Enolizable Carbonyls and N,O-Acetals: A Rational Approach for Room-Temperature Lewis Superacid-Catalyzed Direct α-Amidoalkylation of Ketones and Aldehydes

Article abstract of DOI:10.1002/chem.201504772

An efficient catalytic room-temperature direct α-amidoalkylation of carbonyl donors, that is, ketones and aldehydes with unbiased N,O-acetals, is described. Sn(NTf₂)₄ is an optimal catalyst to promote this challenging transformation at low loading and the reaction shows promising scope. A comprehensive and rational evaluation of this reaction has led to the establishment of an empirical scale of nucleophilic reactivity for a broad set of ketones that should be helpful in the synthetic design and development of carbonyl α-functionalization methods. Magic Lewis superacid: A room-temperature catalytic direct α-amidoalkylation reaction of ketones and aldehydes that operates via N-acyliminium ions generated in situ is reported (see scheme). This reaction is efficiently catalyzed by 0.5-2 mol % of the Lewis superacid Sn(NTf₂)₄ and exhibits a large scope. An evaluation of this Mannich coupling led to the establishment of a reactivity scale of enolizable ketones.

Full text of DOI:10.1002/chem.201504772

DOI: 10.1002/chem.201504772
Full Paper
&
Alkylation
Enolizable Carbonyls and N,O-Acetals: A Rational Approach for
Room-Temperature Lewis Superacid-Catalyzed Direct
a-Amidoalkylation of Ketones and Aldehydes
Bahria Touati,[a, c] Abderrahman El Bouakher,[a] Catherine Taillier,[a] Raja Ben Othman,[b]
Malika Trabelsi-Ayadi,[c] Sylvain Antoniotti,[d] Elisabet DuÇach,[d] and Vincent Dalla*[a]
Abstract: An efficient catalytic room-temperature direct a-
amidoalkylation of carbonyl donors, that is, ketones and al-
dehydes with unbiased N,O-acetals, is described. Sn(NTf2)4 is
an optimal catalyst to promote this challenging transforma-
tion at low loading and the reaction shows promising scope.
A comprehensive and rational evaluation of this reaction has
led to the establishment of an empirical scale of nucleophilic
reactivity for a broad set of ketones that should be helpful
in the synthetic design and development of carbonyl a-func-
tionalization methods.
by an electron-withdrawing group.[12b,c] The extension to more
challenging pro-electrophiles such as the less reactive allylic
and propargylic alcohols, that is, precursors of more electro-
philic carbenium intermediates,[15] has been made possible
only by means of more powerful second-generation strategies
merging organocatalysis and metal catalysts.[7–9] Despite these
considerable advancements, some limitations persist. By way
of example, highly electrophilic carbenium ions, such as regu-
lar (electronically unbiased) benzhydrylium and secondary ben-
zylic cations generated in situ, have not yet succumbed to the
stereoselective alkylation, and their coupling has so far been
effected only under the influence of strong (achiral) Lewis
acidic catalysts,[14a,b] transition-metal-based catalysts,[14d] or me-
diators,[14c] sometimes under drastic conditions.[14b] As these
latter contributions show, beside the asymmetric alkylation of
well-stabilized carbenium ions, the evaluation of less-biased
counterparts, for example, carbocation precursors, which are
less prone to ionization[15] and/or which are likely to follow
competing pathways such as b-elimination, although of funda-
mental importance, is lacking. Heteroatom-stabilized carboca-
tions, namely oxonium ions[7c] and N-acyliminium ions,[16] are
synthetically very useful transient intermediates. Our interest in
developing catalytic methods for their functionalization,[17] mo-
tivated us to evaluate their potential as coupling partners for
unmodified ketones and aldehydes (Scheme 1).[18] The electro-
philicity parameters of N-acyliminium ions are not reported in
Mayr’s electrophilicity scale,[15] but these ions are expected to
stand as a whole in the top region of this scale,[19] thus sup-
porting the view that they constitute stimulating targets in the
catalytic alkylation with enolisable carbonyl nucleophiles. How-
ever, their clear structural modularity is assumed to translate
into variable electrophilicity inside this zone, which should
result in subtle differences in their reactivity (strength of the
acid catalyst to be used, catalyst loading, reaction kinetics).
This predication is illustrated with the approximate empirical
Introduction
The development of catalytic strategies for the direct a-alkyla-
tion of unmodified donor carbonyl components, that is, ke-
tones and aldehydes, is a current topic of great interest in or-
ganic synthesis.[1–14] The field was pioneered by the develop-
ment of purely organocatalytic methods,[2–6] but, despite the
unquestionable breakthroughs achieved as a result of these
studies, the available methods still suffer from unavoidable lim-
itations, and, in most cases, are only suitable for some specific
nucleophile/electrophile combinations. As far as SN1-type alky-
lations are concerned, only highly activated and therefore
somewhat biased pro-electrophiles were compatible, as
a result of the inherent weak acidity associated with the orga-
nocatalytic modes of activation used in these strategies.[3–6]
The same holds true with the photoredox approaches devel-
oped by MacMillan, which are limited to the alkylation of
capto-radicals that are generated in situ and that are stabilized
[a] B. Touati, Dr. A. El Bouakher, Dr. C. Taillier, Prof. Dr. V. Dalla
Normandie Univ., UNIHAVRE, CNRS, URCOM
76600 Le Havre (France)
Fax: (+33)2-32-74-43-91
E-mail: vincent.dalla@univ-lehavre.fr
[b] Dr. R. B. Othman
Ecole supØrieure des sciences et de technologie de Hammam Sousse
Rue Lamine Abassi 4011H Sousse (Tunisie)
[c] B. Touati, Prof. Dr. M. Trabelsi-Ayadi
Laboratoire d’Application de la Chimie aux Ressources et
Substances Naturelles et à l’Environnement-
LACReSNE-FacultØ des Sciences de Bizerte- 7021 Bizerte
UniversitØ de Carthage (Tunisie)
[d] Dr. S. Antoniotti, Dr. E. DuÇach
Institut de Chimie de Nice, UMR 7272, UniversitØ de Nice Sophia Antipolis
C.N.R.S., Parc Valrose, 06108 Nice cedex 2 (France)
Supporting information for this article and ORCID for one of the authors
are available on the WWW under http://dx.doi.org/10.1002/
chem.201504772.
Chem. Eur. J. 2016, 22, 6012 – 6022
6012
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
needs of this study, this scale was completed with inclusion of
N-carbamoyl quinolinium/isoquinolinium salts[20] and their N-
aryl iminium[21] counterparts, which we positioned with regard
to our ordinary N-acyliminiums, on the basis of the qualitative
analysis briefly stated previously[20] and discussed in more
detail as follows.
The empirical scale of reactivity became a useful guide that
could be used to choose, in a rational fashion, numerous N,O-
acetal/nucleophile combinations that could be applied in a di-
verse range of studies; this strategy proved to be particularly
relevant in the present case. Our reactivity-guided design strat-
egy was based on the following considerations: It appeared
clear that the N-acyl (iso)quinoliniums[20] display generally sig-
nificantly higher reactivity and applicability than more tradi-
tional N,O-acetals[16,17] under the usual conditions of electro-
philic catalysis. Consequently, we assumed that these cations
are too biased to be used to characterize properly the poten-
tial of the catalytic direct amidoalkylation of ketones. Con-
versely, we anticipated that substrates with reduced ionization
capability, which will generate “more electrophilic” ions are
more suited for our aim. Therefore, we focused in this study
mainly on a selection of such precursors (see structures out-
lined in bold in Figure 1). In this respect, we shall briefly con-
sider the use of “biased” N,O-acetal substrates, typified by a tet-
rahydroisoquinoline derivative 5 (see below), only for the pur-
pose of comparison.
Scheme 1. Outline of the work.
Our interest in developing a direct amidoalkylation reaction
mainly comes from our previous discovery that triflimidate-
based catalysts display high catalytic activity in the alkylation
of silyl enol ethers.[17a–c] Following these preliminary results, we
were logically eager to expand the chemistry to the most chal-
lenging use of ketones and aldehydes and were pleased to
report the first examples of direct a-amidoalkylation of a small
range of ketones using 1 mol% Sn(NTf2)4.[17b] While this prelimi-
nary report documented a few results of high temperature
(60–708C) experiments that were restricted to random combi-
nations of isoindolic hemiaminals and a small set of ketones
[Scheme 1, Eq. (1)], various control experiments carried out at
room temperature (RT) were found to be promising. This
prompted us to investigate the reaction in more detail with
the aim of establishing in a more rational manner the identity
of the reaction partners that are able to productively combine
in this milder and more appealing room temperature variant,
reactivity scale shown in Figure 1 that we were able to build
over the years by following the accumulation of a great
number of results on diverse amidoalkylation reactions, in the
context of our continuing interest in catalytic N-acyliminium
ion chemistry. This scale primarily includes a representative set
of N,O-acetals that are routinely used in our laboratory, along
with their ancillary N-acyliminium ions, given the inextricable
relationship between the ionization ability of N,O-acetals and
the electrophilicity of the incipient N-acyliminiums. For the
thereby providing
a breakthrough in N-acyliminium ion
chemistry [Scheme 1, Eq. (2)].[22]
Herein, we advance the results of a rationally designed
study that demonstrates the feasibility of this objective, with
the Mannich-type coupling of many combinations of N,O-ace-
tals and ketones/aldehydes proceeding successfully at room
temperature and low catalyst loading, to provide access to
ring-substituted b-amido ketones. Moreover, a comprehensive
evaluation of this method has resulted in the determination of
an interesting empirical reactivity scale of ketone nucleophilici-
ty, which, hopefully, will guide organic chemists in the design
and development of new alkylation processes and carbonyl
a-functionalization methods beyond this N-acyliminium ion
chemistry.
Figure 1. Empirical reactivity scale for some representative N-O-acetals used
in our laboratory and their incipient N-acyliminiums, along with the expect-
ed comparison with N-carbamoyl (iso)quinoliniums and N-aryl tetrahydroiso-
quinoliniums.
