to ca. 8 mL of a 15% aqueous acidic solution), which could
negatively affect the formation and the stability of the
ketimine.
Table 1. Radical Addition of Formamide to Ketimines
Generated in Situ from PMP-NH2 1a and Cyclohexanone 2aa
To verify this hypothesis and improve progressively the
efficiency of the protocol, we added Zn metal (powder) to a
lower amount of TiCl3 solution (entry 2) and then completely
replaced TiCl3 with stoichiometric amounts of an anhydrous
TiCl4 solution in CH2Cl2. In doing so we succeeded in
increasing the yield up to 74% (entry 3), whereas in the
absence of Zn metal no reaction occurred (entry 4).
In a generic procedure, a homogeneous solution of
formamide (10 mL) containing 1a-c (2 mmol), 2a-v (12
mmol), and TiCl4 (2.5 mL of a 1 M CH2Cl2 solution, 2.5
mmol) was stirred at 0 °C under N2 atmosphere. After 30
min, Zn powder (300 mg, ca. 5 mmol) was suspended in
the reaction medium, and an aqueous 35 wt % H2O2 solution
(ca. 0.5 mL, 5 mmol), diluted in 4.5 mL of formamide, was
added dropwise over 3 h. Analogous results were observed
by reducing the amount of TiCl4 (1.5 mmol) in favor of the
cheaper Zn metal (8 mmol), after addition of a 37% HCl
aqueous solution (Table 1, entries 6-9). The reaction
proceeds like a titration, with periodic changes of color from
orange to violet, until a pale orange is barely maintained
also upon further addition of Zn. On the basis of these
observations we suggest the possible mechanism depicted
in Scheme 2.
While the changes of color prove the periodic variation
of the oxidation state of titanium ion in solution, the lack of
conversion in the desired products in the absence of titanium
salts (Table 1, entry 5) makes it obvious that zerovalent Zn
metal has the sole role to continuously convert Ti(IV)
(orange) to Ti(III) (violet) (paths i and iii). Instead, titanium
species play a multiple key role. Ti(III) acts both as radical
initiator, generating the hydroxyl radical (path ii), which is
in turn responsible for the hydrogen abstraction from
formamide (path iV), and as radical terminator, causing the
reduction of the aminium radical intermediate (path Vii). In
both cases, Ti(III) is reoxidized to Ti(IV), thus prolonging
the redox cycle and justifying the color oscillation. At the
same time, Ti(IV), as a Lewis acid, increases the electro-
Zn,
HCl 37%, 3a yield,
b
entry TiCl3, mmol TiCl4, mmol mmol
mL
%
1
2
3
4
5
6
7
8
9
8
2
23
47
74
5
5
2.5
2.5
5
5
5
8
8
1.5
1.5
1.5
1.5
33
72
71
60
1
1
2
a 2 mmol of 1a was reacted with 12 mmol of 2a in 10 mL of formamide.
Yield determined by H NMR with 2-methyl-benzylalcohol added as an
internal standard to the crude reaction mixture.
b
1
this lack in the literature can be ascribed to the lower
reactivity (due to the poor electrophilicity of the CdN bond)
combined with the lower stability of ketimines (anhydrous
conditions are usually required because they easily undergo
hydrolysis). However, free-radical addition to ketone-derived
imino compounds is particularly intriguing as it could provide
an unique route to the synthesis of tert-alkyl-amino-deriva-
tives not conventionally prepared via ionic chemistry.
We have recently reported that a novel system, based on
TiCl3 and hydroperoxides (t-BuOOH or H2O2), promotes
one-pot bond-forming transformations via the radical addition
of ethers,5 formamide,6 and alcohols7 to aldimines, generated
in situ under aqueous conditions.8 As a part of our ongoing
interest in this field, we wanted to verify whether the free-
radical addition to aldimines, promoted by our system in
formamide6 as a pivotal subtrate, could be also applied to
ketimines generated in situ for the one-pot synthesis of R,R-
disubstituted-R-aminoamides. This procedure would repre-
sent a convenient route to quaternary R-amino acids, which
are considered important building blocks in the design of
bioactive peptides with enhanced properties, as they are able
to introduce conformational constraints and consequent
structure stabilization.9
(3) For some leading references, see: (a) Ueda, M.; Miyabe, H.; Sugino,
H.; Naito, T. Org. Biol. Chem 2005, 3, 1124–1128. (b) Yamada, K.;
Yamamoto, Y.; Maekawa, M.; Akindele, T.; Umeki, H.; Tomioka, K. Org.
Lett. 2006, 8, 87–89. (c) Akindele, T.; Yamamoto, Y.; Maekawa, M.; Umeki,
H.; Yamada, K.; Tomioka, K. Org. Lett. 2006, 8, 5729–5732. (d) Yamada,
K.; Nakano, M.; Maekawa, Akindele, T.; Tomioka, K. Org. Lett. 2008, 10,
3805–3808.
(4) (a) Torrente, S.; Alonso, R. Org. Lett. 2001, 3, 1985–1987. (b)
Miyabe, H.; Yamaoka, Y.; Takemoto, Y. J. Org. Chem. 2005, 70, 3324–
3327. (c) Miyabe, H.; Yamaoka, Y.; Takemoto, Y. J. Org. Chem. 2006,
71, 2099–2106. (d) Friestad, G. K.; Ji, A. Org. Lett. 2008, 10, 2311–2313.
(5) Clerici, A.; Cannella, R.; Pastori, N.; Panzeri, W.; Porta, O.
Tetrahedron 2006, 62, 5986–5994.
Initial experiments, devoted to the optimization of the
protocol, were conducted under different operating conditions
by reacting p-methoxyaniline (PMP-NH2) 1a in the presence
of cyclohexanone 2a. Disappointingly, preliminary attempts
to extend the free radical addition to ketimines by using our
classic TiCl3/H2O2 system and HCONH2 gave poor results,
affording only 23% of the desired product 3a (Table 1, entry
1). We ascribed this behavior to the high amount of water
present in the reaction medium (8 mmol of TiCl3 corresponds
(6) Cannella, R.; Clerici, A.; Panzeri, W.; Punta, C.; Porta, O. J. Am.
Chem. Soc. 2006, 128, 5358–5359.
(7) (a) Clerici, A.; Ghilardi, A.; Pastori, N.; Punta, C.; Porta, O. Org.
Lett. 2008, 10, 5063–5066. (b) Spaccini, R.; Ghilardi, A.; Pastori, N.; Clerici,
A.; Punta, C.; Porta, O. Tetrahedron 2010, 66, 2044–2052.
(8) A review: Pastori, N.; Gambarotti, C.; Punta, C. Mini-ReV. Org.
Chem. 2009, 6, 184–195.
(9) For reviews in the field, see: (a) Gro¨ger, H. Chem. ReV. 2003, 103,
2795–2827. (b) Cativiela, C.; D´ıaz-de-Villegas, M. Tetrahedron: Asymmetry
2007, 18, 569–623. (c) Cativiela, C.; Ordo´n˜ez, M. Tetrahedron: Asymmetry
2009, 20, 1–63.
Org. Lett., Vol. 12, No. 17, 2010
3899