA method
for the synthesis of alkoxyamines have been developed in order
to avoid any metal in both products and effluents. The two
methods are based on the photodecomposition of either azo
compounds or dithiocarbamates and led to quantitative yields
with respect to the nitroxide. The first method was easier to
handle but needed an excess of reactant. The second one
required only a stoichiometric amount of reactant but used
sulfurous compounds. The two methods could be combined to
prepare highly labile alkoxyamines with only a slight excess of
reactant and sulfurous compounds as intermediates.
31
powder. Mp: 122-125 °C. P NMR (CDCl
3
, 121.49 MHz): δ 27.63
1
(
s). H NMR (CDCl
3
, 300.13 MHz): δ 1.13 (s, 9 H); 1.22 (s, 9 H);
1
1.30 (t, J
H-H
) 7 Hz, 6 H); 1.58 (s, 3 H); 1.76 (s, 3 H); 3.34 (d,
1
13
JH-P ) 27.22 Hz, 1 H); 3.90-4.30 (m, 4 H). C NMR (CDCl ,
3
75,47 MHz): δ 6.1-16.5 (CH
3
); 24.1 (CH
); 36.1 (CH); 60.2-60.4 (CH ); 63.3 (C
4.9 (C ); 176.8 (C ). Anal. Calcd for C17
.51; N, 3.67. Found: C, 53.78; H, 9.57; N, 3.69.
3
); 28.8 (CH
); 69.0-70.8 (CH);
36 6
H O N: C, 53.53; H,
3
); 29.8
(CH
3
2
q
8
9
q
q
Acknowledgment. ARKEMA, the University of Provence,
and the CNRS are thanked for their financial support.
Supporting Information Available: Characterization of the
alkoxyamines 1, 3, 4, BlocBuilder, and the intermediates used
such as dithiocarbamate 6 and 2,2′-azobisisobutyric acid sodium
salt 5. Copies of the NMR spectra for new products. Cage effect
modeling. This material is available free of charge via the
Internet at http://pubs.acs.org.
Experimental Section
Ethyl 2-Methyl-2-[N-tert-butyl-N-(1-diethoxyphosphoryl-2,2-
dimethylpropyl)aminoxy]propionoate (DEAB-SG1) Alkoxyamine
2
2. A N -bubbled solution of ethyl 2,2′-azobisisobutyrate DEAB (6.5
g, 24 mmol) and SG1 (2.8 g, 8.1 mmol) in THF (100 mL) was
irradiated in a 100 mL double-jacketed reactor equipped at the top
with a specific optic fiber (320-400 nm) linked to a Lumatec
JO800422A
J. Org. Chem. Vol. 73, No. 12, 2008 4731