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6013
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ing from partial hydrolysis of the starting acetate 2a, no
longer evolves towards the formation of the adduct 4aa. Ac-
cordingly, when hydroxy lactam 1a or its methoxy counterpart
were reacted with cyclohexanone (3a) under the optimized re-
action conditions, either no or very low (ca. 5%) conversions
were observed, respectively (results not shown). The lower re-
activity of 1a versus 2a is quite consistent with many other
observations in our laboratory,[17b–d] and also with several re-
ported examples illustrating a much higher propensity of ace-
tates to undergo ionization compared with the parent alco-
hols, particularly when using weakly acidic catalysts[26] or sub-
strates carrying a relatively low degree of stabilizing sub-
units.[27] We are not aware of a similar direct comparison be-
tween alkoxy versus ester as a leaving group in acid-catalyzed
SN1 alkylations reported, but our results reveal that the reactivi-
ty of alkoxy lactams in this phthalimide series is much closer to
that of hydroxy lactams than in that of acetates, thereby illus-
trating the importance of an ester leaving group in this specific
case.[28] It is worth mentioning that, in reactions of other N,O-
acetals of higher ranking in our empirical reactivity scale
(Figure 1), such ester leaving groups are generally no longer
required in catalytic a-amidoalkylations, and this is also the
case in this tin (IV)-catalyzed direct amidoalkylation of ketones.
In those cases, the use of a simple methoxy, or sometimes
even of a hydroxy, can be sufficient (see below). The impor-
tance of leaving groups, and the choice of the best of them
(OC(O)R vs. OR vs. OH) in catalytic SN1 alkylation reactions is
then, quite logically, closely and primarily correlated with the
ability of the cation precursor to undergo ionization and thus
to the identity of the N,O-acetal.
Results and Discussion
Ketones have been relatively little studied as C-nucleophiles in
catalytic alkylative chemistry[13] compared with aldehydes, de-
spite their ability to undergo enolization; therefore, we chose
to focus mainly on their use. Following our analysis summar-
ized in Figure 1 (see above), feasibility and optimization studies
were then undertaken to explore the alkylation of 2-allyl-3-oxo-
isoindolin-1-yl acetate (2a)[23] with cyclohexanone (3a). To
clearly characterize the various parameters determining the
outcome of the reaction, and also to increase the sustainability
of the method, attempts were made over 24 h using only
1 mol% of several different Lewis and ꢀ5 mol% of Brønsted
superacids (Table 1). Conversions were estimated after work-up
Table 1. Catalyst evaluation.
Entry
Catalyst (mol%)
3a [equiv]
Time [h]
4aa/1a[a,b]
1
2
3
4
5
6
7
8
9
Sn(NTf2)4·4DMSO (1)
Sn(NTf2)4·4DMSO (1)
Sn(NTf2)4·4DMSO (1)
In(NTf2)3 (1)
Cu(NTf2)2 (1)
Tf2NH (5)
3
3
2
3
3
3
3
3
2
24
72
24
39
24
24
24
24
24
70:30[c,d]
70:30
65:35
65:35
15:85
60:40
56:44
65:35
19:81
TfOH (5)
C6F5ÀCH(Tf)2 (2)
With this good procedure in hand, we started an evaluation
of the scope of the reaction. Recognizing that the nature of
the nitrogen substituent might significantly influence the
course of the reaction, by sequestrating the catalyst to variable
extent, we first evaluated a set of acetoxy lactams 2a–j, differ-
ing in their N-substituent, in the alkylation of 3a (Scheme 2,
top). In general, the reaction took place with either no or
modest diastereocontrol.[29]
C6F5ÀCH(Tf)2 (2)
[a] Products ratios were determined based on 1H NMR spectroscopic anal-
ysis of crude reaction mixtures. Internal standards were not used, but
these ratios can be likened to conversions because the reactions were
very clean and the isolated yields were always very close to the propor-
1
tions of adducts estimated by H NMR spectroscopic analysis of the crude
reaction mixtures. [b] d.r. not shown because reactions proceed with
either no to modest diastereocontrol. [c] Average value obtained from
three experiments. [d] Reactions carried out under the conditions of
entry 1 in other solvents gave inferior 4aa/1a ratios (55:45 in THF, 45:55
in MeNO2, 33:67 in both toluene and dichloroethane).
Based on the variable conversions observed, the reaction ef-
ficiency exhibited, as anticipated, a strong dependence on the
identity of the protecting group, with the conversion ranging
from 25% for substrate 2b, with a p-Lewis basic isoprenyl
group, to 80%, for substrate 2j, bearing the phenethyl subu-
nit. Although we have no rational explanation for the variation
in reactivity imparted by these different groups, it seems that
the weakest conversions observed (for 2b–e) could be as-
cribed to the Lewis base properties of the groups on nitrogen
and therefore to their propensity to interfere with the catalyst.
For instance, the isoprenyl group located in substrate 2b is
known to participate in cycloisomerization processes catalyzed
by Lewis super acids,[30] whereas the ortho-dimethoxy aryl
group in 2d might bind tin(IV) through a bidentate complexa-
tion mode, hence resulting in moderate levels of sequestration
of the catalyst and reducing the catalytic activity. Conversely,
the best reactivity exerted by 2j is noteworthy and even aston-
ishing, given the potentially concurrent intramolecular Friedel–
Crafts-type amidoalkylation, which, actually was not observed
by measuring the adduct 4/OH lactam 1 ratios from the
1H NMR spectra of the crude mixtures.[24] This optimization
study established the use of three equivalents of ketone in
acetonitrile as the best reaction system. In accordance with
our previous observation on relative transformations with sili-
con-based nucleophiles,[17b] tin(IV) triflimidate appeared by far
superior to a set of triflate and triflimidate salts. Interestingly,
use of 5 and 2 mol%, respectively, of the Brønsted superacids
[25]
Tf2NH and C6F5CH(Tf)2
gave promising conversions, but
Sn(NTf2)4 displayed substantially better catalytic efficiency, relia-
bly furnishing approximately 70% conversion into the desired
compound 4aa (Table 1).
It should be noted that increasing the reaction time to 72 h
was also tested without any benefit (Table 1, entries 2 vs. 1).
This suggests that, once formed, the hydroxy lactam 1a, result-
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6014
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
cut discrimination between ketones, allowing us to again es-
tablish an interesting empirical reactivity scale that can be di-
vided in four categories (Scheme 2, structures depicted in
black, orange, blue, and red) arbitrarily set on the basis of reac-
tion efficiency.
We were delighted to observe that 2-tetralone 3l and, to
a lesser extent, 2-indanone 3k, displayed an outstanding reac-
tivity, giving complete conversion into products 4ak–al within
short reaction times. Phenylacetone 3j afforded a conversion
of more than than 85% towards the formation of compound
4aj. 1-Indanone 3h and phenylacetophenone (2-deoxyben-
zoin) 3i (giving 4ah and 4ai, respectively) also appeared to be
interesting donor components in this catalytic direct alkylation
event, displaying slightly better reactivity than cyclohexanone
(3a), whereas cyclopentanone (3g) performed less well, pro-
viding about 50% conversion into 4ag. Somewhat surprisingly,
ketones conjugated to an aryl (3c–d and 3 f) or a styrenyl unit
(3e) appeared far less effective than 3a under our standard
conditions, affording 4ac, 4ad, 4af, and 4ae, respectively, in
only 15–48% conversion.[32] Acyclic ketones devoid of an aro-
matic substituent, exemplified here by methoxyacetone 3b (al-
kylation product 4ab), exhibited poor reactivity,[32] albeit with
complete branched selectivity.
As expected given their good reactivity, 3i and 3j could also
afford quantitative conversion of 2a in shorter reaction times
(24 h, unoptimized) with a catalyst loading of 2 mol% (see
Scheme 2). This slight change in the procedure enhanced the
conversion of slightly less reactive ketones like 3a, and those
of the “orange” family or 3e–h, which in contrast to 3i–j,
remain poorly responsive to the use of extended reaction time
(see below). Indeed, approximately 20% gain in conversion
was achieved in most of the reactions at 2 mol% loading, and
4aa and 4af–ai were isolated in synthetically useful yields
ranging from 62 to 90%. Remarkably, a considerable and
unique improvement was seen for the reluctant coupling be-
tween cyclohexanone (3a) and 2b, bearing the isoprenyl sub-
stituent on nitrogen. When 2 mol% Sn(NTf2)4 was used, the
poor 30% conversion initially observed at 1 mol% loading
(Scheme 2, top chart) could be raised to an appreciable 78%,
and 4ba was isolated in a synthetically useful yield of 69%.
The good to excellent reactivity exhibited by using deoxy-
benzoin (3i), benzylacetone (3j), 2-tetralone (3l), and 2-inda-
none (3k), as well as by the exclusive obtention of the
branched adduct 4ab in the reaction of methoxyacetone (3b)
are supportive of a critical role of thermodynamic factors; that
is, the proportion of the (more stable) enol forms at equilibri-
um.[31] Likewise, the order of reactivity: cyclohexanone (3a)>
cyclopentanone (3g)> acetophenone (3d)> methoxyacetone
(3b) (taken from the global reactivity scale in Scheme 2), is
also consistent with the reported relative keto–enol equilibri-
um constant values.[31]
Scheme 2. Comparative reactivity of various phthalimide-derived acetoxy
lactams 2a–j and representative ketones 3a–l. [a] The letters attached to
each product number refer to the starting acetoxy lactam and to the ketone
used. [b] Conversions determined by 1H NMR spectroscopic analysis are
given; values in parentheses are isolated yields. The recovered starting ma-
terial is the hydroxy lactam 1a–j. [c] Reactivity classification: low (black),
moderate (purple), good (blue), high (red). [d] Isolated yield=69%. [e] Data
for reactions catalyzed by Sn(NTf2)4·4DMSO (2 mol%), 4af–ah and 4aj deter-
mined after 40 h, 4aa and 4ai determined after 64 h. [f] The conversion was
estimated because of the complexity of the 1H NMR spectrum of the crude
mixture (keto/enol forms). [g] Isolated yield given for a reaction completed
within 2 h at 1 mol% loading. [h] Isolated yield given for a reaction complet-
ed within 1 h at 1 mol% loading.
at all. Along with its quantitative value, this set of results has
a qualitative interest by establishing an empirical reactivity
scale, with the emergence of three classes of N-groups
(Scheme 2, structures depicted in black, blue, and red).
With the view that the reactivity should also be directly re-
lated to the ability of ketones to equilibrate with their enol
form,[31] a large array of carbonyl compounds were then sur-
veyed in the alkylation of the model substrate 2a under the re-
action conditions described above. A representative set of ex-
amples involving 12 compounds (3a–l) is depicted in
Scheme 2 (bottom); the complete study is detailed in the Sup-
porting Information. The mild reaction conditions associated
with the relatively strong electrophilicity of N-acyliminium de-
rived from 2a[19] (Figure 1) was clearly critical to enable a clear-
The striking differences in reactivity observed in this study
by varying the ketone structures contrast strongly with the re-
lated direct cross-dehydrogenative couplings of tertiary
amines, which proceed through the in situ generation and
trapping of the weakly electrophilic iminium[15,33] intermediates
and tolerate a large variety of ketones (cyclic and acyclic) with
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6015
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
almost equal efficiency.[34] The branched-selectivity of methyl
alkyl ketones of the general formula CH3C(=O)CH2R, exempli-
fied here by the reaction of methoxyacetone (3b) but which
includes other counterparts (not shown) also differs from ex-
periments of alkylation of iminiums, which generally favor the
linear aminoalkylation products.[34,35] All in all, our study clearly
shows how influencial the strength of an electrophilic species
on the global outcome of direct a-alkylations of ketones can
be, exemplified in the case of our very electrophilic N-acylimi-
niums not only by the drawing of an unusual discrimination
frame of ketones, but also with an inversion of regioselectivity
as opposed to similar reactions on iminiums.[34] Nonetheless, it
is worth mentioning that N,O-acetals with robustness equal to
phthalimidic derivatives but of higher ionization capacity, with
respect to the production of more stabilized N-acyliminium
ions with electrophilicity closest to iminiums (for example, sub-
strate 5; see Schemes 4 and 5 below for structure), give results
in better agreement with the trends of the iminium series (see
below).
The two reactivity scales that were established empirically,
as well as meticulous observations made on the outcome of
the reaction of these preliminary experiments, were then uti-
lized to demonstrate the potential of this approach by ad-
dressing particularly challenging catalytic amidoalkylations. As
an illustration of our rational approach, we considered the
direct alkylation of parent hemiaminals 1, which, as stated
above, are intrinsically much less reactive than their acetoxy
counterparts 2 under acid catalysis at room temperature.[17c,d,23]
Indeed, in our feasibility experiments we often observed in-
complete conversions caused by the competitive formation of
1 in the course of the reaction, even with ketones of promising
reactivity such as 3a and 3e–j (giving products 4aa and 4ae–
aj in Scheme 2). Given their remarkable reactivity in the reac-
tions with acetoxy lactam 2a, we anticipated that 2-tetralone
(3l) and 2-indanone (3k) could override this limitation, ena-
bling the direct use of hydroxy lactams as electrophilic sub-
strates. A shown on Scheme 3, Eq. (1), both ketones not only
gave high yields of the N-allylated and N-phenethyl Mannich
adducts 4ak–al and 4jk, but their high nucleophilic reactivity
in this catalytic direct alkylation also overcame the deactivating
effect caused by the “worse” N-isoprenyl, N-methallyl (see
Scheme 2), and N-homoallyl groups, as shown by the excellent
yields recorded for products 4bl, 4el, and 4kl, respectively.
The efficient formation of 4kl is noticeably illustrative of this
spectacular “2-tetralone–2-indanone effect”, because the ace-
toxy lactam 2k [N-(CH2)2CH=CH2] completely failed to alkylate
cyclohexanone 3a (result not shown).[36] These couplings are
all the more remarkable because hemi-N,O-acetals derived
from phthalimide are, in most cases, not alkylated by common
nucleophiles, which may be considered better than 2-tetralone
and 2-indanone (b-diketones, poly-methoxy arenes, enoxy si-
lanes etc.) under catalytic conditions similar to those used in
this work.[17b–d] They indeed show that the intrinsic trends in
the reactivity of leaving groups in catalytic alkylative SN1
chemistry, which are primarily imposed by the substrate back-
bone as mentioned previously, can be upset by the nature of
specific nucleophiles. Overall, these couplings constitute a re-
Scheme 3. Challenging room temperature catalytic amidoalkylations.
markably mild, atom-economical, and powerful direct catalytic
amidoalkylation method that we initially believed beyond our
reach, and which strengthen the attractiveness of our ap-
proach.
Conversely and despite their intrinsically good reactivity (see
Scheme 2), we noted the inability of 2-deoxybenzoin (3i) and
phenylacetone (3j) to achieve a similar atom-economical alky-
lation of substrates 1, but conversely observed their unique
propensity to consume the small portions of 1 that are contin-
uously released during the catalytic alkylation of the acetoxy
lactams 2. Pursuing our methodical development of this room
temperature catalytic direct a- amidoalkylation of ketones, by
simply prolonging the reaction time to 72 h we were able to
establish half-gram scale reactions catalyzed by only 0.5 mol%
Sn(NTf2)4 with these two ketones with the acetoxy lactams 2i
and 2h bearing the “innocent” N-benzyl and N-PMB motifs,
thereby offering a viable option to reach almost quantitative
yields of the corresponding Mannich adducts 4ij, 4ii, and 4hj
[Scheme 3, Eq. (2)]. By using 3j, this procedure also gave good
yield of 4dj (66%) in the alkylation of substrate 2d bearing
a “worse” N-3,4-dimethoxyphenyl substituent, which invariably
stopped at about 40% conversion in its reaction with cyclo-
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6016
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
hexanone (3a). This simple improvement of the reaction effi-
ciency by prolonging the reaction time nicely complements
the procedure utilizing 2 mol% catalyst (see above).
Indeed, an exquisite reactivity was observed with various ke-
tones (3a, 3d, 3e, 3g, 3j, and 3m) notably including acetone
3m and acetophenone 3d, which performed poorly with ace-
toxy lactams 2 under our standard conditions (Scheme 5). In-
terestingly, by using our standard procedure, all six ketones
were alkylated with complete conversion, affording the corre-
sponding Mannich adducts 6 in high chemical yields (83–91%)
over a 4–48 h time frame. This set of results offers unambigu-
ous support for the paradigm mentioned above that N-acylimi-
nium ions in the quinolinium/isoquinolinium series distinguish
themselves from “usual” N-acyliminiums, by showing a much
better reactivity and a wider applicability.
Having investigated the ketones and isoindolic N,O-acetals
to forge the first part of this study, we next sought to diversify
the method and, in particular, to make it amenable to a large
range of substrates. In early 2013, Klussmann and co-workers
documented an original sequence starting with a thermal
metal-free direct CÀH oxidation of N-Cbz THIQ (THIQ = tetra-
hydroisoquinoline) and coupling of the resulting tert-hydroper-
oxyaminal adduct 5 with a variety of nucleophiles using
10 mol% MsOH as an activator in AcOH as a quite unusual re-
action medium.[37,38] In this study, only the Mannich coupling
of three ketones was reported. For comparison, reactions of
both tert-hydroperoxyaminal 5 and acetoxy lactam 2a with cy-
clopentanone (3g) as a model ketone were then evaluated
under Klussmann’s alkylation conditions [ketone (5 equiv),
MsOH (10 mol%), acetic acid, RT] and with our own optimized
reaction conditions [ketone (3 equiv), Sn(NTf2)4 (1 mol%),
CH3CN, RT] (Scheme 4). As anticipated (see Figure 1) the THIQ-
Scheme 5. Global reactivity of N,O-acetal 5: extending our reactivity scale.
Given that reaction times required to reach complete con-
version in these reactions were roughly in compliance with our
empirical scale previously established, we sought to gain more
precise data to strengthen this unrefined vision. To this end,
the direct amidoalkylation of these six model ketones were re-
peated and quenched after only 4 h and the crude mixtures
1
were evaluated by H NMR spectroscopic analysis (Scheme 5).
In contrast to what was previously observed in the isoindolone
series (Scheme 2), cyclohexanone (3a) was found to alkylate 5
faster than phenylacetone (3j), whereas acetone displayed
a much improved reactivity than observed previously (no reac-
tion took place with 2a), and even surpassed that of acetophe-
none (3d). This set of comparative results seems to indicate
a greater contribution of steric effects to the kinetics of the re-
action in the case of this N-CBz THIQ N,O-acetal 5. The other
ketones showed relative reactivities that were consistent with
those described in Scheme 2 (although the differences are
smoothed because of the intrinsically greater reactivity of the
THIQ system as mentioned previously), thus confirming the vi-
ability of our empirical reactivity scale of ketones.
Scheme 4. Comparison of Klussmann’s reaction conditions against ours with
substrate 2a against 5.
derived N,O-acetal 5 demonstrated a significantly higher reac-
tivity compared with 2a. Indeed, irrespective of the conditions
used, high conversions of the THIQ substrate 5 into 6g were
observed after short reaction times (3–4 h). Contrasting results
were obtained with the acetoxy lactam 2a depending on the
reaction conditions. Whereas only hydrolysis of 2a into 1a was
observed after prolonged reaction time under Klussmann’s
conditions, smooth formation of the alkylated ketone 4ag was
observed by using our standard procedure, thereby showcas-
ing the distinct value and larger applicability of the Sn(NTf2)4
method.
To complete this work, we were interested in evaluating an
assortment of N,O-acetals of reactivity that was intermediate
between that of the aminal THIQ 5 and the phthalimidic ace-
toxy lactames 2, which were studied in the first part of this
work (Figure 1). Indeed, the N-Cbz N,O-acetals 7a–b and their
N-tosyl counterparts 9a–b, derived from pyrrolidin-2-one, as
well as succinimidic N-PMB N,O-acetals 11 a–b were also exam-
ined (Scheme 6).
In line with this preliminary result and with the intrinsically
high reactivity generally displayed by THIQ-based pro-electro-
philes in amino- and amidoalkylation processes,[20,33,37–39]
5
demonstrated a reactivity coverage to ketones that was signifi-
cantly wider than that of the model phthalimidic N,O-acetals 2.
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6017
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Finally, despite some potential for side-reactions such as
aldol self-condensation under our conditions of super Lewis
acid catalysis, the branched a-phenyl propionaldehyde 13
turned out to be a competent nucleophile to afford the Man-
nich adduct 14 bearing two contiguous quaternary and tertiary
stereocenters in good yield (Scheme 7). The inherent reactivity
Scheme 7. Catalytic a-amidoalkylation of branched a-phenyl propionalde-
hyde 13 and oxyalkylation of ketone 3d with acetal 15.
displayed by 13 during these feasibility experiments places it
almost at the level of cyclohexanone (3a) in our empirical reac-
tivity scale. This exciting discovery arguably expands the syn-
thetic relevance of the method and sets the stage for future
evaluation of a broader combination of aldehydes/N,O-acetals.
Moreover, the alkylation of the unbiased (by reference to chro-
mene-based proelectrophiles) acetal 15 by the moderately re-
active acetophenone 3d (using only 1.3 equivalents) to afford
the pseudo-aldol adduct 16 in decent yield underlines the
high potential of the most electrophilic oxoniums in this
Sn(NTf2)4-catalyzed alkylative method.[35,40]
Scheme 6. Reaction scope: catalytic a-amidoalkylations of N,O-acetals 7a–b,
9a–b and 11 a–b with cyclohexanone (3a) and 2-tetralone (3l).
Given that these components are less robust substrates, and
all are susceptible to b-proton loss (E1 elimination), they con-
stitute a suitable means to probe the generality of the
method. In this preliminary study, reactions were mainly limit-
ed to the use of cyclohexanone (3a) and 2-tetralone (3l) as
benchmark ketones (Scheme 6). We were pleased to observe
that, despite the possibility of a competitive b-elimination, the
N-acyliminium ion precursors 7, 9, and 11 were found to alky-
late 3a to afford the desired Mannich adducts 8a, 10a, and
12a in moderate to good yields. The reaction kinetics of this
set of substrates were generally faster than those of acetates
2, thus confirming their greater reactivity according to
Figure 1. In contrast, and expectedly, their alkylation with 3l
proceeded remarkably well in every case, providing the desired
Mannich products 8l, 10l, and 12l in good to excellent yields
ranging from 80 to 93%. In line with the results shown in
Scheme 3, compounds 10l and 12l were isolated in high
yields from the less reactive but more atom-economical hemia-
minal substrates 7a and 9a, whereas the use of methoxy
aminal 7b was mandatory for a high yielding access to the
Mannich adduct 8b. From a general point of view, the fruitful
alkylation reactions of this set of N,O-acetals carrying rather
poor MeO and OH leaving groups (which also applies for sub-
strate 5) testify to the flexibility to adapt the choice of the
leaving group to the nature of the substrate in alkylative catal-
ysis, according to the comments noted above related to
Table 1 and Scheme 3. It is pleasant to see that this paradigm
can even apply in the case of challenging alkylations, as dem-
onstrated here with the use of enolizable ketones at room
temperature. Overall and importantly, this set of results also il-
lustrates the flexibility of our method.
Conclusion
We have developed a room-temperature catalytic direct a-ami-
doalkylation of ketone donors.[22,37] The reaction proceeds
under remarkably mild reaction conditions under the guidance
of 0.5–2 mol% of the Lewis superacid Sn(NTf2)4 as an optimal
catalyst without any amino additive. The mild reaction condi-
tions made possible a comprehensive survey of the scope of
the reaction, and this has resulted in the construction of em-
pirical reactivity scales both for the nitrogen substituents of
lactamic N,O-acetals and, more importantly, the carbonyl nucle-
ophiles. These reactivity scales, which are expected to be
useful to researchers in the field of alkylative and ketone a-
functionalization chemistry, were applied to fully explore the
scope of the reaction and notably define the borders of our
catalytic procedure. This catalytic coupling extends well
beyond the robust isoindolic N,O-acetals, and also allows alky-
lation of more difficult counterparts derived from succinimide
and pyrrolidinone, which are all sensitive to the loss of proton
(E1 elimination) when they are exposed to weak nucleophiles.
Likewise, oxoniums (via the parent acetals) and a-branched al-
dehydes were also demonstrated to be viable reaction partners
in this method. We believe that this study improves the chemi-
cal applicability of the venerable N-acyliminium ions, and also
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6018
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(69%); viscous oil; two diastereoisomers (d.r. 52:48); Rf =0.37 (cy-
provides a better understanding of their catalytic reactivity
with weak nucleophiles. The potential of the fascinating and
easily available Sn(NTf2)4 Lewis superacid catalyst is also broad-
ened,[41] and we hope that the specific and general information
gained from this study will be useful to the scientific communi-
ty.[42]
clohexane/EtOAc, 70:30); IR (CHCl3): n˜ =1773, 1678 cmÀ1 1H NMR
;
(300 MHz, CDCl3, 208C): d (mixture of diastereoisomers)=0.74 (m,
0.48H; 1H d2), 1.22–1.95 (m, 4.52H), 1.72 (s, 1.44H; 6H d2), 1.75 (s,
1.56H; 6H d1), 2.01–2.14 (m, 1H), 2.30–2.47 (m, 1H), 2.52–2.64 (m,
1H), 2.80 (dd, J=7.2, 11.5 Hz, 0.52H; 1H d1), 2.98 (m, 0.48H; 1H d2),
3.71 (dd, J=4.6, 15.4 Hz, 0.52H; 1H d1), 3.80 (dd, J=7.1, 15.4 Hz,
0.48H; 1H d2), 4.43 (dd, J=7.7, 15.4 Hz, 0.52H; 1H d1), 4.53 (dd, J=
7.1, 15,4 Hz, 0.48H; 1H d2), 5.24 (t, J=7.0 Hz, 0.48H; 1H d2), 5.30 (d,
J=2.6 Hz, 0.48H; 1H d2), 5.35 (m, 0.52H; 1H d1), 5.47 (s, 0.52H; 1H
d1), 7.30–7.64 (m, 3H), 7.77–7.86 ppm (m, 1H); 13C NMR (75 MHz,
CDCl3, 208C): d (mixture of diastereoisomers)=18.0, 18.1, 24.4,
24.4, 25.4, 25.7, 25.8, 25.8, 26.5, 38.1, 40.1, 41.6, 42.2, 50.4, 53.7,
57.4, 58.4, 119.1, 119.8, 121.1, 123.4, 123.5, 124.4, 128.0, 128.1,
131.3, 131.4, 133.0, 133.1, 135.5, 136.6, 143.5, 145.1, 168.4, 169.0,
209.0, 210.7 ppm; MS (EI): m/z (%): 297 (13) [M]+·, 228 (25), 200
(49), 132 (100); HRMS (EI): m/z calcd for C19H23NNaO2: 298.1802
[M+H]+; found: 298.1808.
Experimental Section
General remarks: All reactions were carried out under argon at-
mosphere in oven-dried glassware with magnetic stirring. Commer-
cially available compounds were used without further purification.
Solvents were distilled prior to use, taking precaution to exclude
moisture by heating to reflux over CaH2 (CH3CN, CH2Cl2) under an
inert atmosphere. The tin triflimidate salts were prepared as previ-
ously reported[43] and used as such, as solvates. Sn(NTf2)4 was then
used in the form Sn(NTf2)4·4DMSO, or Sn(NTf2)4·8DMSO.[43] Analyti-
cal thin-layer chromatography (TLC) was performed on silica gel 60
with fluorescent indicator UV254 TLC plates. The spots were visual-
ized with UV light (254 nm) and stained with a solution of p-anisal-
dehyde, followed by heating. Flash column chromatography was
performed by using silica gel 60 (particle size 0.040–0.063 mm) typ-
ically with a cyclohexane/ethyl acetate eluent system. FTIR spectra
were recorded with a PerkinElmer Frontier. NMR spectra were mea-
2-Allyl-3-(1-methoxy-2-oxo-2-phenylethyl)isoindolin-1-one (4af):
Prepared from acetoxy lactam 2a (0.25 mmol) and 2-methoxyace-
tophenone (3 f) by following the general procedure at RT with
Sn(NTf2)4·4DMSO (2 mol%). Reaction time: 40 h. Yield: 52 mg
(65%); colorless viscous oil; two diastereoisomers (d.r. 56:44); Rf =
0.38 (cyclohexane/EtOAc, 50:50); IR (CHCl3): n˜ =1772, 1711,
1616 cmÀ1; 1H NMR (300 MHz, CDCl3, 208C): d (mixture of diastereo-
isomers)=3.36 (s, 1.32H; 3H d2), 3.39 (s, 1.68H; 3H d1), 3.83–4.02
(m, 1H), 4.08 (m, 0.44H; 1H d2), 4.65–4.83 (m, 1.56H), 5.02–5.39 (m,
3H), 5.71–5.97 (m, 1H), 7.16–7.68 (m, 6H); 7.69–7.98 ppm (m, 3H);
13C NMR (75 MHz, CDCl3, 208C): d (mixture of diastereoisomers)=
43.5, 44.4, 58.7, 58.8, 60.6, 61.1, 76.6, 76.7, 77.0, 77.2, 77.4, 82.5,
85.6, 117.8, 118.1, 123.2, 123.6, 127.5, 128.3, 128.4, 128.5, 128.6,
128.7, 128.8, 129.5, 130.3, 131.3, 131.3, 132.3, 132.5, 132.9, 133.5,
133.7, 133.8, 133.8, 133.9, 135.8, 136.1, 140.9, 141.3, 168.5,
198.0 ppm; HRMS (ESI): m/z calcd for C20H20NO3: 322.1438 [M+H]+;
found: 322.1446.
1
sured at RT with a Bruker Avance 300 spectrometer. H NMR spec-
tra were recorded at 300 MHz. Chemical shifts (d) are reported in
ppm using residual solvent peaks as reference (CHCl3: d=
7.26 ppm). Data are reported as follows: chemical shift, multiplicity
(s: singlet, d: doublet, t: triplet, q: quartet, qui: quintuplet, m: mul-
tiplet), coupling constant (J in Hz) and integration. 13C NMR spectra
were recorded at 75 MHz using broadband proton decoupling,
and chemical shifts are reported in ppm using residual solvent
peaks as reference (CHCl3: d=77.0 ppm). High-resolution mass
spectra were recorded with a 6530 Q-TOF (Agilent Technologies).
The Q-TOF MS instrument was operated under the following con-
dition: Ion source ESI+ Agilent Jet Stream or APCI both in positive
ionization mode. Melting points were measured using open capil-
lary tubes with a Stuart Scientific analyzer and are uncorrected.
2-Allyl-3-(2-oxo-2,3-dihydro-1H-inden-1-yl)isoindolin-1-one
(4ak): Prepared from hydroxy lactam 1a (0.25 mmol) and 2-inda-
none (3k) by following the general procedure at RT with
Sn(NTf2)4·4DMSO (1 mol%). Reaction time: 12 h. Yield: 70 mg
(93%); viscous oil; two diastereoisomers (d.r. 53:47); Rf =0.30 (cy-
Sn(NTf2)4-catalyzed direct a-amidoalkylation reactions of ke-
tones and aldehydes; General procedure: Ketone (3 equiv) and
Sn(NTf2)4·xDMSO (0.5 to 2 mol%) were successively added to a solu-
tion of N,O-acetal in anhydrous acetonitrile (2 mLmmolÀ1 of sub-
strate) under an argon atmosphere. The reaction mixture was
stirred at RT (see reaction time specified for each compound) and
conversion of the starting material was monitored by TLC and/or
1H NMR spectroscopic analysis. The reaction mixture was directly
purified by flash column chromatography on silica gel (cyclohex-
ane/ethyl acetate), to afford the corresponding a-amidoalkylated
ketone.
clohexane/EtOAc, 70:30); IR (CHCl3): n˜ =1746, 1681, 1616 cmÀ1
;
1H NMR (300 MHz, CDCl3, 208C): d (mixture of diastereoisomers)=
3.24 (d, J=23.0 Hz, 0.53H; 1H d1), 3.48–3.76 (m, 2H), 3.86 (dd, J=
7.5, 15.5 Hz, 0.47H; 1H d2), 4.15 (s, 0.47H; 1H d2), 4.24 (d, J=3.8 Hz,
0.53H; 1H d1), 4.57 (m, 0.47H; 1H d2), 4.95 (m, 0.53H; 1H d1), 5.10–
5.39 (m, 3H), 5.74 (m, 0.47H; 1H d2), 5.94 (m, 0.53H; 1H d1), 6.59
(d, J=7.5 Hz, 0.47H; 1H d2), 6.80–7.91 ppm (m, 7.53H); 13C NMR
(75 MHz, CDCl3, 258C): d (mixture of diastereoisomers)=28.1, 29.3,
43.1, 43.8, 60.2, 60.6, 63.2, 64.2, 118.5, 118.8, 122.4, 123.0, 123.6,
123.9, 124.1, 129.2, 129.3, 131.8, 132.0, 132.6, 132.9, 133.2, 134.3,
140.4, 141.8, 168.3, 168.5, 201.7, 203.7 ppm; HRMS (ESI): m/z calcd
for C20H18NO2: 304.1332 [M+H]+; found: 304.1342.
Analytical data: This experimental section gives the analytical de-
scription of the isolated products for which a yield was recorded.
Products 4ca–ia were formed in moderate conversions in the con-
text of our optimization study (see Schemes 1and 2), and their pu-
rification gave only low amounts of analytically pure samples,
which are all fully described in the Supporting Information. The
conversion into products 4k and 4l was too low and purification
was not attempted, hence these two compounds are not de-
scribed.
2-Allyl-3-(2-oxo-1,2,3,4-tetrahydronaphthalen-1-yl)isoindolin-1-
one (4al): Prepared from hydroxy lactam 1a (0.25 mmol) and 2-tet-
ralone (3l) by following the general procedure at RT with
Sn(NTf2)4·4DMSO (1 mol%). Reaction time: 12 h. Yield: 69 mg
(88%); yellow viscous oil; two diastereoisomers (d.r. 58:42); Rf =
0.34 (cyclohexane/EtOAc, 60:40); IR (CHCl3): n˜ =1679, 1617 cmÀ1
;
1H NMR (300 MHz, CDCl3, 208C): d (mixture of diastereoisomers)=
2.01 (m, 0.58H; 1H d1), 2.41 (m, 0.42H; 1H d2), 2.65–3.03 (m, 3H),
3.78 (t, J=8.6 Hz, 0.42H; 1H d2), 3.83 (t, J=8.6 Hz, 0.58H; 1H d1),
4.13 (d, J=3.3 Hz, 0.58H; 1H d1), 4.21 (d, J=4.3 Hz, 0.42H; 1H d2),
4.86–5.01 (m, 1H), 5.19–5.41 (m, 2.58H), 5.67 (d, J=3.3 Hz, 0.42H;
2-(3-Methylbut-2-enyl)-3-(2-oxocyclohexyl)isoindolin-1-one
(4ba): Prepared from acetoxy lactam 2b (0.25 mmol) and cyclohex-
anone 3a by following the general procedure at RT with
Sn(NTf2)4·4DMSO (2 mol%). Reaction time: 64 h. Yield: 51 mg
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6019
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1H d1), 5.82–5.98 (m, 1H), 6.59 (d, J=7.5 Hz, 0.42H; 1H d2), 6.82 (m,
0.58H; 1H d1), 6.92–7.52 (m, 6H), 7.64 (d, J=7.1 Hz, 0.58H; 1H d1),
7.84 ppm (d, J=7.4 Hz, 0.42H; 1H d2); 13C NMR (75 MHz, CDCl3,
208C): d (mixture of diastereoisomers)=28.1, 28.7, 39.1, 39.9, 43.2,
43.3, 51.0, 51.2, 60.9, 62.0, 118.5, 118.6, 122.2, 122.7, 123.5, 124.0,
126.7, 127.2, 127.3, 127.8, 128.1, 128.3, 128.4, 128.8, 128.8, 131.1,
131.3, 131.8, 132.4, 132.5, 132.9, 133.4, 133.4, 136.2, 136.7, 142.7,
143.0, 167.6, 168.6, 207.4, 210.0 ppm; HRMS (ESI): m/z calcd for
C21H20NO2: 318.1489 [M+H]+; found: 318.1491.
N-Benzylcarboxy-2-(2-oxo-1,2,3,4-tetrahydronaphthalen-1-yl)pyr-
rolidine (8l): Prepared from methoxy lactam 7b (59 mg,
0.25 mmol) by following the general procedure with 2-tetralone
(3l) at RT with Sn(NTf2)4·4DMSO (1 mol%). Reaction time: 3 h.
Yield: 76 mg (87%); orange oil; two diastereoisomers (d.r. not de-
1
termined due to complication of both H and 13C NMR spectra by
the presence of rotamers); Rf =0.50 (cyclohexane/EtOAc, 70:30); IR
;
(CHCl3): n˜ =1690, 1410 cmÀ1 1H NMR (300 MHz, CDCl3, 258C): d
(mixture of diastereomers)=1.49–3.18 (m, 7.34H), 3.27–3.70 (m,
3.40H), 3.76 (m, 0.13H), 4.26 (m, 0.13H), 4.53–4.34 (m, 1H), 4.89–
5.36 (m, 2H), 6.91–7.50 ppm (m, 9H); 13C NMR (75 MHz, CDCl3,
208C): d=22.3, 22.9, 23.5, 26.9, 27.1, 27.3, 27.4, 27.6, 28.2, 28.7,
29.0, 29.3, 37.9, 38.2, 38.6, 46.3, 46.4, 46.5, 46.9, 56.6, 57.0, 58.0,
58.3, 58.6, 59.5, 59.8, 60.1, 66.5, 66.8, 67.0, 67.2, 126.5, 126.7, 126.9,
127.1, 127.2, 127.4, 127.7, 127.8, 127.9, 128.1, 128.2, 128.3, 128.4,
128.5, 128.6, 129.3, 129.6, 129.8, 130.0, 134.7, 134.9, 135.1, 136.3,
136.9, 137.3, 138.1, 154.5, 155.3, 210.7, 211.0 ppm; HRMS (ESI): m/z
calcd for C22H23NO3Na: 372.1570 [M+Na]+; found: 372.1569.
2-(But-3-en-1-yl)-3-(2-oxo-1,2,3,4-tetrahydronaphthalen-1-yl)i-
soindo-lin-1-one (4kl): Prepared from the N-homoallyl phthali-
mide-derived hydroxy lactam 1k (0.25 mmol) by following the gen-
eral procedure with 2-tetralone (3l) at RT with Sn(NTf2)4·4DMSO
(1 mol%). Reaction time: 12 h. Yield: 76 mg (92%); colorless oil;
two diastereoisomers (d.r. 58:42); Rf =0.25 (cyclohexane/EtOAc,
70:30); IR (CHCl3): n˜ =1700, 1666, 1600 cmÀ1 1H NMR (300 MHz,
;
CDCl3, 208C): d (mixture of diastereoisomers)=2.15 (m, 0.42H; 1H
d2), 2.34–2.57 (m, 2.42H), 2.63–3.03 (m, 3H), 3.09 (m, 0.58H; 1H d1),
3.24 (m, 0.58H; 1H d1), 4.07–4.27 (m, 1.42H), 4.39 (m, 0.58H; 1H
d1), 4.99–5.19 (m, 2H), 5.40 (d, J=3.9 Hz, 0.42H; 1H d2), 5.68 (d, J=
3.4 Hz, 0.58H; 1H d1), 5.74–5.92 (m, 1H), 6.71–7.88 ppm (m, 8H);
13C NMR (75 MHz, CDCl3, 208C): d (mixture of diastereoisomers)=
28.0, 28.8, 32.4, 32.7, 39.1, 39.7, 39.8, 40.0, 51.0, 51.7, 61.2, 61.5,
117.3, 117.6, 122.0, 122.5, 123.4, 123.9, 126.7, 127.1, 127.3, 127.8,
128.0, 128.1, 128.1, 128.4, 128.7, 128.8, 131.0, 131.3, 131.5, 132.5,
132.9, 132.9, 134.6, 135.1, 136.1, 136.6, 142.7, 143.2, 168.2, 168.7,
207.7, 210.3 ppm; HRMS (ESI): m/z calcd for C22H22NO2: 332.1645
[M+H]+; found: 332.1650.
N-Tosyl-2-(2-oxocyclohexyl)pyrrolidine (10a): Prepared from me-
thoxy compound 9b (64 mg, 0.25 mmol) by following the general
procedure with cyclohexanone (3a) at RT with Sn(NTf2)4·4DMSO
(1 mol%). Reaction time: 72 h. Yield: 69 mg (86%); white solid;
two diastereoisomers (d.r. 54:46); Rf =0.45 (cyclohexane/EtOAc,
70:30); IR (CHCl3): n˜ =1703, 1339, 1155 cmÀ1 1H NMR (300 MHz,
;
CDCl3, 258C): d (mixture of diastereomers)=1.22–1.90 (m, 7H),
1.91–2.16 (m, 2H), 2.21–2.50 (m, 3.46H), 2.43 (s, 3H), 3.09 (s,
0.54H), 3.17–3.44 (m, 2H), 3.98–4.08 (m, 1H), 7.29–7.35 (m, 2H),
7.67–7.74 ppm (m, 2H); 13C NMR (75 MHz, CDCl3, 208C): d (mixture
of diastereomers)=21.5, 23.8, 24.6, 24.8, 25.1, 26.5, 27.3, 28.2, 28.2,
30.5, 32.0, 42.2, 43.0, 48.6, 49.8, 54.8, 55.4, 58.6, 60.2, 127.6, 127.7,
129.7, 133.3, 135.0, 143.4, 143.5, 211.9, 212.3 ppm; HRMS (ESI): m/z
calcd for C17H23NO3SNa: 344.1291 [M+Na]+; found: 344.1289.
2-(4-Methoxybenzyl)-3-(2-oxo-1-phenylpropyl)isoindolin-1-one
(4hj): Prepared from acetoxy lactam 2h (500 mg, 1.61 mmol) by
following the general procedure with 2-phenylacetone (3j) at RT
with Sn(NTf2)4·4DMSO (0.5 mol%). Reaction time: 72 h. Yield:
583 mg (94%); two diastereoisomers (d.r. 68:32); Rf =0.375 (cyclo-
N-Tosyl-2-(2-oxo-1,2,3,4-tetrahydronaphthalen-1-yl)pyrrolidine
(10l): Prepared from hydroxy lactam 9a (60 mg, 0.25 mmol) by fol-
lowing the general procedure with 2-tetralone (3l) at RT with
Sn(NTf2)4·4DMSO (1 mol%). Reaction time: 3.5 h. Yield: 80 mg
(87%); white solid; m.p. 1658C; two diastereoisomers (d.r. 80:20);
Rf =0.45 (cyclohexane/EtOAc, 70:30); IR (CHCl3): n˜ =1704, 1342,
hexane/EtOAc, 70:30); IR (CHCl3): n˜ =1690, 1611 cmÀ1 1H NMR
;
(300 MHz, CDCl3, 208C): d (mixture of diastereoisomers)=1.80 (s,
2.04H; 3H d1), 2.07 (s, 0.96H; 3H d2), 3.15 (d, J=15.1 Hz, 0.32H; 1H
d2), 3.76 (s, 0.96H; 3H d2), 3.80 (s, 2.04H; 3H d1), 3.90 (d, J=6.9 Hz;
1H), 4.59 (d, J=15.3 Hz, 0.68H; 1H d1), 5.00 (d, J=15.3 Hz, 0.68H;
1H d1), 5.14 (m, 0.64H; 2H d2), 5.29 (d, J=6.9 Hz, 0.68H; 1H d1),
6.59 (d, J=7.6 Hz, 0.68H; 1H d1), 6.76 (m, 0.68H; 1H d1), 6.82–6.97
(m, 3.32H), 7.11–7.56 (m, 7.32H), 7.75 (d, J=7.5 Hz, 0.68H; 1H d1),
7.85 ppm (m, 0.32H; 1H d2); 13C NMR (75 MHz, CDCl3, 208C): d (mix-
ture of diastereoisomers)=29.3, 30.4, 44.8, 55.2, 55.3, 59.7, 60.2,
62.1, 64.7, 113.8, 114.1, 123.0, 123.5, 124.1, 124.7, 128.2, 128.3,
128.5, 128.6, 128.8, 129.1, 129.2, 129.2, 129.3, 129.4, 129.6, 130.8,
131.7, 132.0, 132.4, 133.6, 134.0, 143.3, 145.4, 158.8, 159.0, 169.0,
169.7, 206.4, 206.8 ppm. HRMS (ESI): m/z calcd for C25H24NO3:
386.1751 [M+H]+; found: 386.1653.
1
1153 cmÀ1; H NMR (300 MHz, CDCl3, 208C): d (mixture of diastereo-
isomers)=1.16–1.38 (m, 3H), 4.41–1.55 (m, 1H), 2.41 (s, 0.6H; 3H
d2), 2.44 (s, 2.4H; 3H d1), 2.46–2.61 (m, 1H), 2.73–2.86 (m, 1H),
2.89–3.02 (m, 1.2H), 3.06–3.32 (m, 2.8H), 3.58 (m, 0.2H; 1H d2), 4.10
(d, J=5.6 Hz, 0.8H; 1H d1), 4.35–4.46 (m, 1H), 7.08–7.38 (m, 3.8H),
7.34 (d, J=8.0 Hz, 1.6H; 2H d1), 7.52–7.64 (m, 1H), 7.80 ppm (d, J=
8.0 Hz, 1.6H; 2H d1); 13C NMR (75 MHz, CDCl3, 208C): d (mixture of
diastereoisomers)=21.6, 23.8, 24.1, 27.6, 27.9, 28.8, 29.8, 38.4, 39.1,
49.2, 49.4, 57.1, 59.0, 62.1, 62.4, 126.5, 126.9, 127.3, 127.6, 127.8,
128.1, 128.3, 129.7, 129.8, 130.8, 134.1, 134.2, 137.2, 141.8, 143.6,
143.7, 210.9, 211.8 ppm; HRMS (ESI): m/z calcd for C21H23NO3SNa:
392.1291 [M+Na]+; found: 392.1295.
(E)-2-Benzylcarboxy-1-(2-oxo-4-phenylbut-3-en-1-yl)-3,4-dihy-
droiso-quinoline (6e): Prepared from THIQ-derived N,O-acetal 5
(89 mg, 0.25 mmol) by following the general procedure with (E)-4-
phenylbut-3-en-2-one (3e) at RT with Sn(NTf2)4·4DMSO (1 mol%).
Reaction time: 48 h. Yield: 87 mg (85%); colorless oil; Rf =0.55 (cy-
N-(4-Methoxybenzyl)-5-(2-oxo-1,2,3,4-tetrahydronaphthalen-1-
yl)py-rrolidin-2-one (12l): Prepared from hydroxy lactam 11 a
(55 mg, 0.25 mmol) by following the general procedure with 2-tet-
ralone (3l) at RT with Sn(NTf2)4·4DMSO (1 mol%). Reaction time:
15 h. Yield: 81 mg (93%); brown solid; two diastereoisomers (d.r.
68:32); Rf =0.25 (cyclohexane/EtOAc, 70:30); IR (CHCl3): n˜ =1704,
clohexane/EtOAc, 70:30); IR (CHCl3): n˜ =1689, 1607, 1423 cmÀ1
;
1H NMR (300 MHz, CDCl3, 208C): d=2.77–3.34 (m, 4H), 3.39–3.67
(m, 1H), 3.95–4.08 (m, 0.5H), 4.19–4.34 (m, 0.5H), 5.07–5.25 (m,
2H), 5.75–5.90 (m, 2H), 6.69 (d, J=16.1 Hz, 0.5H), 6.86 (d, J=
16.1 Hz, 0.5H), 7.06–7.78 ppm (m, 14H); 13C NMR (75 MHz, CDCl3,
208C): d=28.3, 28.6, 38.6, 39.4, 48.3, 48.8, 51.8, 52.2, 67.3, 67.4,
126.1, 126.5, 127.0, 127.1, 127.9, 128.0, 128.5, 128.8, 128.9, 129.1,
130.6, 134.2, 134.3, 134.5, 136.5, 143.2, 155.1, 155.3, 197.1,
197.4 ppm; HRMS (ESI): m/z calcd for C27H25NO3Na: 434.1727
[M+Na]+; found: 434.1729.
1
1680, 1611 cmÀ1; H NMR (300 MHz, CDCl3, 208C): d (mixture of dia-
stereoisomers)=1.18–1.56 (m, 1.5H), 1.80–2.17 (m, 2H), 2.87–2.94
(m, 2.5H), 3.25 (d, J=14.8 Hz, 0.32H; 1H d2), 3.65 (d, J=6.9 Hz,
0.32H; 1H d2), 3.76–3.90 (m, 1H), 3.82 (s, 0.96H; 3H d2), 3.83 (s,
2.04H; 3H d1), 3.96 (d, J=14.7 Hz, 0.68H; 1H d1), 4.39 (m, 0.68H;
1H d1), 4.97 (d, J=14.7 Hz, 0.32H; 1H d2), 5.21 (d, J=14.7 Hz,
0.68H; 1H d1), 6.83–7.02 (m, 3H), 7.21–7.32 ppm (m, 5H); 13C NMR
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6020
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Capdevila, P. G. Cozzi, Angew. Chem. Int. Ed. 2011, 50, 7842; Angew.
Chem. 2011, 123, 7988; b) A. Gualandi, D. Petruziello, E. Emer, P. G.
Cozzi, Chem. Commun. 2012, 48, 3614.
(75 MHz, CDCl3, 208C): d (mixture of diastereoisomers)=21.8, 22.8,
27.4, 27.9, 29.6, 29.7, 38.2, 39.7, 44.0, 44.9, 52.2, 55.3, 58.1, 59.7,
76.6, 77.0, 77.2, 77.5, 114.1, 114.2, 127.2, 127.5, 127.7, 127.8, 128.1,
128.2, 128.6, 129.1, 129.3, 129.4, 129.7, 132.9, 136.7, 136.9, 159.1,
175.8, 210.2, 210.6 ppm; HRMS (ESI): m/z calcd for C22H23NO3Na:
372.1570 [M+Na]+; found: 372.1576.
[7] For organocatalytic stereoselective SN1 alkylations of aldehydes through
enantioselective combined aminocatalysis/Lewis acid catalysis, see: a) A.
Gualandi, L. Mengozzi, C. M. Wilson, P. G. Cozzi, Synthesis 2014, 1321;
b) M. G. Capdevila, F. Benfatti, L. Zoli, M. Stenta, P. G. Cozzi, Chem. Eur. J.
2010, 16, 11237; c) K. Motoyama, M. Ikeda, Y. Miyake, Y. Nishibayashi,
Eur. J. Org. Chem. 2011, 2239.
[8] For organocatalytic stereoselective SN1 allylations of aldehydes through
enantioselective combined aminocatalysis/transition metal catalysis,
see: a) S. Afewerki, I. Ibrahem, J. Rydfjord, P. Breistein, A. Cordova,
Chem. Eur. J. 2012, 18, 2972; b) M. Yoshida, E. Masaki, T. Terumine, S.
Hara, Synthesis 2014, 1367; c) M. Chiarucci, M. di Lillo, A. Romanello,
P. G. Cozzi, G. Cera, M. Bandini, Chem. Sci. 2012, 3, 2859; d) S. Krautwald,
D. Sarlah, M. Schrafroth, E. M. Carreira, Science 2013, 340, 1065; for
a similar approach in propargylation reactions, see: e) Y. Nishibayashi,
Synthesis 2012, 489.
[9] For organocatalytic stereoselective SN1 allylations of aldehydes through
enantioselective combined Brønsted acid catalysis/transition-metal cat-
alysis, see: G. Jiang, B. List, Angew. Chem. Int. Ed. 2011, 50, 9471; Angew.
Chem. 2011, 123, 9643.
[10] For cascade reactions that include an organocatalytic stereoselective
SN1 alkylation of aldehydes, see: J. Stiller, A. J. Vorholt, K. A. Ostrowski,
A. Behr, M. Christmann, Chem. Eur. J. 2012, 18, 9496.
[11] For the organocatalytic stereoselective alkylation of aldehydes through
cross dehydrogenative couplings, see: a) F. Benfatti, M. G. Capdevila, L.
Zoli, E. Benedetto, P. G. Cozzi, Chem. Commun. 2009, 5919; b) B. Zhang,
S.-K. Xiang, L.-H. Zhang, Y. Cui, N. Jiao, Org. Lett. 2011, 13, 5212.
[12] For the organocatalytic stereoselective alkylation of aldehydes by radi-
calar approaches; SOMO-initiated processes, see: a) T. D. Beeson, A.
Mastracchio, J.-B. Hong, K. Ashton, D. W. C. MacMillan, Science 2007,
316, 582; for photoredox initiated transformations, see: b) D. A. Nice-
wicz, D. W. C. MacMillan, Science 2008, 322, 77; c) E. Arceo, A. Baha-
monde, G. Bergonzini, P. Melchiorre, Chem. Sci. 2014, 5, 2438; d) A. Gua-
landi, M. Marchini, L. Mengozzi, M. Natali, M. Lucarini, P. Ceroni, P. G.
Cozzi, ACS Catal. 2015, 5, 5927.
[13] For few examples of organocatalytic stereoselective SN1 alkylation of
ketones, see: a) Refs. [4,12c]; b) L. Zhang, L. Cui, X. Li, J. Li, S. Luo, J.-P.
Cheng, Eur. J. Org. Chem. 2010, 4876; c) M. Trifonidou, C. G. Kotokos,
Eur. J. Org. Chem. 2012, 1563; d) X. Huo, M. Quan, G. Yang, X. Zhao, D.
Liu, Y. Liu, W. Zhang, Org. Lett. 2014, 16, 1570.
2-(2-Benzyl-3-oxoisoindolin-1-yl)-2-phenylpropanal (14): Pre-
pared from acetoxy lactam 2i (70 mg, 0.25 mmol) by following the
general procedure with 2-phenylpropanal (13) (101 mL, 0.750
mmol, 3 equiv) at RT with Sn(NTf2)4·8DMSO (2 mol%). Reaction
time: 72 h. Yield: 67 mg (76%); white solid; two diastereoisomers
(d.r. 55:45); Rf =0.50 (cyclohexane/EtOAc, 70:30); IR (neat): n˜ =
1
1720, 1682, 1614 cmÀ1; H NMR (300 MHz, CDCl3, 208C): d (mixture
of diastereoisomers)=1.16 (s, 1.35H; 3H d2), 1.20 (s, 1.65H; 3H d1),
2.74 (d, J=15.7 Hz, 0.55H; 1H d1), 4.09 (d, J=15.7 Hz, 0.45H; 1H
d2), 5.15 (d, J=15.7 Hz, 0.55H; 1H d1), 5.40 (s, 0.55H; 1H d1), 5.45
(s, 0.45H; 1H d2), 5.46–5.55 (m, 1H), 6.57–6.65 (m, 1H), 7.05–7.22
(m, 2H), 7.29–7.58 (m, 10H), 7.85–8.00 (m, 1H), 9.55 (s, 0.55H; 1H
d1), 9.70 ppm (s, 0.45H; 1H d2); 13C NMR (75 MHz, CDCl3, 208C): d
(mixture of diastereoisomers)=11.3, 12.3, 45.0, 45.4, 57.1, 58.0,
62.5, 63.0, 123.7, 123.8, 124.1, 124.3, 127.2, 127.3, 127.6, 128.0,
128.2, 128.3, 128.3, 128.5, 128.6, 128.7, 129.2, 129.5, 130.9, 131.9,
132.7, 135.3, 135.7, 136.4, 136.5, 142.2, 144.0, 169.8, 170.5, 199.6,
200.0 ppm; HRMS (ESI): m/z calcd for C24H21NO2Na: 378.1465
[M+Na]+; found: 378.1460.
Acknowledgements
B.T. is grateful to the Tunisian government for a scholarship.
A.E.B. thanks the University of Le Havre for an ATER position.
Dr Radouane Affani is acknowledged for his contribution at
the early stage of this study. We are grateful to Dr Nicolas
Hucher and Ms Marie-JosØ Tranchant for their assistance with
analytical measurements.
[14] For the non-enantioselective direct catalytic alkylation of aldehydes and
ketones with very reactive carbocations catalyzed by strong achiral
Brønsted and Lewis acids, see: a) C. Xing, H. Sun, J. Zhang, G. Li, R. Y.
Chi, Chem. Eur. J. 2011, 17, 12272; b) X. Pan, M. Li, Y. Gu, Chem. Asian J.
2014, 9, 268; c) for a stoichiometric approach under extremely mild re-
action conditions, see: S. R. Koppolu, N. Naveen, R. Balamurugan, J. Org.
Chem. 2014, 79, 6069; d) Y. Inada, Y. Nishibayashi, S. Uemura, Angew.
Chem. Int. Ed. 2005, 44, 7715; Angew. Chem. 2005, 117, 7893.
Keywords: alkylation
superacidic systems
· enols · lactams · Lewis acids ·
[1] For reviews, highlights, and features articles, see: a) P. Melchiorre,
Angew. Chem. Int. Ed. 2009, 48, 1360; Angew. Chem. 2009, 121, 1386;
b) A. Gualandi, P. G. Cozzi, Synlett 2013, 281; c) D. M. Hodgson, A. Charl-
ton, Tetrahedron 2014, 70, 2207.
[15] H. Mayr, B. Kempf, A. R. Ofial, Acc. Chem. Res. 2003, 36, 66.
[16] For a recent review on organocatalytic enantioselective transformations
of N-acyliminiums, see: a) Y.-S. Lee, Md. M. Alam, R. S. Keri, Chem. Asian
J. 2013, 8, 2906; for seminal contributions in catalytic asymmetric ami-
doalkylations with chiral Lewis acids, see: b) M. Takamura, K. Funabashi,
M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 2000, 122, 6327; c) O. Ono-
mura, Y. Kanda, Y. Nakamura, T. Maki, Y. Matsumura, Tetrahedron Lett.
2002, 43, 3229; by organocatalysis: d) I. T. Raheem, P. S. Thiara, E. A. Pe-
terson, E. N. Jacobsen, J. Am. Chem. Soc. 2007, 129, 13404; e) M. C. Mur-
atore, C. A. Holloway, A. W. Pilling, R. I. Storer, G. Trevitt, D. J. Dixon, J.
Am. Chem. Soc. 2009, 131, 10796.
[17] a) R. Ben Othman, T. Bousquet, M. Othman, V. Dalla, Org. Lett. 2005, 7,
5335; b) R. Ben Othman, R. Affani, M. J. Tranchant, S. Antoniotti, V. Dalla,
E. DuÇach, Angew. Chem. Int. Ed. 2010, 49, 776; Angew. Chem. 2010,
122, 788; c) A. Devineau, G. Pousse, C. Taillier, J. Rouden, J. Blanchet, V.
Dalla, Adv. Synth. Catal. 2010, 352, 2881; d) M. Hamon, N. Dickinson, A.
Devineau, D. Bolien, M. J. Tranchant, C. Taillier, I. Jabin, D.-C. Harrowven,
R. Whitby, A. Ganesan, V. Dalla, J. Org. Chem. 2014, 79, 1900.
[2] For stereoselective SN2 alkylations of aldehydes catalyzed by chiral ami-
ˇ
´
nocatalysts, see: B. List, I. Coric, O. O. Grygorenko, P. S. J. Kaib, I. Komar-
ov, A. Lee, M. Leutzsch, S. C. Pan, A. V. Tymtsunik, M. van Gemmeren,
Angew. Chem. Int. Ed. 2014, 53, 282; Angew. Chem. 2014, 126, 286, and
references therein.
[3] For stereoselective SN1 alkylations of alcohols catalyzed by chiral amino-
catalysts and/or their protic salts, see: a) R. R. Shaikh, A. Mazzanti, M.
Petrini, G. Bartoli, P. Melchiorre, Angew. Chem. Int. Ed. 2008, 47, 8707;
Angew. Chem. 2008, 120, 8835; b) P.-G. Cozzi, F. Benfatti, L. Zoli, Angew.
Chem. Int. Ed. 2009, 48, 1313; Angew. Chem. 2009, 121, 1339.
[4] For organocatalytic stereoselective SN1 alkylations of ketones catalyzed
by chiral Brønsted acids, see: L. Song, Q. X. Guo, X. C. Li, J. Tian, Y. G.
Peng, Angew. Chem. Int. Ed. 2012, 51, 1899; Angew. Chem. 2012, 124,
1935.
[5] For organocatalytic stereoselective SN1 alkylations of aldehydes cata-
lyzed by bifunctional amino-hydrogen bond donating organocatalysts,
see: a) A. R. Brown, W.-H. Kuo, E. N. Jacobsen, J. Am. Chem. Soc. 2010,
132, 9286; b) B. Xu, Z. L. Guo, W.-Y. Jin, Z.-P. Wang, Y.-G. Peng, Q.-X. Guo,
Angew. Chem. Int. Ed. 2012, 51, 1059; Angew. Chem. 2012, 124, 1083.
[6] For organocatalytic stereoselective alkylations of aldehydes with stable
and readily available carbenium ions, see: a) A. Gualandi, E. Emer, M. G.
[18] As far as we are aware, a single report concerning the catalytic intermo-
lecular a-amidoalkylation of aldehydes and ketones was known at the
outset of this project beside our own contributions (see Ref. [18b]). The
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6021
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
coupling used harsh reaction conditions (PTSA catal., benzene, azeo-
tropic conditions) and a very particular type of N,O-acetal; see: a) G. Bo-
bovsky, J. Shavel, J. Heterocycl. Chem. 1980, 17, 277; for a few examples
of known intramolecular a-amidoalkylations of ketones mediated by an
acidic promoter, see: b) T. Fujita, H. Nagasawa, Y. Uto, T. Hashimoto, Y.
Asakawa, H. Hori, Org. Lett. 2004, 6, 827; c) W. G. Earley, T. Oh, L. E. Over-
man, Tetrahedron Lett. 1988, 29, 3785; d) I. H. Sµnchez, F. J. López, H. J.
Flores, M. I. Larraza, Heterocycles 1983, 20, 247; e) for more recent exam-
ples, see Refs. [17,30].
tion with the same efficiency as their acetate counterpart does in some
other acid-catalyzed amidoalkylations. We did not use this type of ester
here.
[29] The diastereomeric ratios are not shown in the manuscript for clarity
but are given in the Supporting Information.
[30] a) F. Grau, A. Heumann, E. DuÇach, Angew. Chem. Int. Ed. 2006, 45,
7285; Angew. Chem. 2006, 118, 7443; b) J. Godeau, F. Fontaine-Vive, S.
Antoniotti, E. DuÇach, Chem. Eur. J. 2012, 18, 16815.
[31] J. Toullec, Tetrahedron Lett. 1984, 25, 4401; we noticed that the reported
constant values were measured in water, which is a totally different en-
vironment from our reaction media.
[32] Ketones that exhibit poor reactivity in this room temperature alkylation
procedure are easily amidoalkylated at temperature of 708C and
above; results will be published in due course.
[33] E. Boess, C. Schmitz, M. Klussmann, J. Am. Chem. Soc. 2012, 134, 5317.
[34] This comparative analysis must be considered carefully because the
CDC reactions in the papers listed below were executed at tempera-
tures of 408C or above, which constitute conditions that probably con-
tribute to smoothing possible discriminations between the different ke-
tones that were used; see: a) Y. M. Shen, M. Li, S. Z. Wang, T. G. Zhan, Z.
Tan, C. C. Guo, Chem. Commun. 2009, 953; b) . PintØr, A. Sud, D. Sure-
shkumar, M. Klussmann, Angew. Chem. Int. Ed. 2010, 49, 5004; Angew.
Chem. 2010, 122, 5124; c) J. Xie, H. Li, J. Zhou, Y. Cheng, C. Zhu, Angew.
Chem. Int. Ed. 2012, 51, 1252; Angew. Chem. 2012, 124, 1278.
[35] Direct alkylations of non-aromatic acyclic ketones by oxonium ions gen-
erated from ethers through CÀH activation-based methods also demon-
strated branched (although incomplete) selectivities, suggesting that
the branched versus linear selectivity in these alkylations correlates to
the electrophilicity of the cation; see: a) Y. Zhang, C.-J. Li, J. Am. Chem.
Soc. 2006, 128, 4242; b) H. Richter, R. Rohlmann, O. Garcia-Manchano,
Chem. Eur. J. 2011, 17, 11622.
[19] This assumption is qualitative and based on comparative acid-catalyzed
alkylation experiments performed in our laboratory using N-acylimini-
ums and p-stabilized carbocations of known electrophilicity parame-
ters.
[20] Very recently, organocatalytic asymmetric direct amidoalkylations be-
tween aldehydes and N,O-acetals using synergistic acid/enamine cataly-
sis were reported; however, only particular N-acyliminium precursors
with high ionization susceptibility, mainly in the quinoline/isoquinoline
family, were tolerated; see: a) L. Mengozzi, A. Gualandi, P. G. Cozzi,
Chem. Sci. 2014, 5, 3915; b) S. Sun, Y. Mao, H. Lou, L. Liu, Chem.
Commun. 2015, 51, 10691; c) F. Berti, F. Malossi, F. Marchetti, M. Pineschi,
Chem. Commun. 2015, 51, 13694; less biased N,O-acetals were clearly
shown not to react (Ref. [20c]); similar unsuccessful (unpublished) re-
sults were observed in our laboratory, thereby evidencing the peculiar
reactivity of N-carbamoyl (iso)quinolinium ions, which apparently tends
to converge on that of the weakest and related N-aryl iminiums (Ref.
[33]).
[21] Similar direct aminoalkylation of enolizable aldehydes/ketones with N-
aryliminiums produced in situ from N-aryl tetrahydroisoquinolines were
also recently reported; oxidative CÀH and aminocatalytic activation
modes were merged in a synergistic way; see: a) J. Zhang, B. Tiwari, C.
Xing, X. Chen, Y. R. Chi, Angew. Chem. Int. Ed. 2012, 51, 3649; Angew.
Chem. 2012, 124, 3709; b) G. Zhang, Y. Ma, S. Wang, W. Kong, R. Wang,
Chem. Sci. 2013, 4, 2645; for a selection of papers documenting similar
racemic chemistry, see Ref. [34].
[22] Recently, Kanai and co-workers reported an impressive catalytic enan-
tioselective coupling between cyclic N-Boc aminals and ketones; how-
ever, this transformation proceeded by a distinct mechanism that does
not involve an N-acyliminium ion; see: S.-L. Shi, X.-F. Wei, Y. Shimizu, M.
Kanai, J. Am. Chem. Soc. 2012, 134, 17019.
[23] The optimization study has revealed the acetoxy lactam 2a to be, as
anticipated, much more reactive than its methoxy and hydroxy (1a)
counterparts in this room-temperature catalytic coupling (see com-
ments in the main text related to Table 1 and Refs. [17b–d]).
[36] Details are given in the Supporting Information.
[37] B. Schweitzer-Chaput, M. Klussmann, Eur. J. Org. Chem. 2013, 666.
[38] For further extension of this chemistry by means of approaches for the
direct but noncatalytic oxidative cross deoxygenative couplings (CDCs)
between N-Cbz THIQs and ketones or aldehydes, see: a) X. Liu, B. Sun,
Z. Xie, X. Qin, L. Liu, H. Lou, J. Org. Chem. 2013, 78, 3104; b) Z. Xie, L.
Liu, W. Chen, H. Zheng, Q. Xu, H. Yuan, H. Lou, Angew. Chem. Int. Ed.
2014, 53, 3904; Angew. Chem. 2014, 126, 3985; c) W. Chen, H. Zheng, X.
Pan, Z. Xie, X. Zan, B. Sun, L. Liu, H. Lou, Tetrahedron Lett. 2014, 55,
2879.
[39] a) A. Gualandi, L. Mengozzi, E. Manoni, P. G. Cozzi, Catal. Lett. 2015, 145,
398.
[24] The unreacted N,O-acetals were systematically recovered in the form of
their hydroxy lactams 1, which were occasionally accompanied by low
amounts of the starting acetoxy lactam.
[25] K. Ishihara, A. Hasegawa, H. Yamamoto, Angew. Chem. Int. Ed. 2001, 40,
4077; Angew. Chem. 2001, 113, 4201.
[40] a) M. Rueping, C. M. R. Volla, I. Atodiresei, Org. Lett. 2012, 14, 4642;
b) Ref. [38a]; c) Z. Meng, S. Sun, H. Yuan, H. Lou, L. Liu, Angew. Chem.
Int. Ed. 2014, 53, 543; Angew. Chem. 2014, 126, 553; d) Z.-J. Wu, J. Qian,
T.-T. Wang, Z.-Z. Huang, Synlett 2014, 2072; e) S. Manojveer, R. Balamur-
ugan, Org. Lett. 2015, 17, 3600.
[26] a) I. Matsuda, K.-I. Komori, K. Itoh, J. Am. Chem. Soc. 2002, 124, 9072; K.-
I. Komori, K. Itoh, J. Am. Chem. Soc. 2002, 124, 9072; b) K. Mertins, I.
Novel, J. Kischel, A. Zapf, M. Beller, Angew. Chem. Int. Ed. 2005, 44, 238;
Angew. Chem. 2005, 117, 242; c) Z.-p. Zhan, S.-p. Wang, X.-b. Cai, H.-j.
Liu, J.-l. Yu, Y.-y. Cui, Adv. Synth. Catal. 2007, 349, 2097.
[27] a) P. A. Grieco, S. T. Handy, Tetrahedron Lett. 1997, 38, 2645; b) M. Rubin,
V. Gevorgyan, Org. Lett. 2001, 3, 2705; c) T. Schwier, M. Rubin, V. Ge-
vorgyan, Org. Lett. 2004, 6, 1999; d) S. H. Kim, C. Shim, A. N. Pau, H. Y.
Koh, M. M. Chang, B. Y. Chung, Y. S. Cho, Synthesis 2004, 1581; e) O.
Mendoza, G. Rossey, L. Ghosez, Tetrahedron Lett. 2010, 51, 2571.
[28] Previous results from our laboratory demonstrated that substrates
equipped with ester leaving groups, such as pivalate or benzoate, func-
[41] Tin tetrakis(trifluoromethanesulfonimide)-DMSO complex is now com-
mercially available from Sigma–Aldrich, CAS number 919356-21-1.
[42] While this manuscript was in preparation, Odell and co-workers pub-
lished an excellent paper on the direct amidoalkylation of ketones by
N-acyliminiums generated in situ by a dehydrative intramolecular uree/
aldehyde condensation approach; see: R. T. Sawant, M. Y. Stevens, L. R.
Odell, Eur. J. Org. Chem. 2015, 7743.
[43] S. Antoniotti, E. DuÇach, Chem. Commun. 2008, 993.
Received: November 26, 2015
Published online on March 15, 2016
Chem. Eur. J. 2016, 22, 6012 – 6022
www.chemeurj.org
6022
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim