Stereoselective aminoxylation of biradical titanium enolates with TEMPO

Article abstract of DOI:10.1002/chem.201402127

A highly efficient and straightforward aminoxylation of titanium(IV) enolates from (S)-N-acyl-4-benzyl-5,5-dimethyl-1,3-oxazolidin-2-ones with TEMPO has been developed. A wide array of functional groups on the acyl moiety, including alkyl and aryl substituents, olefins, esters, or α-cyclopropyl, as well as α-trifluoromethyl groups, are well tolerated. This transformation can therefore produce the α-aminoxylated adducts in excellent yields with high diastereomeric ratios (d.r.). In turn, parallel additions to the α,β-unsaturated N-acyl counterparts give the corresponding γ-adducts with complete regioselectivity in moderate to good yields. Removal of the piperidinyl moiety or the chiral auxiliary converts the resultant adducts into enantiomerically pure α-hydroxy carboxyl derivatives, alcohols, or esters in high yields under mild conditions. Finally, a new mechanistic model based on the biradical character of the titanium(IV) enolates has been proposed.

Full text of DOI:10.1002/chem.201402127

DOI: 10.1002/chem.201402127
Full Paper
&
Synthetic Methods
Stereoselective Aminoxylation of Biradical Titanium Enolates with
TEMPO
Alejandro Gꢀmez-Palomino,[a] Miquel Pellicena,[a] Juan Manuel Romo,[a] Ricard Solꢁ,[a]
Pedro Romea,*[a] Fꢂlix Urpꢃ,*[a] and Mercꢂ Font-Bardia[b]
Abstract: A highly efficient and straightforward aminoxyla-
tion of titanium(IV) enolates from (S)-N-acyl-4-benzyl-5,5-di-
methyl-1,3-oxazolidin-2-ones with TEMPO has been devel-
oped. A wide array of functional groups on the acyl moiety,
including alkyl and aryl substituents, olefins, esters, or a-cy-
clopropyl, as well as a-trifluoromethyl groups, are well toler-
ated. This transformation can therefore produce the a-ami-
noxylated adducts in excellent yields with high diastereo-
meric ratios (d.r.). In turn, parallel additions to the a,b-unsa-
turated N-acyl counterparts give the corresponding g-ad-
ducts with complete regioselectivity in moderate to good
yields. Removal of the piperidinyl moiety or the chiral auxili-
ary converts the resultant adducts into enantiomerically
pure a-hydroxy carboxyl derivatives, alcohols, or esters in
high yields under mild conditions. Finally, a new mechanistic
model based on the biradical character of the titanium(IV)
enolates has been proposed.
Introduction
action followed by stereoselective free radical reduction of the
resultant a-bromo or a-seleno esters.[5] Apart from this reaction
path, the classical dimerization of metal enolates[6–8] has recent-
ly been updated and some ingenious methods based on the
oxidation of metal enolates and subsequent homo- as well as
heterocoupling of the resultant enolyl radicals have been de-
vised.[9] This radical chemistry was further advanced through
the introduction by MacMillan’s group of SOMO-organocataly-
sis concepts, whereby one-electron oxidation of a transient
chiral enamine derived from an aldehyde provides a cation
radical that can undergo highly enantioselective transforma-
tions.[10] Despite the tremendous advancement in asymmetric
synthesis facilitated by these ideas, the requirement of a stoi-
chiometric amount of an oxidant to generate the reactive in-
termediate is a major drawback in terms of atom economy.[11]
This hurdle has occasionally been overcome by merging pho-
toredox catalysis with organocatalysis.[12] Indeed, upon irradia-
tion, ruthenium(II) photoredox catalysts trigger the formation
of the enamine cation radical, which then reacts with other
radical intermediates produced by the ruthenium(I) species.
Such an ingenious combination of two independent catalytic
cycles permits the enantioselective intermolecular a-alkylation
of aldehydes without the need for an additional oxidant.[13,14]
In this context, we revealed the unconventional biradical
character of the titanium(IV) enolates,[15] which might mean
that they can participate directly in homolytic transformations
without any additional reagents in a highly economic manner.
Zakarian proved the feasibility of this new reaction paradigm
by developing the radical haloalkylation of the titanium(IV)
enolates of chiral N-acyl oxazolidinones catalyzed by rutheni-
um(II) complexes.[16a] Later, the method was expanded to zirco-
nium(IV) enolates,[16b] and it was also found that it could be
carried out using catalytic amounts of TiCl4.[16c] Thus, the biradi-
The development over the last decades of highly chemo-,
regio-, and stereoselective procedures for the enolization of
carbonyl compounds has meant that metal enolates are now
among the most important carbon nucleophiles. This has
paved the way for the use of metal enolates in a wide array of
organic transformations and, nowadays, a significant number
of stereoselective bond-forming reactions can only be under-
stood through considering the contribution of lithium, boron,
titanium(IV), or tin(II) enolates as structurally defined and very
reactive nucleophilic species.[1] Running parallel to this hetero-
lytic profile, attention has also been focused on the exploita-
tion of the homolytic reactivity of a-carbonyl radicals (enolyl
radicals), which can participate in highly stereoselective trans-
formations.[2] To date, a-halo carbonyl compounds have com-
monly been used as the source of such intermediates. For in-
stance, Sibi disclosed highly stereocontrolled radical alkylations
of chiral a-bromo N-acyl oxazolidinones,[3] and Porter reported
related transformations promoted by chiral Lewis acids.[4] In
turn, Guindon has developed a general strategy for polypropi-
onate synthesis based on a sequence of a Mukaiyama aldol re-
[a] A. Gꢀmez-Palomino, Dr. M. Pellicena, J. M. Romo, R. Solꢁ, Prof. P. Romea,
Prof. F. Urpꢂ
Departament de Quꢂmica Orgꢁnica, Universitat de Barcelona
Carrer Martꢂ i Franquꢃs 1–11, 08028 Barcelona, Catalonia (Spain)
Fax: (+34)93-3397878
E-mail: pedro.romea@ub.edu
felix.urpi@ub.edu
[b] Dr. M. Font-Bardia
Unitat de Difracciꢀ de RX, CCiTUB, Universitat de Barcelona
Carrer Solꢃ i Sabarꢂs 1–3, 08028 Barcelona, Catalonia (Spain)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201402127.
