Radical aminomethylation of unactivated alkenes

Article abstract of DOI:10.1021/ol801162m

(Chemical Equation Presented) S-Succinimidomethyl- and S-phthalimidomethyl xanthates are added efficiently to various alkenes resulting in an overall aminomethylation process. In contrast, under similar conditions, S-pyrrolidonyl xanthate gives rise mostl

Full text of DOI:10.1021/ol801162m

ORGANIC
LETTERS
2008
Vol. 10, No. 15
3279-3282
Radical Aminomethylation of
Unactivated Alkenes
Be´atrice Quiclet-Sire and Samir Z. Zard*
Laboratoire de Synthe`se Organique, CNRS UMR 7652 Ecole Polytechnique,
91128 Palaiseau Cedex, France
zard@poly.polytechnique.fr
Received May 21, 2008
ABSTRACT
S-Succinimidomethyl- and S-phthalimidomethyl xanthates are added efficiently to various alkenes resulting in an overall aminomethylation
process. In contrast, under similar conditions, S-pyrrolidonyl xanthate gives rise mostly to oligomers. This unexpected difference in reactivity
is attributed to the more important allylic character of the intermediate radical in the case of the imide derivative as compared to the lactam.
In the course of our work on the radical xanthate transfer
reaction,1 we made a puzzling observation: while clean
Scheme 1
radical additions to N-vinyl pyrrolidone 1 could be achieved
in the usual manner,2 similar reactions with the synthetically
more interesting N-vinyl phthalimide 3 furnished mostly
oligomers 7, instead of the desired adducts 6 (Scheme 1;
PhthN ) phthalimido).3 Only with xanthates such as 8,
leading to highly stabilized radicals (e.g., 9), could we obtain
a reasonable yield of the desired monoadduct (10 in this
case).
The extensive formation of oligomers is often the result
of the adduct radical 11 being more stable than radical R•
derived from the starting xanthate 5, as can be seen in the
simplified mechanistic manifold of Scheme 2. The relative
stabilities of the starting and adduct radicals in the xanthate
addition process is a key element for success.1 Intermediate
radical 12 has to fragment preferentially into starting radical
(1) For reviews of the xanthate transfer chemistry, see: (a) Zard, S. Z.
R• and not back to adduct radical 11, because otherwise the
Angew. Chem., Int. Ed. Engl. 1997, 36, 672. (b) Zard, S. Z. In Radicals in
latter will accumulate, causing oligomer formation by further
additions to the alkene component.
Thus, the easy, and for our purposes unwanted, oligomer
formation in the case of N-vinyl phthalimide pointed to a
significantly greater stability of adduct radical 4 as compared
with the structurally closely related N-vinyl pyrrolidone
Organic Synthesis; Renaud, P., ; Sibi, M. P., Eds; Wiley-VCH: Weinheim,
2001; Vol. 1, p 90. (c) Quiclet-Sire, B.; Zard, S. Z. Chem.-Eur. J. 2006,
12, 6002. (d) Quiclet-Sire, B.; Zard, S. Z. Top. Curr. Chem. 2006, 264,
201. (e) Zard, S. Z. Org. Biomol. Chem. 2007, 5, 205.
(2) See for example: Gagosz, F.; Zard, S. Z. Org. Lett. 2003, 5, 2655.
(3) N-Vinylphthalimide has very recently been polymerized using
xanthates by what is now called the RAFT/MADIX process: Maki, Y.; Mori,
H.; Endo, T. Macromol. Chem. Phys. 2007, 208, 2589.
10.1021/ol801162m CCC: $40.75
Published on Web 06/27/2008
2008 American Chemical Society

likelihood, small but sufficient to allow control of the radical
process and to curtail the formation of unwanted oligomers.
Furthermore, phthalimidomethyl xanthate 17 proved only
marginally more reactive (a factor of 1.3) than the succin-
imido analogue 15, as shown by a competition experiment
between the two reagents. The stabilization due to the second
carbonyl group thus appears to be more important than the
presence of the aromatic ring in the phthalimido derivative.
Finally, pivaloyl xanthate 21 reacted sluggishly with allyl
cyanide to furnish mostly oligomers and returned starting
material but almost no monoadduct 22. This militates against
polar factors playing a predominant role in the success of
the imide derivatives,6 since radical 23 should exhibit a
similar, if not more, electron-withdrawing character as
compared to 20, where X ) O.7
We had occasion, in the context of a different project, to
perform radical additions starting with the isopropyl phthal-
imido xanthate 19,8 but because the radical generated in this
case is a tertiary radical, and therefore necessarily more stable
than the (usually) secondary adduct radical, its addition to a
nonactivated alkene did not surprise us. The fact that the
addition is successful even with the parent primary xanthate
17 opens a major avenue in synthesis because of the
enormous importance of primary amines in organic and
medicinal chemistry.
Scheme 2
derived radical 2. This is opposite to what one would have
predicted upon a superficial examination since the lone pair
on the nitrogen should be more available to stabilize the
radical in 2 than in 4.4
This unanticipated difference in stability was confirmed
when the radical addition of xanthate 13 to allyl cyanide
was found to proceed sluggishly, giving a poor yield of the
corresponding adduct 14 and important quantities of oligo-
mers, which made purification difficult (Scheme 3). In
Scheme 3
The radical additions of reagents 15 or 17 can be viewed
as the overall synthetic equivalent of a hydroaminomethy-
lation of an alkene, a transformation traditionally ac-
complished through the tandem combination of a hydro-
formylation reaction with a reductive amination (eq 1,
Scheme 4).9 The hydroaminomethylation and related trans-
(5) It is interesting to note that, in contrast to the corresponding lactams,
cyclic imides exhibit two carbonyl stretching frequencies in the infra red
spectrum, which may be separated by as much as 70 cm-1, and which have
been attributed to the existence of a strong coupling between the two
carbonyl groups due to conjugation of the type pictured in structure 20b.
Furthermore, the stretching frequencies are sensitive to the substituents on
the nitrogen atom indicating a possible interaction of the type shown in
structure 20c. See: Hargreaves, M. K.; Pritchard, J. G.; Dave, H. R. Chem.
ReV. 1970, 70, 439. and references cited therein.
(6) In its recent use in a RAFT/MADIX controlled radical polymeri-
zation, the relative efficiency of xanthate 17 was attributed simply to the
electrophilic nature of radical 20a: (a) Postma, A.; Davis, T. P.; Evans,
R. A.; Li, G.; Moad, G.; O’Shea, M. S. Macromolecules 2006, 39, 5293.
(b) Postma, A.; Davis, T. P.; Evans, R. A.; Li, G.; Moad, G.; O’Shea, M. S.
Macromolecules 2006, 39, 5307.
(7) The 1H chemical shift signal of the N-CH2-S protons in 17 appears
at δ 5.35 ppm, whereas that of the O-CH2-S in 21 appears at δ 5.65 ppm,
which is 0.30 ppm more downfield, reflecting the slightly stronger electron-
withdrawing effect of the pivalate in comparison to that of the phthalimido
group.
dramatic contrast, the addition of succinimido- and phthal-
imido-methyl xanthates 15 and 17 took place smoothly and
cleanly to furnish the respective adducts 16 and 18 in good
yield and without untoward formation of oligomers. The most
plausible rationalization appears to be a more significant
contribution of canonical forms 20b and 20c when X ) O
than when X ) H2,5 causing a significant increase in the
allylic character radical in the former case and therefore an
increase in its stability. This difference in stability is, in all
(8) This xanthate was generated in situ by decarbonylation of the
corresponding S-acyl xanthate: Heinrich, M.; Zard, S. Z. Org. Lett. 2004,
6, 4969.
(9) (a) Eilbracht, P.; Ba¨rfacker, L.; Buss, C.; Hollmann, C.; Kitsos-
Rzychon, B. E.; Kranemann, C. L.; Rishe, T.; Roggenbuck, R.; Schmidt,
A. Chem. ReV. 1999, 99, 3329. For some recent studies, see: (b) Briggs,
J. R.; Klosin, J.; Whiteker, G. T. Org. Lett. 2005, 7, 4795. (c) Beller, M.;
Seayad, J.; Tillack, A.; Jiao, H. Angew. Chem., Int. Ed. 2004, 43, 3368. (d)
Ahmed, M.; Seayad, A. M.; Jackstell, R.; Beller, M. J. Am. Chem. Soc.
2003, 125, 10311. (e) Angelovski, G.; Eilbracht, P. Tetrahedron 2003, 59,
8265. (f) Seayad, A. M.; Ahmed, M.; Klein, H.; Jackstell, R.; Gross, T.;
Beller, M. Science 2002, 297, 1676. (g) Rische, T.; Eilbracht, P.; Muller,
K.-S. Tetrahedron 1999, 55, 9801. (h) Rische, T.; Eilbracht, P. Tetrahedron
1999, 55, 1915–1920. (i) Bergmann, D. J.; Campi, E. M.; Jackson, W. R.;
Patti, A. F.; Saylik, D. Tetrahedron Lett. 1999, 40, 5597. (j) Zimmermann,
B.; Herwig, B.; Beller, M. Angew. Chem., Int. Ed. 1999, 38, 2372. (k) Ojima,
I.; Tzamarioudaki, M.; Eguchi, M. J. Org. Chem. 1995, 60, 7078.
(4) The stabilization of aminoalkyl radicals is due to a favorable two-
orbital-three-electron interaction. When the availability of the lone pair
on nitrogen is decreased, for example, by protonation, the result is a
destabilization of the aminoalkyl radical. See: Mayer, P. M.; Glukhovtsev,
M. N.; Gauld, J. W.; Radom, L. J. Am. Chem. Soc. 1997, 119, 12889.
3280
Org. Lett., Vol. 10, No. 15, 2008

Scheme 4
Table 1. Radical Aminomethylation of Alkenes
formations apply, with very rare exceptions,9,9b to simple
alkenes, and the use of ammonia in these processes, which
would lead to primary amines in principle, is highly
problematic in practice because of the rapid reaction of the
primary amine product with the aldehyde generated in the
hydroformylation step.9,9i A further problem is the frequent
obtention of mixtures of linear and branched products as
implied by eq 1 in Scheme 4. The utility of this route to the
synthesis of complex, densely functionalized structures is
thus severely limited.
The present, radical-based approach10 complements, and
often surpasses, the organometallic route, owing to the
compatibility of radical processes with many functional
groups and the mildness of the reaction conditions.11 For
the remainder of this study, we used xanthate 17 since it is
nicely crystalline and readily prepared in essentially quantita-
tive yield in one step from commercially available and cheap
N-chloromethylphthalimide.12 If the xanthate group in adduct
24 is reductively removed and the phthalimide cleaved off
with hydrazine or other reagents, the overall process becomes
strictly equivalent to a hydroaminomethylation of an alkene
(eq 2, Scheme 4).
The presence of the xanthate in 24 can of course be
exploited in numerous other ways to introduce further
functionality (Z * H in 25).1 The varied examples collected
in Table 1 are self-explanatory and demonstrate unequivo-
cally the tremendous power of this technology. Yields in
parentheses correspond to yields based on recovered starting
material. In the case of compounds 30, 33, 36, 39, 45, and
46, approximately 1:1 mixtures of diastereomers were
obtained.
The xanthate group in the products can be removed by
various methods.1 We introduced a few years ago a reductive
dexanthylation procedure involving lauroyl peroxide/isopro-
panol.13 An application illustrating its compatibility with the
phthalimido group, as well as the synthesis of new hexaflu-
(10) Intermolecular radical hydroaminoalkylations of alkenes with simple
amines under initiation with peroxides or UV light are known but are often
low yielding and very limited in scope, working best, not unexpectedly,
with electron-poor alkenes. See: (a) Hoffmann, N. Chem. ReV. 2008, 108,
1052. (b) Renaud, P.; Giraud, L. Synthesis 1996, 913.
orinated protected amino alcohol 47, is displayed in Scheme
5. The isolation of the initial addition product 29 is not
necessary: the two operations can be carried out in one pot
(11) Radical 20 has been generated from the corresponding bromide
and iodide using stannane chemistry, and two examples of its addition to
electrophilic olefins have been reported: (a) Campbell, E. F.; Park, A. K.;
Kinney, W. A.; Fengl, R. W.; Liebeskind, L. S. J. Org. Chem. 1995, 60,
1470. (b) Receveur, J.-M.; Guiramand, J.; Re´casens, M.; Roumestant, M.-
L.; Viallefont, P.; Martinez, J. Bioorg. Med. Chem. Lett. 1998, 8, 127.
(12) C.-P., Lo T. J. Org. Chem. 1961, 26, 3591.
(13) Liard, A.; Quiclet-Sire, B.; Zard, S. Z. Tetrahedron Lett. 1996, 37,
5877.
Org. Lett., Vol. 10, No. 15, 2008
3281

primary diamines are relatively uncommon. They are gener-
ally made by reduction of the corresponding dinitrile,16 and
it is necessary to perform the reduction sequentially to be
able to distinguish between the two amino groups.12 The
present approach is in contrast straightforward and concise.
Finally, polyamines can be produced by multiple additions
of reagent 17 to poly(1,2-butadiene) 55 leading to 56, which
was isolated by precipitation in 46%.17 Essentially, all the
terminal olefins in the polymer can be saturated.
Scheme 5
In summary, we have now in hand a very practical,
efficient, and quite general process for the regioselective
aminomethylation of alkenes. On a more fundamental level,
we have uncovered a hitherto unsuspected mechanism for
stabilizing an aminoalkyl radical that exploits the coupling
of the two carbonyls in an imide structure through mesomeric
forms such as 20b,c. The success of the present method and
all the consequent synthetic implications hinge on a small,
but nevertheless decisive, difference in stabilization energy
provided by the imide group. This effect should also apply,
in principle, to structures related to imides (e.g., imidazo-
linediones), but whether its magnitude in these cases will
be sufficient to tip the balance in the desired direction remains
to be seen.
using 2-propanol as the solvent. The possibility of obtaining
differentially protected diamine derivatives such as 32-34
is worth underscoring.
Acknowledgment. We dedicate this paper to the memory
of Jacques Levisalles and Jean-Louis Gras.
Other types of substituted diamines can be obtained as
shown in Scheme 5. Thus, addition to N-allylsulfanilide 48
can be followed by ring closure to the aromatic ring to give
indoline 50, a structure that is useful as a precursor of
melatonin.14 In the case of addition to sulfonamide 51,
acetylation of primary adduct 52 and further treatment with
lauroyl peroxide in isopropanol cause a shift of the aromatic
ring15 to give orthogonally protected diamine 54. Such
Supporting Information Available: Experimental pro-
1
cedures, full spectroscopic data, and copies of H and 13C
NMR spectra for all new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL801162M
(16) (a) Harley-Mason, J.; Jackson, A. H. J. Chem. Soc. 1954, 1165.
(b) White, J. D.; Yager, K. M.; Yakura, T. J. Am. Chem. Soc. 1994, 116,
1831. (c) Lebarbier, C.; Carreaux, F.; Carboni, B. Tetrahedron Lett. 1999,
40, 6233.
(14) Quiclet-Sire, B.; Sortais, B.; Zard, S. Z. Chem. Commun. 2002,
1692.
(17) Poly(1,2-butadiene), purchased from Aldrich, is constituted of 62
mol % 1,2-addition. In structure 55, partial cyclization is possible but is
not indicated for clarity.
(15) Georghe, A.; Quiclet-Sire, B.; Vila, X.; Zard, S. Z. Org. Lett. 2005,
7, 1653.
3282
Org. Lett., Vol. 10, No. 15, 2008