Chem. Eur. J. 2014, 20, 10153 – 10159
10153
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
cal character of titanium(IV) enolates could be an excellent
platform from which to take advantage of the radical chemis-
try of enolyl-like intermediates.[17]
These precedents and our ongoing interest in the reactivity
of the titanium(IV) enolates[18] led us to explore new stereose-
lective transformations in which they could act as radicals.[19]
Initially, we chose the 2,2,6,6-tetramethylpiperidine-N-oxyl radi-
cal (TEMPO), a commercially available stable free radical,[20] to
probe this new reactivity that might deliver a-oxygenated car-
bonyl compounds in a straightforward and stereocontrolled
manner. a-Hydroxylated carbonyl compounds are traditionally
prepared by treating the corresponding enolates with a variety
of oxidizing reagents, including oxygen, peroxides, hypervalent
iodine complexes, and chiral N-sulfonyloxazaridines.[21] More re-
cently, this set of procedures has been enlarged through the
addition of several enantioselective organocatalytic methods
based on oxidation with nitrosobenzene,[22,23] peroxides,[24]
oxygen,[25] and TEMPO.[26] Unfortunately, these methods can
only be applied to aldehydes, so the quest for more general
and efficient methods for the synthesis of a-hydroxy carbonyl
compounds remains active.
Scheme 1. a-Aminoxylation of titanium enolates from ethyl ketones.
ing on the use of TiCl3(iPrO) or TiCl4 as the titanium(IV) Lewis
acids, respectively (Eq. (1) in Scheme 1). Encouraged by these
findings, we then applied the same experimental conditions to
(S)-2-benzyloxy-3-pentanone (3), a lactate-derived chiral ketone
(Eq. (2) in Scheme 1). Unfortunately, this substrate-controlled
transformation yielded only small amounts of a-aminoxylated
adduct 4 and, most significantly, with only moderate diastereo-
selectivity (d.r. 75:25). None of our efforts to improve these re-
sults were successful. Nevertheless, the study demonstrated
the feasibility of the reaction and indicated that electron-do-
nating ligands bound to the metal atom lower the reactivity of
the resultant enolates and should therefore be avoided. More-
over, the moderate 1,3-asymmetric induction observed for a-
benzyloxy ketone 3 suggested that the reaction proceeds
through an open transition state in which the C1-methyl
group hardly differentiates between the two faces of the
enolate.[35]
The direct oxidation of enolates with TEMPO could meet
these challenges, but a-aminoxylation of metal enolates with
TEMPO usually requires the generation of N-oxoammonium
salts in situ through the use of an external oxidant.[27–29]
Renaud and Studer overcame this constraint by using ketone-
derived catecholboron enolates, the reaction of which with
TEMPO gave the corresponding a-carbonyl radicals, which
could then be trapped with a second equivalent of TEMPO to
provide the a-aminoxylated carbonyl products.[30] Jahn has
convincingly proved that the single-electron oxidation of lithi-
um enolates from ketones, esters, and amides with ferroceni-
um hexafluorophosphate (Cp2FePF6) affords the desired enolyl
radicals, which then couple with TEMPO in a straightforward
manner.[31] Finally, Li has recently proposed that copper-cata-
lyzed a-aminoxylation of a-alkoxy ketones with TEMPO may
also proceed through enolyl radicals.[32] All of these methods
provide the a-aminoxylated derivatives in good yields, but
their lack of stereocontrol thwarts advanced synthetic applica-
tions. This restriction has been successfully addressed by Za-
karian, who has reported the asymmetric radical addition of
TEMPO to titanium(IV) enolates from chiral oxazolidinones.[33]
This very recent advance has prompted us to disclose herein
our findings, which confirm Zakarian’s results and show that ti-
tanium(IV) enolates from a wide range of carbonyl and carbox-
yl compounds react under very mild conditions with TEMPO to
provide the corresponding a-aminoxylated compounds in
good yields and with moderate to excellent diastereo-
selectivities.
It was thus clear that the asymmetric a-aminoxylation of ti-
tanium enolates represented a genuine opportunity, provided
that sufficiently reactive and structurally rigid enolates were
used. These two requirements led us to center our attention
on titanium enolates from N-propanoyl oxazolidinone 5a (see
Table 1), because they had already permitted highly stereocon-
Table 1. Preliminary studies of the addition of TEMPO to the titanium
enolate of N-propanoyl oxazolidinone, 5a.
[b]
TEMPO[a]
T [8C]
t [h]
d.r.
Yield [%][c]
1
2
3
4
5
6
7
8
9
1.2
1.2
1.7
2.2
2.2
2.2
2.2
2.1
2.1
20
20
20
20
À78
À20
0
2
15
2
2
2
2
2
1
0.5
94:6
93:7
94:6
94:6
92:8
94:6
94:6
94:6
94:6
46
41
77
94
58
90
94
93
91
Results and Discussion
Our experience with substrate-controlled aldol reactions[18,34]
initially led us to assess the addition of TEMPO to different tita-
nium(IV) enolates of 2-methyl-3-pentanone (1). We were
pleased to observe that the addition occurred and produced
the desired a-aminoxylated adduct 2 in 15–47% yield, depend-
0
0
[a] Equivalents of TEMPO. [b] Established by 1H NMR analysis of the reac-
tion mixture. [c] Overall isolated yield.
Chem. Eur. J. 2014, 20, 10153 – 10159
www.chemeurj.org
10154
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
trolled additions of radicals to putative enolyl intermedi-
ates.[3,16,36] Exploratory experiments confirmed that our choice
was right. Indeed, treatment of the TiCl4-mediated enolate of
5a with 1.2 equivalents of TEMPO at room temperature for 2 h
afforded a-aminoxylated adduct 6a in 46% yield with a 94:6
diastereomeric ratio (see entry 1 in Table 1). Remarkably, the
conversion proved to be closely related to the amount of
TEMPO used, and it was finally established that an additional
equivalent is necessary to obtain 6a in a yield up to 94%
(compare entries 1–4 in Table 1). Low temperatures did not im-
prove the diastereoselectivity, and 6a was consistently ob-
tained with an excellent diastereomeric ratio at both room
temperature and À788C (compare entries 4–7 in Table 1)
whereas the addition progressed very rapidly at 08C to the
point that 6a was isolated in 91–94% yield after 0.5–2 h (com-
pare entries 7–9 in Table 1).[37] In light of these results and
taking into account the need for a general procedure, we
chose the conditions summarized in entry 8 as optimal; 6a
was obtained in a highly stereocontrolled manner (d.r. 94:6) in
93% yield by treating the TiCl4-enolate of 5a with 2.1 equiva-
lents of TEMPO for 1 h at 08C.
Next, we examined the scope of the reaction by applying
the optimal conditions to a wide range of N-acyl oxazolidi-
nones 5. The results, summarized in Scheme 2, show that a-
aminoxylated adducts 6 were consistently obtained in a highly
chemoselective and stereoselective manner. Indeed, N-acyl ox-
azolidinones 5a–d produced excellent yields and diastereose-
lectivities, irrespective of the steric bulk of R. Moreover, the
presence of functional groups such as an olefin, a methyl ester,
or a cyclopropyl ring in oxazolidinones 5e–g was well tolerat-
ed, and the corresponding adducts 6e–g were isolated in 94–
95% yield with diastereomeric ratios higher than 93:7
(Scheme 2). Conversely, N-trifluoropropanoyl oxazolidinone 5h
did not produce the desired adduct under the general condi-
tions, probably because of the instability of the titanium eno-
late at 08C.[38] Thus, we were pleased to isolate adduct 6h in
65% yield as a single diastereomer (d.r. > 95:5)[39] by carrying
out the enolization at À788C and allowing the reaction mix-
ture to slowly warm for 1 h. This provides new and enantiose-
lective access to small chiral fragments containing a trifluoro-
methyl group that could be very useful in medicinal
chemistry.[40]
Moreover, the reactions of a,b-unsaturated N-acyl oxazolidi-
nones 5i and 5j turned out to be completely regioselective
and the g-adducts 6i and 6j were isolated in moderate to
good yields without observing the formation of the alternative
a-OTEMP adducts (Scheme 3).[41] Unfortunately, the g-aminoxyl-
ated adduct 6j was obtained as a 1:1 mixture of (E)-a,b-unsa-
turated diastereomers.[42] This lack of stereocontrol may be due
to the conformational flexibility of the b,g-conjugated enolate
or the inability of the C4-benzyl group to shield the upper face
of the p-system at the g-position.
The configuration of these a-aminoxylated adducts was es-
tablished through X-ray analysis of 6b (Figure 1).[43] It was later
confirmed by removal of the chiral auxiliary from 6a and the
correlation of the resultant alcohol (see ref. [53]).
Having applied the asymmetric addition of TEMPO to a wide
range of N-acyl oxazolidinones, 5, we next studied the origin
of this successful transformation. Initially, we surveyed other
common chiral auxiliaries to gain a better understanding of
the structural elements that determine the stereochemical out-
come of the process. We were especially interested in deter-
mining the influence of geminal methyl groups at C5 and the
outcome of the reaction using a sulfur-based chiral auxiliary.
Therefore, we evaluated the additions of TEMPO to the titani-
um enolates from chiral N-propanoyl oxazolidinone 7 and
Scheme 2. Additions of TEMPO to the titanium enolates of N-acyl oxazolidinones 5a–h.
Chem. Eur. J. 2014, 20, 10153 – 10159 www.chemeurj.org
10155
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 3. Additions of TEMPO to the titanium enolates of a,b-unsaturated N-acyl oxazolidinones 5i,j.
proceeds through a transition state in which the benzyl group
at the chiral auxiliary hinders the a-approach of TEMPO to the
Re face of the corresponding chelated titanium enolate. More-
over, the absolute regioselectivity observed for the aminoxyla-
tion of 5i and 5j suggests orbital-controlled addition, as
occurs in Lewis acid-mediated vinylogous additions of electro-
philes to conjugated silyl enol ethers.[45] Without detracting
from the importance of these features, the crux of such addi-
tions is the titanium atom.[46] In this context, the poor oxidative
capacity of titanium(IV) complexes[47] and the need for two
equivalents of TEMPO indicate that it is unlikely that the titani-
um enolates from 5 would oxidize TEMPO to generate the cor-
responding oxoammonium salt. In contrast, this addition
might proceed through a radical pathway in which the biradi-
cal character of titanium enolates represented by II (Figure 2)
would play a crucial role.
Figure 1. X-ray crystal structure of 6b. Hydrogen atoms at non-stereogenic
centers have been removed for clarity.
Figure 2. Titanium enolate derived from 5.
Scheme 4. Additions of TEMPO to titanium enolates from 7 and 8.
The way the enolate form II acts cannot be understood by
overlapping the reactivity of two radicals; rather, we require
a valence tautomeric[17,48,49] paradigm that interweaves the TiIII/
TiIV and the enolyl-like chemistries. One of the crucial issues re-
lated to this model involves the beginning of the reaction se-
quence. This might take advantage of the reducing capacity of
TiIII to produce an enolyl radical, which would subsequently be
trapped by a TEMPO radical in the classical way. Conversely,
the initial attack of a TEMPO radical at the a-position (or the g-
position in a,b-unsaturated N-acyl oxazolidinones) of II might
be followed by the trapping of the resultant TiIII complex with
a second molecule of TEMPO. Zakarian favored the former
pathway based on its mechanistic similarity with catecholbor-
on enolate oxidation by TEMPO, as reported by Studer and
thiazolidinethione 8 (Scheme 4). The former gave the a-
aminoxylated adduct 9 in 93% yield, but in a less stereocon-
trolled manner (d.r. 78:22) than 5a. This proves the key role
played by the geminal methyl groups at C5. In turn, thiazoli-
dinethione 8 produced the corresponding adduct 10 in similar
yield and with comparable diastereoselectivity to 5a (see
entry 5 in Table 1). Thus, replacement of the oxygen-based
chiral auxiliary by a sulfur-based moiety did not modify the
outcome of the addition. Hence, the thiazolidinethione chiral
auxiliary represents an appealing alternative because of its
easy removal under very mild conditions.[44]
As for ketones, all of these results suggest that the a-amino-
xylation of the titanium enolates from N-acyl oxazolidinones 5
Chem. Eur. J. 2014, 20, 10153 – 10159
www.chemeurj.org
10156
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Renaud,[30] together with indirect experimental evidence, pri-
marily isolation of enolate dimerization by-products from cer-
tain reactions.[33] We have never observed the titanium eno-
lates from 5 to induce such dimerizations. In fact, when the ti-
tanium enolate from 5a was stirred at 08C for one day,
quenched, and analyzed, the starting material was quantita-
tively recovered and the 1H NMR spectrum was completely
clear without any other peak. Furthermore, crude reaction mix-
tures from 5g bearing a cyclopropyl ring (radical clock)[50] also
produced clear spectra. Had classical radicals been involved in
these transformations, other aminoxylated isomers would have
been observed.[51]
ed alcohol 12a in 63% yield in addition to a 93% recovery of
the chiral auxiliary 13 (Scheme 6).[53] Unfortunately, such condi-
tions were unsuitable for more hindered adducts such as 6b,
and long reaction times were necessary to attain moderate
yields. Eventually, LiAlH4 turned out to be the most appropriate
reducing agent, affording both 12b and 13 in quantitative
yields (Scheme 6). In turn, treatment of 6a with NH3/MeOH
provided methyl ester 14a in 92% yield in a straightforward
manner under very mild conditions (Scheme 6).[54]
Conclusion
All together, these results suggest that a-carbonyl radicals
are not true intermediates in the process. Instead, experimen-
tal evidence supports a sequence as described in Scheme 5, in
The aminoxylation of titanium(IV) enolates with TEMPO is
a general transformation that proceeds under mild conditions.
In particular, chiral N-acyl-4-benzyl-5,5-dimethyl-1,3-oxazolidin-
2-ones are an appealing plat-
form for obtaining a-aminoxylat-
ed adducts in excellent yields
and diastereomeric ratios. Acyl
components containing olefin,
ester, cyclopropyl, or trifluoro-
methyl groups are well tolerated,
which confers this reaction with
a broad chemoselective profile.
Furthermore, the a,b-unsaturat-
ed N-acyl counterparts afford
the corresponding g-aminoxylat-
ed adducts with outstanding re-
gioselectivity in moderate to
good yields. Therefore, the addi-
tion of TEMPO to the TiCl4-eno-
Scheme 5. Plausible mechanism for the addition of TEMPO to titanium enolates derived from 5.
lates from a wide array of N-acyl-
which the initial Ca addition of TEMPO is followed by the in-
teraction of the resultant TiIII complex with a second molecule
of TEMPO. The initial addition of TEMPO appears to be very
fast and involves a low activation barrier that prevents us from
improving the diastereoselectivity by carrying out the reaction
at low temperatures. In turn, the fate of the resultant titanium-
(III) complex is at present unknown. We speculate that it may
involve a single-electron transfer to produce the corresponding
anion of the hydroxylamine form or the formation of a new
formal TiIV complex in which TEMPO acts as a ligand.
4-benzyl-5,5-dimethyl-1,3-oxazolidin-2-ones produces the ami-
noxylated adducts in a highly efficient and straightforward
manner.
Both the piperidine moiety and the chiral auxiliary can be re-
moved to provide enantioselectively pure a-hydroxy N-acyl ox-
azolidinones, alcohols, or esters in high yields.
Finally, a mechanistic model based on the biradical character
of titanium enolates has been proposed to account for the
Finally, we examined the removal of the piperidine moiety
and the chiral auxiliary. Regarding the first issue, we applied
a common procedure reported in the literature, based on the
use of zinc in hot AcOH or 3:1:1 AcOH/THF/H2O.[52] These con-
ditions turned out to be too harsh for model 6a, producing
a remarkable Ca-epimerization and partial hydrolysis of the re-
sultant hydroxy derivative 11 a (see entries 1 and 2 in Table 2).
Instead, the reduction proceeded smoothly over 6 h in AcOH/
THF/H2O (3:1:1) at room temperature, which permitted us to
isolate diastereomerically pure 11 a in 78% yield and certain
amounts of the chiral auxiliary (compare entries 3 and 4 in
Table 2).
Table 2. Reduction of the OÀN bond in 6a.
[a]
Solvent
T [8C]
t [h]
d.r.
Yield [%][b]
1
2
3
4
AcOH
50
50
20
20
1.5
1.5
19
42:58
32:68
93:7
27 (37)
20 (44)
66 (6)[c]
78 (7)[d]
3:1:1 AcOH/THF/H2O
3:1:1 AcOH/THF/H2O
3:1:1 AcOH/THF/H2O
6
94:6
[a] Established by 1H NMR analysis of the reaction mixture. [b] Isolated
yield of 11 a. In parentheses, isolated yield of the other diastereomer.
[c] 12% of the chiral auxiliary was isolated. [d] 5% of the chiral auxiliary
was isolated.
Simultaneously, we focused our efforts on the removal of
the chiral auxiliary from adducts 6. Thereby, the treatment of
6a with NaBH4 in THF/H2O at room temperature for 4 h afford-
Chem. Eur. J. 2014, 20, 10153 – 10159
www.chemeurj.org
10157
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 6. Removal of the chiral auxiliary from 6.
8501; b) J.-F. Brazeau, P. Mochirian, M. Prꢆvost, Y. Guindon, J. Org. Chem.
2009, 74, 64–74; c) F. Godin, M. Prꢆvost, F. Viens, P. Mochirian, J.-F. Bra-
zeau, S. I. Gorelsky, Y. Guindon, J. Org. Chem. 2013, 78, 6075–6103.
[6] For pioneering studies, see: a) M. W. Rathke, A. Lindert, J. Am. Chem.
Soc. 1971, 93, 4605–4606; b) Y. Ito, T. Konoike, T. Harada, T. Saegusa, J.
Am. Chem. Soc. 1977, 99, 1487–1493.
outcomes of these additions, which might be an example of
a new reaction paradigm.
Experimental Section
[7] For examples of stereoselective oxidative homocouplings of enolates,
see: a) N. Kise, T. Ueda, K. Kumada, Y. Terao, N. Ueda, J. Org. Chem.
2000, 65, 464–468; b) P. Q. Nguyen, H. J. Schꢇfer, Org. Lett. 2001, 3,
2993–2995.
[8] For a review, see: A. G. Csꢈky¨, J. Plumet, Chem. Soc. Rev. 2001, 30, 313–
320.
[9] a) P. S. Baran, M. P. DeMartino, Angew. Chem. 2006, 118, 7241–7244;
Angew. Chem. Int. Ed. 2006, 45, 7083–7086; b) J. M. Richter, B. W. White-
field, T. J. Maimone, D. W. Lin, M. P. Castroviejo, P. S. Baran, J. Am. Chem.
Soc. 2007, 129, 12857–12869; c) M. P. DeMartino, K. Chen, P. S. Baran, J.
Am. Chem. Soc. 2008, 130, 11546–11560.
[10] a) T. D. Beeson, A. Mastracchio, J.-B. Hong, K. Ashton, D. W. C. MacMillan,
Science 2007, 316, 582–585; b) H.-Y. Jang, J.-B. Hong, D. W. C. MacMil-
lan, J. Am. Chem. Soc. 2007, 129, 7004–7005; c) H. Kim, D. W. C. MacMil-
lan, J. Am. Chem. Soc. 2008, 130, 398–399; d) N. T. Jui, E. C. Y. Lee,
D. W. C. MacMillan, J. Am. Chem. Soc. 2010, 132, 10015–10017.
[11] a) B. M. Trost, Science 1991, 254, 1471–1477; b) T. Newhouse, P. S. Baran,
R. W. Hoffmann, Chem. Soc. Rev. 2009, 38, 3010–3021.
General procedure: Neat TiCl4 (60 mL, 0.55 mmol) was added drop-
wise to a solution of N-acyl oxazolidinone 5 (0.50 mmol) in CH2Cl2
(2.0 mL) at 08C. The resulting mixture was stirred for 5 min and
then anhydrous iPr2NEt (96 mL, 0.55 mmol) was added dropwise.
The resulting solution was stirred for 40 min at 08C. A solution of
TEMPO (164 mg, 1.05 mmol) in CH2Cl2 (0.25 mL) was then added
via a cannula (1ꢅ0.25 mL) and stirring was continued for 1 h at
08C. The reaction was then quenched by the addition of saturated
aqueous NH4Cl solution (2 mL) and the mixture was vigorously
stirred at rt for 10 min. It was then partitioned between CH2Cl2
(10 mL) and H2O (10 mL). The layers were separated and the aque-
ous layer was extracted with CH2Cl2 (10 mL). The combined organic
phases were washed with brine (10 mL), dried (MgSO4), and con-
centrated. The residue was analyzed by 1H NMR and purified by
flash column chromatography to afford the aminoxylated adduct
6.
[12] a) D. A. Nicewicz, D. W. C. MacMillan, Science 2008, 322, 77–80; b) D. A.
Nagib, M. E. Scott, D. W. C. MacMillan, J. Am. Chem. Soc. 2009, 131,
10875–10877.
[13] a) P. Renaud, P. Leong, Science 2008, 322, 55–56; b) P. Melchiorre,
Angew. Chem. 2009, 121, 1386–1389; Angew. Chem. Int. Ed. 2009, 48,
1360–1363.
[14] For recent advances in this area, see: a) M. Neumann, S. Fꢉldner, B.
Kçnig, K. Zeitler, Angew. Chem. 2011, 123, 981–985; Angew. Chem. Int.
Ed. 2011, 50, 951–954; b) G. Cecere, C. M. Kçnig, J. L. Alleva, D. W. C.
MacMillan, J. Am. Chem. Soc. 2013, 135, 11521–11524; c) F. R. Petroni-
jevic, M. Nappi, D. W. C. MacMillan, J. Am. Chem. Soc. 2013, 135, 18323–
18326.
Acknowledgements
Financial support from the Spanish Ministerio de Economꢃa y
Competitividad (Grant No. CTQ2012–31034) and the Generali-
tat de Catalunya (2009SGR825) as well as doctorate student-
ships to M.P. and J.M.R. (FPU, Ministerio de Educaciꢀn) are ac-
knowledged.
[15] I. de P. R. Moreira, J. M. Bofill, J. M. Anglada, J. G. Solsona, J. Nebot, P.
Romea, F. Urpꢃ, J. Am. Chem. Soc. 2008, 130, 3242–3243.
Keywords: aminoxylation · asymmetric synthesis · radical
reactions · TEMPO · titanium
[16] a) S. Beaumont, E. A. Ilardi, L. R. Monroe, A. Zakarian, J. Am. Chem. Soc.
2010, 132, 1482–1483; b) A. T. Herrmann, L. L. Smith, A. Zakarian, J. Am.
Chem. Soc. 2012, 134, 6976–6979; c) Z. Gu, A. T. Herrmann, A. Zakarian,
Angew. Chem. 2011, 123, 7274–7277; Angew. Chem. Int. Ed. 2011, 50,
7136–7139.
[1] E. M. Carreira, L. Kvaerno, Classics in Stereoselective Synthesis, Wiley-VCH,
Weinheim, 2009.
[2] a) G. Bar, A. F. Parsons, Chem. Soc. Rev. 2003, 32, 251–263; b) Y.-H. Yang,
M. P. Sibi in Encyclopedia of Radicals in Chemistry, Biology and Materials
(Eds.: C. Chatgilialoglu, A. Studer), 2012, 2, 655–691.
[3] M. P. Sibi, J. Ji, Angew. Chem. 1996, 108, 198–200; Angew. Chem. Int. Ed.
Engl. 1996, 35, 190–192.
[4] N. A. Porter, J. H. Wu, G. Zhang, A. D. Reed, J. Org. Chem. 1997, 62,
6702–6703.
[5] a) Y. Guindon, K. Houde, M. Prꢆvost, B. Cardinal-David, S. R. Landry, B.
Daoust, M. Bencheqroun, B. Guꢆrin, J. Am. Chem. Soc. 2001, 123, 8496–
[17] T. Amatov, U. Jahn, Angew. Chem. 2011, 123, 4636–4638; Angew. Chem.
Int. Ed. 2011, 50, 4542–4544.
[18] For recent contributions, see: a) J. Zambrana, P. Romea, F. Urpꢃ, Chem.
Commun. 2013, 49, 4507–4509; b) S. Alcoberro, A. Gꢀmez-Palomino, R.
Solꢁ, P. Romea, F. Urpꢃ, M. Font-Bardia, Org. Lett. 2014, 16, 584–587.
[19] M. Pellicena, Ph.D. Thesis, Universitat de Barcelona, January 2014.
[20] For the use of TEMPO and other nitroxides in synthesis, see: a) T.
Vogler, A. Studer, Synthesis 2008, 1979–1993; b) L. Tebben, A. Studer,
Chem. Eur. J. 2014, 20, 10153 – 10159
www.chemeurj.org
10158
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Angew. Chem. 2011, 123, 5138–5174; Angew. Chem. Int. Ed. 2011, 50,
5034–5068.
[37] Zakarian has reported that the conversion is quantitative after 5 min at
room temperature (see ref. [33]).
[21] a) B.-C. Chen, P. Zhou, F. A. Davis, E. Ciganek, Org. React. 2003, 62, 4–
356; b) A. M. R. Smith, K. K. Hii, Chem. Rev. 2011, 111, 1637–1656.
[22] For seminal contributions, see: a) G. Zhong, Angew. Chem. 2003, 115,
4379–4382; Angew. Chem. Int. Ed. 2003, 42, 4247–4250; b) S. P. Brown,
M. P. Brochu, C. J. Sinz, D. W. C. MacMillan, J. Am. Chem. Soc. 2003, 125,
10808–10809; c) A. Cꢀrdova, H. Sundꢆn, A. Bøgevig, M. Johansson, F.
Himo, Chem. Eur. J. 2004, 10, 3673–3684; d) Y. Hayashi, J. Yamaguchi, T.
Sumiya, K. Hibino, M. Shoji, J. Org. Chem. 2004, 69, 5966–5973; e) M.
Baidya, K. A. Griffin, H. Yamamoto, J. Am. Chem. Soc. 2012, 134, 18566–
18569.
[38] Defluorination of a-CF3 enolates has traditionally limited the use of
these intermediates, but a few examples of the successful generation
of titanium enolates from a-CF3 carbonyl compounds have been re-
ported: a) Y. Itoh, M. Yamanaka, K. Mikami, J. Am. Chem. Soc. 2004, 126,
13174–13175; b) Y. Itoh, K. Mikami, Org. Lett. 2005, 7, 649–651; c) T.
Shimada, M. Yoshioka, T. Konno, T. Ishihara, Org. Lett. 2006, 8, 1129–
1131; d) X. Franck, B. Seon-Meniel, B. Figadꢂre, Angew. Chem. 2006, 118,
5298–5300; Angew. Chem. Int. Ed. 2006, 45, 5174–5176.
[39] Only one diastereomer was identified in the reaction mixtures by NMR
analysis.
[23] For a recent overview, see: P. Kumar, N. Dwivedi, Acc. Chem. Res. 2013,
46, 289–299.
[40] a) J. Nie, H.-C. Guo, D. Cahard, J.-A. Ma, Chem. Rev. 2011, 111, 455–529;
b) T. Furuya, A. S. Kamlet, T. Ritter, Nature 2011, 473, 470–477.
[41] For a related radical-mediated g-functionalization of a,b-unsaturated
carboxylic amides, see: S. Kim, C. J. Lim, Angew. Chem. 2004, 116, 5492–
5494; Angew. Chem. Int. Ed. 2004, 43, 5378–5380.
[42] The (E)-geometry of the resultant a,b-unsaturated carbonyl adducts 6i,j
was established by analysis of the diagnostic vicinal coupling constants,
3J.
[43] CCDC-984365 contains the supplementary crystallographic data for 6b.
The data can be obtained free of charge from the Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[44] E. Gꢈlvez, P. Romea, F. Urpꢃ, Org. Synth. 2009, 86, 81–91.
[45] S. E. Denmark, J. R. Heemstra, G. L. Beutner, Angew. Chem. 2005, 117,
4760–4777; Angew. Chem. Int. Ed. 2005, 44, 4682–4698.
[24] For recent reports of stereoselective reactions, see: a) T. Kano, H. Mii, K.
Maruoka, J. Am. Chem. Soc. 2009, 131, 3450–3451; b) M. J. P. Vaismaa,
S. C. Yau, N. C. O. Tomkinson, Tetrahedron Lett. 2009, 50, 3625–3627;
c) H. Gotoh, Y. Hayashi, Chem. Commun. 2009, 3083–3085; d) O. Lifchits,
N. Demoulin, B. List, Angew. Chem. 2011, 123, 9854–9857; Angew.
Chem. Int. Ed. 2011, 50, 9680–9683; e) C. Yin, W. Cao, L. Lin, X. Liu, X.
Feng, Adv. Synth. Catal. 2013, 355, 1924–1930.
[25] H. Sundꢆn, M. Engqvist, J. Casas, I. Ibrahem, A. Cꢀrdova, Angew. Chem.
2004, 116, 6694–6697; Angew. Chem. Int. Ed. 2004, 43, 6532–6535; for
a report on the oxidation of sodium enolates with molecular oxygen,
see: H. Lubin, A. Tessier, G. Chaume, J. Pytkowicz, T. Brigaud, Org. Lett.
2010, 12, 1496–1499.
[26] a) M. P. Sibi, M. Hasegawa, J. Am. Chem. Soc. 2007, 129, 4124–4125;
b) T. Kano, H. Mii, K. Maruoka, Angew. Chem. 2010, 122, 6788–6791;
Angew. Chem. Int. Ed. 2010, 49, 6638–6641; c) K. Akagawa, K. Kudo, Org.
Lett. 2011, 13, 3498–3501; d) S. P. Simonovich, J. F. Van Humbeck,
D. W. C. MacMillan, Chem. Sci. 2012, 3, 58–61.
[27] For examples of the aminoxylation of lithium enolates with TEMPO in
the presence of PhI(OAc)2, see K.-H. Ahn, Y. Kim, Synth. Commun. 1999,
29, 4361–4366.
[46] For example, the sodium enolate was completely unreactive.
[47] For studies on the reaction of titanium complexes with TEMPO, see:
a) M. K. Mahanthappa, K.-W. Huang, A. P. Cole, R. M. Waymouth, Chem.
Commun. 2002, 502–503; b) K.-W. Huang, J. H. Han, A. P. Cole, C. B.
Musgrave, R. M. Waymouth, J. Am. Chem. Soc. 2005, 127, 3807–3816.
[48] For the valence tautomerism concept, see: P. Gꢉtlich, A. Dei, Angew.
Chem. 1997, 109, 2852–2855; Angew. Chem. Int. Ed. Engl. 1997, 36,
2734–2736.
[28] The oxidation of sodium enolates with isolated 2,2,6,6-tetramethylpi-
peridine-1-oxoammonium tetrafluoroborate has also been reported,
see: M. Schꢇmann, H. J. Schꢇfer, Synlett 2004, 1601–1603.
[29] For a rare oxidation of electron-rich enols with TEMPO, see: J. Guin, S.
De Sarkar, S. Grimme, A. Studer, Angew. Chem. 2008, 120, 8855–8858;
Angew. Chem. Int. Ed. 2008, 47, 8727–8730.
[49] Valence tautomerism can be viewed as a manifestation of non-innocent
or redox-active ligands. For recent forums on this active area of re-
search in organometallic chemistry, see: a) P. J. Chirik, Inorg. Chem.
2011, 50, 9737–9740; b) K. Hindson, B. De Bruin, Eur. J. Inorg. Chem.
2012, 340–342.
[50] M. Newcomb, Tetrahedron 1993, 49, 1151–1176.
[30] a) M. Pouliot, P. Renaud, K. Schenk, A. Studer, T. Vogler, Angew. Chem.
2009, 121, 6153–6156; Angew. Chem. Int. Ed. 2009, 48, 6037–6040;
b) Y. Li, M. Pouliot, T. Vogler, P. Renaud, A. Studer, Org. Lett. 2012, 14,
4474–4477.
[31] a) U. Jahn, J. Org. Chem. 1998, 63, 7130–7131; b) U. Jahn, P. Hartmann,
I. Dix, P. G. Jones, Eur. J. Org. Chem. 2001, 3333–3355; c) E. Dinca, P.
Hartmann, J. Smrcek, I. Dix, P. G. Jones, U. Jahn, Eur. J. Org. Chem. 2012,
4461–4482.
[51] For a recent example, see: J. M. R. Narayanam, J. W. Tucker, C. R. J. Ste-
phenson, J. Am. Chem. Soc. 2009, 131, 8756–8757.
[52] D. L. Boger, R. M. Garbaccio, Q. Jin, J. Org. Chem. 1997, 62, 8875–8891.
[53] Comparison of the spectroscopic data for alcohol 12a with those re-
ported in ref. [26b] confirmed the configuration of the new stereocen-
ter.
[32] Y.-X. Xie, R.-J. Song, Y. Liu, Y.-Y. Liu, J.-N. Xiang, J.-H. Li, Adv. Synth. Catal.
2013, 355, 3387–3390.
[33] P. J. Mabe, A. Zakarian, Org. Lett. 2014, 16, 516–519.
[34] a) J. Nebot, P. Romea, F. Urpꢃ, J. Org. Chem. 2009, 74, 7518–7521; b) J.
Zambrana, P. Romea, F. Urpꢃ, C. Lujꢈn, J. Org. Chem. 2011, 76, 8575–
8587; c) M. Pellicena, K. Krꢇmer, P. Romea, F. Urpꢃ, Org. Lett. 2011, 13,
5350–5353; d) J. Esteve, C. Jimꢆnez, J. Nebot, J. Velasco, P. Romea, F.
Urpꢃ, Tetrahedron 2011, 67, 6045–6056.
[35] Enholm has reported a highly diastereoselective radical allylation of an
a-tert-butyl a-hydroxy ketone under chelation-controlled conditions, in
which the steric hindrance of the a-tBu group restricts the conforma-
tional freedom of the enolyl radical and hinders the approach of the in-
coming radical: E. J. Enholm, S. Lavieri, T. Cordꢀva, I. Ghiviriga, Tetrahe-
dron Lett. 2003, 44, 531–534.
[54] J. Bian, D. Blakemore, J. S. Warmus, J. Sun, M. Corbett, C. R. Rose, B. M.
Bechle, Org. Lett. 2013, 15, 562–565.
[36] For the synthesis and applications of 4-substituted 5,5-dimethyl-1,3-ox-
azolidin-2-one chiral auxiliaries, see: S. G. Davies, H. J. Sanganee, P.
Szolcsanyi, Tetrahedron 1999, 55, 3337–3354.
Received: February 11, 2014
Published online on July 7, 2014
Chem. Eur. J. 2014, 20, 10153 – 10159
www.chemeurj.org
10159
ꢄ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim