α-Halogenoacetanilides as hydrogen-bonding organocatalysts that activate carbonyl bonds: Fluorine versus chlorine and bromine

Article abstract of DOI:10.1002/chem.201303662

α-Halogenoacetanilides (X=F, Cl, Br) were examined as H-bonding organocatalysts designed for the double activation of Ci￡O bonds through NH and CH donor groups. Depending on the halide substituents, the double H-bond involved a nonconventional Ci-H×××O in

Full text of DOI:10.1002/chem.201303662

DOI: 10.1002/chem.201303662
Full Paper
&
Hydrogen Bonds
a-Halogenoacetanilides as Hydrogen-Bonding Organocatalysts
that Activate Carbonyl Bonds: Fluorine versus Chlorine and
Bromine
Sylvain Koeller,[a] Coralie Thomas,[a, b] Frꢀderic Peruch,[b] Alain Deffieux,[b] Stꢀphane Massip,[c]
Jean-Michel Lꢀger,[c] Jean-Pierre Desvergne,[a] Anne Milet,[d] and Brigitte Bibal*[a]
Abstract: a-Halogenoacetanilides (X=F, Cl, Br) were exam-
ined as H-bonding organocatalysts designed for the double
activation of C=O bonds through NH and CH donor groups.
Depending on the halide substituents, the double H-bond
involved a nonconventional CÀH···O interaction with either
a HÀCXn (n=1–2, X=Cl, Br) or a HÀCAr bond (X=F), as
shown in the solid-state crystal structures and by molecular
modeling. In addition, the catalytic properties of a-halogen-
oacetanilides were evaluated in the ring-opening polymeri-
zation of lactide, in the presence of a tertiary amine as coca-
talyst. The a-dichloro- and a-dibromoacetanilides containing
electron-deficient aromatic groups afforded the most attrac-
tive double H-bonding properties towards C=O bonds, with
a NÀH···O···HÀCX2 interaction.
Introduction
for alternative H-bond donor groups that better activate the
substrate (increased electrophilicity) or enhance a selective ac-
tivation towards a unique substrate in the presence of several
reactants, is still valuable. Indeed, recent interest in one-pot re-
actions involving H-bonding catalysts[3] and the need for dual
H-bonding systems in polymerization[4,2d] (H-bond donor and
acceptor) has spurred on the emergence of new, adjustable H-
bond donor structures that are compatible with other catalysts
and reactants.
Hydrogen-bonding catalysis has developed considerably over
the last decade, as demonstrated by the numerous reviews
dealing with the topic.[1] In particular, activation of carbonyl
bonds by hydrogen-bond (H-Bond) donor compounds is a blos-
soming field of research, due to the extended scope of asym-
metric transformations (Michael type additions, cycloadditions,
radical cyclization, Baylis–Hillman reaction, cyanosilylation) and
because of the current challenges in the controlled polymeri-
zation of cyclic esters.[2] Thus, conventional H-bond donor
groups have been exploited to activate C=O bonds, using
more or less complex structures based on thiourea, alcohol,
phenol, amide, sulfonamide, protonated amine, guanidine and
amidine, and phosphoric acid moieties. However, the search
We previously demonstrated that an amidoindole platform
efficiently activates the C=O bond in the ring-opening poly-
merization (ROP) of lactide, as a model reaction (Scheme 1).[5]
[a] Dr. S. Koeller, Dr. C. Thomas, Dr. J.-P. Desvergne, Dr. B. Bibal
Institut des Sciences Molꢀculaires, UMR CNRS 5255
Universitꢀ de Bordeaux
351 cours de la Libꢀration, 33405 Talence (France)
Fax: (+33)54000-6158
E-mail: b.bibal@ism.u-bordeaux.fr
[b] Dr. C. Thomas, Dr. F. Peruch, Dr. A. Deffieux
Laboratoire de Chimie des Polymꢁres Organiques, UMR 5629
Universitꢀ de Bordeaux and CNRS
Scheme 1. H-Bonding amide catalysts with two donor groups.
16 avenue Pey-Berland, 33600 Pessac (France)
[c] Dr. S. Massip, Prof. Dr. J.-M. Lꢀger
FRE CNRS 3396 Pharmacochimie
Universitꢀ de Bordeaux
The mechanism relied on the formation of a double H-bond
between the C=O of the monomer and the two NH groups of
the catalyst. The acidity of the amide proton was enhanced by
electron-withdrawing groups on the aromatic group, and the
presence of two H-bond donor groups was necessary to better
complex the lactide unit, as demonstrated by comparative cat-
alytic properties. However, some disadvantages were ob-
146 rue Lꢀo Saignat, 33076 Bordeaux (France)
[d] Prof. Dr. A. Milet
Dꢀpartement de Chimie Molꢀculaire, UMR 5250, ICMG FR-2607
Universitꢀ Joseph Fourier-Grenoble I
BP 53, 38041 Grenoble (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201303662.
Chem. Eur. J. 2014, 20, 2849 – 2859
2849
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
served: i) indole as a second H-bond donor was poorly adjusta-
ble thus limiting the scope of activated substrates, and ii) some
amide catalysts tended to interact with the cocatalyst (tertiary
amine) employed in ROP of cyclic esters, rather than with the
substrate. This undesired phenomenon led to poor conver-
sions.[5b]
Weak nonconventional CÀH···O interactions were recently
highlighted in natural and artificial systems that offer comple-
mentary interactions useful for structure and function (protein
folding, enzyme activity, advanced materials, organic synthe-
sis).[6] Concerning their contribution to organocatalysis, the N+
ÀCH moiety within quaternary ammonium groups has been
exploited in asymmetric CÀC bond formations[7] and ROP,[8]
whereas less activated CH groups were reported in a few ex-
amples as probable participants in the stereocontrol of such
reactions.[9]
Taking advantages of the H-bonding properties of anilides
and keeping in mind the need for an adjustable activation of
C=O bonds even in the presence of a cocatalyst or a second
substrate,[3,4] we envisioned that a-halogenoacetanilides pos-
sessing a second activated H-bond donor group such as HÀ
CXn (n=1–2), should be attractive catalysts (Scheme 1). If the
HÀCXn bond is not available for the substrate (steric hindrance,
unfavored conformation), then a secondary but less activated
H-bonding site could be the HÀCAr bond on the electron-defi-
cient aromatic substituent (Scheme 1). All the NH and CH acidi-
ties could be easily modulated, depending on the electronic
nature of the substituents. In addition, both NH and CH
groups can be positioned in a geometry that is similar to
those of thioureas, which are known to be efficient double H-
bonding catalysts. We therefore decided to investigate acetani-
lides 1 and a-halogenoacetanilides 2–9 (Scheme 2), with
a range of aromatic substituents (R=H, CF3, NO2) and XnCH3Àn
groups (n=1–3; X=F, Cl, Br) of different electron-withdrawing
nature.[10] Catalysts 1–9 were evaluated in a model reaction,
that is, the ROP of lactide. In this reaction, cyclohexyldimethyl-
amine (CyNMe2) and (À)-sparteine (Sp) were chosen as cocata-
lysts, taking into account their different strength in H-bond for-
mation, to activate the initiator and the growing chain, which
both hold an alcohol function.[2] Thus, the potential inhibition
of the H-bond donating properties of 1–9 towards the sub-
strate was evaluated in the presence of a tertiary amine.
It is worth noting that a-halogenoacetanilides have, to the
best of our knowledge, never been described as organocata-
lysts. a-Chloroacetanilides have been employed as herbicides,
pesticides, bioactive compounds[11] as well as synthetic precur-
sors of molecules intended for pharmaceutical or biomedical
purposes (b-lactams, oxindoles, MRI contrast agents).[12]
Sigman et al. studied the H-bonding behavior of oxazolines
Scheme 2. Acetanilides 1–9 investigated as H-bonding organocatalysts, co-
catalysts and lactide.
Their properties were also found to be governed by parame-
ters other than acidity. Indeed, O’Hagan and others[14] studied
a-monofluoroamides and their particular arrangement for
which the CÀF bond prefers anti and syn conformations versus
the carbonyl and the NH bond, respectively, mainly as a result
of the interaction between fluorine lone pairs and the NÀH s*
orbital. These results were exploited to control peptide folding
and asymmetric induction.[14d] Smith et al. showed that difluor-
oacetanilides could form an intramolecular double H-bond
with the C=O bond of their carbamate moiety, thus inverting
the CÀF bond towards a pseudo-conformation anti to the NH
bond.[15] This conclusion was based on the X-ray structure of
compounds and quantum calculations. Consequently, dihalo-
genated acetanilides can be considered as promising original
candidates for C=O activation in organocatalysis. In this con-
text, the role of the CH3ÀnXn group in a-halogenoacetanilides
can be limited to an electron-withdrawing group and/or be ex-
tended to an H-bonding moiety, the properties of which are
still unknown. To explore the H-bonding properties of com-
pounds 1–9, the present study encompasses solid-state stud-
ies, molecular modeling in the gas phase, and an investigation
of catalysis in solution.
provided with an alcohol and halogenated acetamide groups Results and Discussion
in an enantioselective hetero Diels–Alder reaction.[13] For these
Synthesis of a-halogenoacetanilides
alkyl-NH-CO-CH3ÀnXn derivatives, the acidity of the amide
proton (CF3 >CCl3 >CHCl2 >CH2F>CH2Cl) was shown to direct-
ly impact both the reaction rate and the enantioselectivity. In
particular, the most acidic oxazolines induced higher conver-
sions. Herein, a-halogenoacetanilides 2–9 were more acidic
than aliphatic acetamides and their structures were simpler.
The preparation of organocatalysts 1–9 was achieved in high
yields (84–98%) from commercial reagents (Scheme 3). Accord-
ing to classical procedures (A or B depending on the availabili-
ty of reactants),[16] acetanilides 1 and a-halogenoacetanilides
2–9 were obtained from the condensation of the correspond-
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2850
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
and their angles are between 150 and 1808, as expected for in-
teractions governed by electrostatics.[6c]
To extrapolate the potential activity of double H-bonding or-
ganocatalysts, a selection of dihalogenated compounds with
electron-withdrawing aromatic substituents were crystallized
and submitted to X-ray analysis. Monocrystals of 3h (X=F), 3i
(X=F), 6d (X=Cl), 6i (X=Cl), and 9i (X=Br) were grown in di-
chloromethane by slow diffusion of pentane. Selected interac-
tions (length and angle) and torsion angles are listed in Table 1
(for detailed crystallographic data see the Supporting Informa-
Scheme 3. Synthesis of a-halogenoacetanilides 1–9.
ing anilines (R=H, CF3, NO2) and freshly distilled acid chlorides
or acyl anhydrides (R1=CH3ÀnXn) in dichloromethane. Among
the series 1–9, the new compounds were fully characterized
(see the Supporting Information). It should be noted
that several a-tribromoacetanilides with electron-
Table 1. Hydrogen bonds and torsion angles in 3h, 3i, 6d, 6i, and 9i monocrystals.[a]
withdrawing groups were also prepared (not shown)
but were too unstable in solution to be properly
evaluated as organocatalysts.
3h
p-NO2
CHF2
3i (50/50)[c]
m,m’-NO2
CHF2
6d
p-CF3
CHCl2
6i
m,m’-NO2 m,m’-NO2
CHCl2 CHBr2
9i
R
R1
C=O···HÀN
C=O···HÀCX
NÀH···XÀC
<HNÀCO>
<OCÀCH>
2.15 (165) 2.07 (168)
none none/2.88 (140)
2.22 (110) 2.31 (106)/3.00 (83) none
2.01 (167) 2.11 (162) 2.10 (165)
2.46 (148) 2.31 (146) 2.26 (145)
Molecular interactions in solution
none
À171
À164
24 (H)
none
À171
À166
22 (H)
Host/guest titrations were conducted on three model
acetanilides to assess their binding ability towards
lactide in CDCl3 (see the Supporting Information).
Compounds with a trifluoromethyl substituent on
the aromatic group were selected to better evaluate
the impact of CX3 and CHX2 groups. A binding con-
stant (Kass) of 4mÀ1 was determined between trifluori-
nated compound 4e and lactide, showing that the
NÀH bond was accessible to lactide, even with
a bulky substituent (COCF3). Dichlorinated catalysts
178
74
180
73/160
À179
176
À3 (H)
<HÀN/CÀA>[b] 14 (F)
9 (F)/À19 (H)
[a] Length of H-bonds in ꢂ and angles in degrees (bracket). [b] Angle of the conforma-
tional tweezer pointing at the O=C bond, either single H-bond donor (A=F) or
double H-bond donor (A=H). [c] Data for two conformers are shown: NH/CF syn con-
formation followed by NH/CH syn conformation.
6e and 6d also associated with lactide, showing binding con-
stants of 2 and 1mÀ1, respectively. Because these binding con-
stants were very low, no valid comparison between the Kass
values and the catalyst structures could be made. However,
previous studies demonstrated that even weak H-bonds could
trigger ROP reactions.[4b,5b,17] Thus, despite the low values of
Kass, the data support the possible activation of a C=O bond by
the aforementioned acetanilides in solution.
tion). All crystals are packed through strong intermolecular H-
bonds between amides, displaying usual distances and angles:
d(NÀH···O=C)=2.01–2.15 ꢂ and (NHO)=162–1688. Aromatic
groups are arranged in more or less ordered layers depending
on the geometry of H-bonds and steric hindrance. Noticeable
differences appear between difluorinated and other dihalogen-
ated acetanilides. Interestingly, difluorinated amide 3h (Fig-
ure 1a) adopts a different conformation to that described for
3a:[15] one fluorine atom interacts with the adjacent NÀH
group (d(F···HÀN)=2.55 ꢂ), as already seen with monofluoro-
acetanilides.[14a–b] The NÀH and CÀF bonds are almost parallel
(HN/CF)=148. The C=O bond also interacts with the NH group
(d=2.15 ꢂ) and the aromatic proton (d(C=O···HÀCAr)=2.56 ꢂ
and (OHC)=1108.
As a result of disorder in the crystal, difluoroacetanilide 3i is
observed in NH/CÀF syn and NH/CÀH syn conformations in
equal proportion (Figure 1b and c, respectively). In the NH/CÀ
F syn conformation (Figure 1b), the NH bond is almost parallel
with the CÀF bond (HN/CF)=98 and then close to the fluorine
atom (d(F···HÀN)=2.31 ꢂ). A single intermolecular H-bond be-
tween the NÀH and C=O bonds is observed (d=2.07 ꢂ). In
contrast to 3h, no interaction between C=O and the acidic ar-
omatic proton is seen in 3i, despite similar conformations. In
the NH/CH syn conformation ((HN/CH)=À148, Figure 1c),
a double H-bond to carbonyl is composed of the same strong
NH···O=C (d=2.07 ꢂ) and weak F2CÀH···O=C interactions (d=
2.88 ꢂ). However, due to the weakness of this secondary inter-
action involving the CHF2 group, it could be considered that
compound 3i mainly interacts with the second molecule
X-Ray diffraction
Whereas reports of a-chloroacetanilides in the solid-state are
common because of their attractive properties,[11,12] no X-ray
structure of a-bromoacetanilide and only a few of a-fluoroace-
tanilides have been described. Thus, only the crystal structures
of monochlorinated 5a[18] and 5h,[19] dichlorinated 6a[20] and
6h,[21] and trichlorinated 7a[22] and 7h[23] acetanilides have
been reported. Interestingly, the authors focused on strong in-
termolecular H-bonds between the NH and the carbonyl
groups stabilizing the packing of a-chloroacetanilides into
polymeric chains (NÀH···O=C distances ranging from 2.04 to
2.23 ꢂ and NHO angles of 155–1788). No comments were
made on possible interactions between the acidic XnCH group
and the carbonyl, even though most structures present parallel
alignments of NH and XnCH bonds. Recently, compound 3a
was fully described with a NÀH···O···HÀC interaction involving
short distances (d(NÀH···O=C)=2.0 ꢂ and d(F2CÀH···O=C)=
2.5 ꢂ).[15] In general, the range of distances for short CÀH···O in-
teractions are d(CÀH···O)=2.1–2.5 ꢂ and d(CÀH···O)=3.1–3.5 ꢂ
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2851
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ometry optimizations at the B3LYP/6-31+G** level were ach-
ieved on l-lactide, a selection of catalysts, and mixtures of
both species, in vacuum.[24] Herein, the most favorable com-
plexes are presented, however, complexes of similar energy
could be present in solution and also activate the substrate. A
series of six a-acetanilides provided with a p-nitrobenzene sub-
stituent were chosen to compare the effect of halide atom(s)
on H-bonding: 10 (R1 =CH3), 2h (CH2F), 3h (CHF2), 4h (CF3),
6h (CHCl2), and 9h (CHBr2). A compound with an m,m’-dinitro-
benzene group (2i; CH2F) was also considered to assess the
impact of a m,m’-NO2 substituent. Although the simulations
are not fully representative of the experimental conditions of
the reaction (solvent, cocatalyst, growing chain), they provide
information on the geometry and energy of the lactide com-
plexes (Figure 2, Table 2) and thus allow the H-bonding proper-
ties of several catalysts towards the same substrate to be com-
pared. To better assess the existence of weak interactions,
some complexes were analyzed through their electron density,
using the NCIPLOT program developed by Yang et al.[25] The
gradient surfaces obtained within the complexes indicate the
location and the strength of any noncovalent interactions,
Figure 1. Hydrogen-bonded a-dihalogenoacetanilides in the crystal: a) 3h,
b) 3i in NH/CF syn conformation, c) 3i in NH/CH syn conformation, d) di-
chlorinated 6d, e) dichlorinated 6i, f) dibrominated 9i.
through a sole NH···O=C interaction, but experience two con-
formations of equal probabilities for the CHF2 group.
Dichlorinated compounds 6d and 6i as well as dibrominat-
ed 9i form intermolecular double H-bonds between the car-
bonyl group and the NH/CH bonds (Figure 1d, e and f, respec-
tively). These nonconventional X2CÀH···O=C interactions are
quite strong, with distances of 2.46, 2.31, and 2.26 ꢂ, respec-
tively and similar CHO angles (145–1488). The NH and CH
bonds are almost parallel (HN/CH)=À2 to +228. Finally, when
compounds 3i, 6i, and 9i, with identical aromatic groups
(m,m’-NO2), are compared, the length of the NH···O=C H-bond
is slightly shorter for 3i (single interaction, 2.07 ꢂ) than for 6i
(double interaction, 2.11 ꢂ), and 9i (double interaction, 2.10 ꢂ).
These results indicate that, in the solid-state, a-dihalogeno-
acetanilides provided with electron-deficient aromatics interact
strongly through intermolecular single or double H-bond(s),
depending on the nature of the halide: a-difluoroacetanilides
present classical NÀH···O=C interaction (and possibly a CArÀ
H···O=C interaction), whereas a-dichloro- and a-dibromo- ace-
tanilides show an original NÀH···O···HÀCX2 interaction. Impor-
tantly, this nonconventional H-bond is particularly strong (d=
2.26–2.46 ꢂ).
Molecular modeling
To the best of our knowledge, intermolecular H-bonding of a-
halogenated acetanilides towards a substrate has never been
modeled. To better predict the activation of the C=O bond, ge-
Figure 2. Molecular modeling at the B3LYP/6-31+G** level of complexes be-
tween l-Lactide and a) 10 (left) with the NCI analysis (right), b) 2h, c) 2i,
d) 3h, e) 4h, f) 6h, and g) 9h.
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2852
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
same aromatic substituent: À5.42, À6.84, and
À7.58 kcalmolÀ1 for 2h, 3h, and 4h, respectively.
Table 2. Hydrogen bonds and energy of interaction in modeled complexes between
l-Lactide and 1h, 2h, 3h, 4h, 6h, and 9h.[a]
Thirdly, dichlorinated and dibrominated 6h and
1h
CH3
2h
CH2F
3h
CHF2
4h
CF3
6h
CHCl2
9h
CHBr2
9h, respectively, both adopt a NH/CH syn conforma-
tion ((NH/CH)=238 and 138, respectively, Figure 2 f
and g) to complex the carbonyl group of l-lactide
through a double H-bond, as seen in the crystal
structures of 6i and 9i. Both lengths are short (C=
O···HÀN distances are 2.00 and 2.02 ꢂ for 6h and 9h,
respectively) and nonconventional C=O···HÀCX2 dis-
tances were 2.39 and 2.37 ꢂ, respectively. Concerning
the 9h/lactide complex, an additional weak interac-
tion was observed between the acidic aromatic
proton of acetanilide and the intracyclic oxygen O2
of lactide (d(CArÀH···O2)=2.86 ꢂ and (CHO2)1618). The
latter weak binding between CArÀH and an intracyclic
oxygen atom was also observed by Schreiner between a thiour-
ea provided with m,m’-CF3 groups and a lactone.[27] The ener-
gies of interaction are stronger for dichloro- and dibromoace-
tanilides complexes than the previous complexes: À8.66 kcal
molÀ1 (for 6h/lactide) and À8.47 kcalmolÀ1 (for 9h/lactide).
Thus, the lowest negative energies of interactions were calcu-
lated for trifluorinated, dichlorinated, and dibrominated com-
pounds 3h>6hꢀ9h, respectively, indicating that these mole-
cules could be the most attractive activators of C=O bonds.
Concerning the m,m’-NO2 derivative (Figure 2c), the calculat-
ed 2i/lactide complex was found to interact through a slightly
shorter double H-bond than 2h/lactide, with d(NÀH···O=C)=
2.15 ꢂ ((NHO)=1568) and d(CArÀH···O=C)=2.20 ꢂ ((HCO)=
1468). Importantly, the meta-NO2 group displays a secondary
H-bond with the methyl group of lactide (d(CÀH4···OÀNO)=
2.79 ꢂ and (HCO)=1658). The distance between the lactide
methyl group and the acidic aromatic proton was calculated
R1
C=O···HÀN
C=O···HÀC
C=O···HÀCAr
NÀH···XÀC
2.09 (171) 2.13 (164) 2.04 (170) 1.96 (176) 2.00 (169) 2.02 (170)
2.97 (139) none none none 2.39 (135) 2.37 (144)
2.58 (141) 2.44 (143) 2.49 (141) 2.66 (136) none
none
–
2.22 (F)
2.1 (F)
À5.42
2.33 (F)
27.3 (F)
À6.84
2.58 (F)
50.8 (F)
À7.58
2.16 (H)
23.5 (H)
À8.66
2.12 (H)
13 (H)
À8.47
<HÀN/CÀA>[b] 0.9 (H)
[c]
DEint
À6.15
[a] Length of H-bonds in ꢂ and angles in degrees (bracket). [b] Angle of the conforma-
tional tweezer pointing at the O=C bond, either single H-bond donor (A=F) or
double H-bond donor (A=H). [c] Energy of interaction in kcalmolÀ1
.
using a color code (blue for strong noncovalent interaction,
green for weak interactions, and red for nonbonded overlap).
The optimized structures of catalysts (without lactide)
showed that all amide moieties are coplanar to the aromatic
ring and in trans conformation. All fluorinated compounds (2h,
2i, 3h, 4h) and the dibrominated compound 9h adopt a NH/
C-X syn conformation, whereas dichlorinated 6h has a NH/CH
syn conformation. In the complexes with l-lactide, the confor-
mations of acetanilides are identical, except for dibrominated
9h, which adopts a NH/CH syn conformation. The cost of the
latter geometry change is +4.00 kcalmolÀ1. The calculated de-
formation is +1–2 kcalmolÀ1 for all other complexes (see the
Supporting Information). All the NH groups of a-acetanilides
are strongly H-bonded[26] to l-lactide (d(NÀH···O=C)=1.96–
2.13 ꢂ and (HNO)=164–1768).
Firstly, as seen in Figure 2a, nonhalogenated compound 10
interacts with the C=O group of l-lactide through its NÀH
bond (d(NÀH···O=C)=2.09 ꢂ) and a CArÀH bond (d(CArÀH···O= to be 2.98 ꢂ. The NCI analysis confirms the existence of H-
C)=2.58 ꢂ), whereas the CÀH2 bond is parallel with the NH
bond and weakly H-bonded to the intracyclic oxygen O2 of
lactide (d(CÀH···O2)=2.90 ꢂ and (CHO2)=1768). The energy of
interaction (DEint) was calculated to be À6.15 kcalmolÀ1. The
noncovalent interaction (NCI) analysis of the 10/lactide com-
plex reveals the same interactions as those detected from the
atomic distance measurements. Logically, the blue-green gradi-
ent isosurfaces involving NH and CArH bonds are typical of
weak H-bonds, whereas the bright-green isosurface involving
the methyl group is characteristic of weaker Van der Waals in-
teractions.
bonds between the carbonyl and the NH/CH tweezer and Van
der Waals forces for the secondary interaction due to the
meta-NO2 group (see the Supporting Information). The interac-
tion energy 2i/lactide is more negative than 2h/lactide (À6.56
vs. À5.42 kcalmolÀ1, respectively), showing the impact of
a m,m’-NO2 group compared with a para-substituent.
For the modeled a-fluoroacetanilides complexes (Figure 2b
and d–e), the CÀF bonds point toward the acidic intracyclic
protons of the lactide without making realistic H-bonds (dis-
tances greater than 3.00 ꢂ: d(CÀF···H3ÀC)=3.09, 3.51, and
3.45 ꢂ for 2h, 3h, and 4h, respectively). This effect results in
the unparallel positioning of the lactide plane towards acetani-
lides, and an eclipsed conformation (HÀN/CÀF=518) appears
in the 4h/lactide complex. In addition, the unusual preferred
approach of lactide can also account for the interaction energy
of 2h/lactide, which is less negative (À5.42 kcalmolÀ1) than
one of the 10/lactide complex (À6.15 kcalmolÀ1). Thus, the ge-
ometry of these lactide complexes is affected by very weak CÀ
F···HÀC interactions.[28] Notably, no CÀX···HÀC interactions were
observed with chlorinated or brominated catalysts that were
designed for double H-bonding a carbonyl group through HÀ
CXn groups.
Secondly, a-fluoroacetanilides 2h, 3h and 4h (Figure 2b
and d–e) exhibit a double H-bond with the carbonyl of lactide,
involving their NÀH group (d(NÀH···O=C)=2.13, 2.04 and
1.96 ꢂ, respectively) and a CArÀH bond (d(CArÀH···O=C)=2.44,
2.49 and 2.66 ꢂ, respectively). Due to the NH/CF conformation,
no interaction between the HÀCFn bond and the C=O bond of
lactide was observed. The strength of the main NÀH···O=C H-
bond between fluorinated catalysts and lactide increases with
the number of fluorine atoms, whereas the strength of secon-
dary NH···HÀCAr and NH···FÀC interactions decrease. The values
of the interaction energy also became more negative when
the number of fluorine atoms were increase while keeping the
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2853
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Finally, B3LYP/6-31+G** calculations on a-halogenoacetani-
lide/l-lactide complexes in vacuum predict that the C=O acti-
vation occurs through different H-bonds depending on the
nature of the halide; a-dichloro- and a-dibromoacetanilides
adopt a NH/CH syn conformation that favor a strong double H-
bond with C=O group, whereas a-fluoroacetanilides display
a double H-bond involving the NH group and an aromatic
proton. According to the calculations, weak secondary interac-
tions (CÀF···HÀC or ONÀO···HÀCH2) can influence the geometry
of the complex. Modeling confirms the solid-state study reveal-
ing the favored conformations of a-halogenoacetanilides to-
wards C=O bonds and also supports the binding experiments
in solution, highlighting the C=O bond activation even with tri-
halogenoacetanilides. Although the calculations do not take
into account the complexity of the reaction medium (solvent,
other H-bonding compounds), a catalytic activity for the a-hal-
ogenoacetanilides can be envisioned in the ROP of lactide in
dichloromethane.
lyst, and 4 ꢂ MS) did not trigger the polymerization independ-
ently. Only Sp allowed a 10–15% of conversion in the presence
of an initiator and in the absence of H-bond donor.[5b] The H-
Bond donor/H-Bond acceptor/initiator system was employed in
5:5:5 mol% ratio versus the monomer to initiate and propa-
gate the reaction in a controlled fashion.[17b] The crude polyes-
1
ters were analyzed by H NMR (% conv.), size-exclusion chro-
matography [molar mass (Mn), Dispersity (ꢃ)], and MALDI-TOF
(chain-end fidelity, distribution) (see the Supporting Informa-
tion). In accordance with a living-like mechanism, the average
molar masses were close to the theoretical values and the dis-
persity was very narrow (ꢃ<1.1). The percentage conversion
for acetanilides 1, a-fluoroacetanilides 2–4, a-chloroacetani-
lides 5–7, and a-bromoacetanilides 8–9 are presented in
Table 3.
As a first general observation (Table 3), all reactions conduct-
ed in the presence of ortho-substituted catalysts by a CF3 (1b,
2b, 3b, 4b, 5b, 6b) or a NO2 group (1 f, 2 f, 3 f, 4 f, 5 f, 6 f),
display low conversions (13–36%) irrespective of which cocata-
lyst was used, compared with the corresponding meta- and
para-substituted catalysts (27–100% conv.). Strong intramolec-
ular H-bonds are proposed to disable the activation of lactide
for all ortho-substituted acetanilides. Indeed, some acetanilides
provided with accepting ortho-substituents (NO2, CO2R,
SO2NR2) were shown to present an intramolecular H-bond be-
tween the NH amide and its aromatic substituent.[29] This is the
first time that a CF3 group has been reported as an intramolec-
ular accepting group in compounds 1b, 2b, 3b, 4b, 5b, and
6b. Furthermore, the impact of meta- versus para- substituents
on the conversion yield was similar, with a Æ5–10% difference
(within the range of experimental error), except for 1g versus
1h (42 vs. 67%). The reason for the large difference in catalytic
activity of the latter is not yet clear. Because acetanilides 1 are
less efficient catalysts, we focused on the halogenated com-
pounds and will present this case later. However, a detailed
H-Bonding catalysis
In contrast to the solid-state and modeling studies, the catalyt-
ic activity of a-halogenoacetanilides was also evaluated for the
lactide in the presence of a tertiary amine as a cocatalyst.
Thus, two factors can influence the outcome of the reaction:
i) The H-bonding properties of 1–9 towards the substrate, and
ii) the acid-base interactions between 1–9 and the cocatalyst.
The following discussion will take into account both influen-
ces.
The ROP of lactide was undertaken under classical condi-
tions with lactide (0.7m) in dichloromethane, at 208C, in the
presence of 4-biphenylmethanol as the initiator, CyNMe2 or
(À)-sparteine (Sp) as activator of the initiator/polymer growing
chain, and 4 ꢂ molecular sieves.[5,8,17] Preliminary experiments
indicated that each component of the system (initiator, cata-
Table 3. Catalytic properties of acetanilides 1–9 in the ROP of dl-Lactide; percentage of conversion in the presence of Sp (or CyNMe2).[a,b]
R/R1
CH2Me
CH2F
CHF2
CF3
CH2Cl
CHCl2
CH2Br
CHBr2
H
o-CF3
m-CF3
p-CF3
m,m’-CF3
o-NO2
m-NO2
p-NO2
m,m’-NO2
1a, 29
1b, 25 (14)
1c, 27
1d, 29
1e, 56
1 f, 25 (13)
1g, 42
1h, 67
2a, 36
n.d.[c]
2c, 44
2d, 37
2e, 94
n.d.[c]
2g, 59
2h, 51
2i, 100 (40)
3a, 62 (20)
3b, 23 (19)
3c, 96 (37)
3d, 95 (27)
3e, 100 (46)
3 f, 19 (15)
3g, 97 (49)
3h, 96 (44)
3i, 100 (79)
4a, 29 (10)
n.d.[c]
5a, 58
5b, 22
6a, 88 (36)
6b, 36 (28)
6c, 98 (62)
6d, 99 (40)
6e, 100 (66)
6 f, 32
6g, 100 (87)
6h, 100 (90)
6i, 100 (100)
8a, 73
9a, 84 (28)
9b, n.d.[c]
9c, 97 (53)
9d, 97 (51)
9e, 100 (85)
9 f, n.d.
9g, 100 (65)
9h, 100 (65)
9i, 100 (100)
8b, n.d.[c]
8c, 96
n.d.[c]
5c, 87 (49)
5d, 81 (24)
5e, 100 (40)
5 f, 22
5g, 91 (61)
5h, 92 (59)
5i, 100 (82)
n.d.[c]
8d, 93
4e, 15 (10)
n.d.[c]
n.d.[c]
n.d.[c]
8e, 100 (50)
8 f, n.d.c
8g, 97
8h, 97
8i, 100 (90)
1i, 100 (34)
4i, 10 (15)
[a] Conditions of ROP: dl-Lactide (0.7m in dichloromethane), H-bond donor catalyst (5 mol%), Sp (or CyNMe2) cocatalyst (5 mol%), 4-biphenylmethanol as
initiator (5 mol%), 4 ꢂ molecular sieves, 208C, 24 h. [b] Determined by 1H NMR spectroscopic analysis; percent conversion in the presence of CyNMe2 is
given in parenthesis. [c] n.d.: not determined.
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2854
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
analysis of m- versus p-substituted compound conversions in-
dicates that for the moderately activated compounds (among
2c–6d), m-substituted catalysts induce slightly higher conver-
sion than the corresponding p-substituted derivatives, in a sys-
tematic manner: thus, in the presence of Sp, 2c versus 2d (44
vs. 37%), 2g versus 2h (59 vs. 51%), and 5c versus 5d (87 vs.
81%). The phenomenon is clear in the presence of CyNMe2,
with 3c versus 3d (37 vs. 27%), 5c versus 5d (49 vs. 24%),
and 6c versus 6d (62 vs. 40%). Although, in some cases, the
difference in percent conversion was within the range of the
experimental error, repeated experiments (eight times) con-
firmed the same tendency (m->p-) was found for the quoted
compounds. This interesting phenomenon could be attributed
to i) a slight difference in amide pKa between m- and p- deriva-
tives, as reported for a thiourea,[30] which might result in slight-
ly different H-bonding properties, or ii) to secondary interac-
tions between m-NO2/CF3 and the methyl group of lactide, as
seen in the computed 2i/lactide complex (Figure 2c). Further
experiments are in progress to explain and exploit the H-bond
accepting properties of m-substituted catalysts with appropri-
ate substrates, which has been poorly documented so far.
As a second general remark, based on conversion, the elec-
tron-withdrawing properties of CF3 and NO2 groups followed
the expectations (CF3 <NO2). In both cases, two electron-with-
drawing groups on the aromatic function (instead of only one
group) are required to reach the highest catalytic activities.
However, the effect of halide atoms is more subtle (see below).
The unique general tendency is that dihalogenated acetani-
lides are more active than the monohalogenated acetanilides.
As a third general point, the activating power of Sp is
known to be higher than that of CyNMe2, thus leading to
higher conversions under the same conditions. An exception
was found when Sp interacts with the H-bond donor catalyst
(thus inducing lower conversion than with CyNMe2) or when
the H-bond donor catalyst is not active.[5b] Interestingly, tri-
fluorinated 4b–i and trichlorinated 7a–i catalysts (see the Sup-
porting Information) undergo 5–15% conversion irrespective
of which cocatalyst was used. As shown by modeling and the
binding constant, these catalysts can activate the C=O bond of
lactide. Consequently, steric hindrance or electronic factors do
not account for the lack of activation. Here, the acidity of the
trihalogenoacetanilide NH protons is the highest among 1–9
and, accordingly, is proposed to induce a preferential interac-
tion with the cocatalyst, thus significantly reducing its catalytic
properties (see below). In summary, acidic a-trihalogenated
acetanilides poorly activate carbonyl groups in the presence of
a H-bond acceptor cocatalyst.
hibit a large difference in catalytic activity in the presence of
Sp. This point is not yet clearly understood and requires fur-
ther investigations that are out of the scope of the present
study.
Secondly, the H-bonding behavior of a-fluoroacetanilides 2–
4 is dependent on the number of fluorine atoms. The catalytic
properties of monofluorinated compounds 2a–g with Sp (36–
94% conv.) are slightly higher than the corresponding nonha-
logenated compounds 1a–g with Sp (25–56% conv.), which, in
connection with the electron-withdrawing properties of fluo-
rine, better activates the amide bond, despite the less advanta-
geous NH/CF syn conformation. The 2i/Sp system induced the
highest conversion in the series (100% conv.) but 2i/CyNMe2
was far slower (40% conv.) under the same conditions.
Concerning a-difluoroacetanilides 3, their catalytic proper-
ties follow the expected electronic effects: by increasing the
electron-withdrawing power of R1 (CHF2 vs. CH2F) and R (NO2
vs. CF3) substituents, the percentage of lactide conversion in-
creased. Thus, difluorinated catalysts 3 are more efficient than
monofluorinated 2, and compound 3i is the most forceful H-
bond donor of the series: 100% in the presence of Sp and
79% when associated with CyNMe2.
a-Trifluoroacetanilides
4 were inefficient irrespective of
which cocatalyst was used (10–29% conv.). As discussed in the
general comments, an extra H-bond between the donor and
acceptor catalysts is probably responsible for these low con-
versions. This hypothesis is corroborated by the fact that the
less acidic compound 4a (R=H), provided with a phenyl group,
is the best activator (29% conv.) of the series (10–15% for
4b–i).
Catalysis with a-chloro- and a-bromoacetanilides
Concerning the monochlorinated 5a–h/Sp systems, the con-
versions range from 58 to 100% (except for ortho-substituted
compounds), whereas monofluorinated 2a–h/Sp trigger 36–
59% conversion. Monobrominated 8a–h/Sp is able to convert
73–100% lactide. With similar derivatives, the following order
of efficiency was found: CH2F<CH2Cl<CH2Br. The same com-
parisons in efficiency can be made with the dihalogenated sys-
tems: difluorinated 3a–h/Sp (62–96% conv.)<dichlorinated
6a–h/Sp (88–100% conv.)ꢀdibrominated 9a–h/Sp (84–100%
conv.). Again, the catalysts provided with the strongest elec-
tron-withdrawing groups (m,m’-CF3, m,m’-NO2) undergo the
best conversions: 100% with 6i or 9i irrespective of the coca-
talyst. These unprecedented total conversions obtained in the
presence of CyNMe2 are remarkable, because the latter tertiary
amine is recognized as a moderate H-bond acceptor. As pre-
dicted by molecular modeling (Figure 2 f and g), the geometry
and stability of the 6h/lactide and 9h/lactide complexes are in
the same range, revealing comparable catalytic properties. As
already observed for trifluorinated compounds 4, trichlorinated
derivatives 7a–i are ineffective when associated with Sp (5–
15% conv.) or CyNMe2 (10–23% conv.; see the Supporting In-
formation). Again, the electron-withdrawing power of the
COCX3 group on the amide is probably too high, inducing pre-
Catalysis with a-fluoroacetanilides
Firstly, aryl-propionamides 1a–h are moderately active (25–
67% conv.) in the presence of Sp, except when they are pro-
vided with the most powerful electron-withdrawing group
(m,m’-NO2). Compound 1i with Sp undergoes 100% conver-
sion, compared with m,m’-CF3 compound 1e with Sp, which
only leads to 56% conversion. As mentioned above, m-NO2
and p-NO2 substituted 1g (42% conv.) and 1h (67% conv.) ex-
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2855
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ferred interactions with the cocatalyst that hampers the H-
bonding catalysis.
The living-like aspect of the reaction is controlled by running
chain extension experiments with 6i/Sp (see Table S5 in the
Supporting Information). Indeed the catalytic activity of the H-
bonding system is maintained after three successive additions
of monomer (100% conv., 24 h after each loading). Compared
with the most popular catalytic system for ROP of lactide, that
is, thiourea TU/Sp,[2] which allows full conversion in 24 h under
the same conditions, the presented systems based on 3i, 6i
and 9i are at least as efficient. Importantly, the H-bond donor
catalysts 6i and 9i are the first to allow full conversion of lac-
tide in 24 h in the presence of CyNMe2. Interestingly, the use
of CyNMe2 has two advantages compared with classical amine-
based cocatalysts: a) a lower basicity (pKa ꢀ18 in CH3CN) than
Sp (pKa =21.6), 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) (pKa =
24.3), and TBD (pKa =26), and b) a lower nucleophilicity than 4-
Figure 3. Molecular modeling at the B3LYP/6-31+G** level of 6h/Sp.
less negative than those of 6h/lactide (Eint =À8.66 kcalmolÀ1).
Multiple short distances were measured between the sub-
strates: a double H-bond in which the NH group is pinched be-
tween the two nitrogen atoms of Sp (d(NÀH···N)=2.12 and
2.55 ꢂ) and the acidic aromatic protons are in short contact
with CH groups in proximity to the Sp nitrogen atoms (d(CArÀ
H···HÀCN)=2.24, 2.64 and 2.70 ꢂ). However, to better fit its
host, dichloroacetanilide 6h adopts an unusual conformation
in which the amide group is not coplanar with the aromatic
ring (<(O)CNCArCA > =438) and C=O and NÀH bonds are in
syn conformation. The calculated energy of deformation for 6h
to fit Sp is +9.13 kcalmolÀ1 compared with +1–2 kcalmolÀ1
for the other complexes (see the Supporting Information).
Thus, despite a strong energy of interaction, this complex
would not be favored in solution due to the high cost in
energy required to move from the stable trans C(O)ÀNH bond
coplanar with the aromatic ring, which is observed in the solid
and calculated for all the previous complexes. Catalyst 6h
therefore appeared to interact poorly with the Sp cocatalyst.
This assumption can be reasonably extended to the other di-
halogenated catalysts. Concerning the weaker H-bond donor
CyNMe2, some acid-base interaction between catalysts are re-
vealed but to a lesser extent.
(N,N-dimethylamino)pyridine
(DMAP),
1,8-diazabicyclo-
[5.4.0]undec-7-ene (DBU), and 1,4-diazabicyclo[2.2.2]octane
(DABCO). Accordingly, a-dihalogenoacetanilides with m,m’-NO2
substituents opens new perspectives on organocatalyzed reac-
tions relying on CyNMe2 as a cocatalyst.
Finally, the conversion obtained in the presence of CyNMe2
is informative regarding comparative catalytic activity as most
of activate catalysts led to total conversion in the presence of
Sp. Except for ortho-substituted compounds and trihalogenat-
ed acetanilides, a classification depending on aromatic sub-
stituents for the H-bonding catalysts in this ROP reaction with
CyNMe2 emerges: i) for m,m’-CF3 dihalogenated compounds,
3e<6e<9e (46, 66, 85% conv.); ii) for m,m’-NO2 monohalo-
genated catalysts, 2i<5i<8i (40, 82, 90% conv.), and iii) for
m,m’-NO2 dihalogenated compounds, 3i<6i=9i (79, 100,
100% conv.). The impact of aromatic and halide substituents
on the conversion is strong. The most important information
for the powerful a-dihalogenoacetanilides/cocatalyst systems
in the ROP of lactide is the following classification: CHF2 <
CHCl2 <CHBr2.
Based on solid-state and molecular modeling, we propose
that the H-bonding mode of a-halogenoacetanilides towards
the C=O bond of lactide depends on the nature (and the
number) of halide substituents (Scheme 4): i) a-mono- and a-
di-fluoroacetanilides provide a preferred H-bond to lactide
through the amide group and the acidic CArÀH bond, and in-
teract with the cocatalyst, in a larger propotion than less acidic
chlorinated and brominated catalysts, and ii) a-dichloroacetani-
lides and a-dibromoacetanilides display a remarkable double
H-bond to a C=O group through their activated and available
NH and X2CH groups, whereas their acid-base interaction with
the cocatalyst is weak. Therefore, as a result of preferential
conformations and appropriate acidities, adjustable a-dichloro-
and a-dibromoacetanilides are efficient H-bonding catalysts
that act through double H-bonds involving a nonconventional
XnCÀH···O interaction, even in the presence of tertiary amines
as cocatalysts.
H-Bonding and lactide activation
To rationalize the halide dependent conversions observed in
the ROP of lactide (CHF2 <CHCl2 <CHBr2), two hypotheses are
proposed: i) due to their increasing acidity (CHF2 >CHCl2 >
CHBr2),[13] the catalysts interact with both the substrate and the
basic cocatalysts with different strengths, or ii) the interaction
between catalysts is minor and the scale of reactivity is linked
to another phenomenon, such as different modes of H-bond-
ing, as suggested by solid-state packing and modeling.
Indeed, all the acetanilides interact with the cocatalysts with
differing strengths depending on their acidity. For instance, the
most acidic trifluorinated 4 and trichlorinated 7 compounds
were shown to be inactive in C=O activation due to strong in-
teraction with the cocatalysts. To evaluate the dihalogenated
structures, we modeled the complexes between one of the
more efficient catalysts (i.e., 6h) and the basic cocatalyst Sp
(Figure 3). It was found that 6h/Sp forms a strong complex
with an interaction energy of À8.28 kcalmolÀ1, which is slightly
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2856
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Procedure A:
A
mixture of the corresponding acyl chloride
(4.0 mmol), potassium carbonate (4.0 mmol), corresponding aniline
derivative (2.0 mmol), and CH2Cl2 (10 mL) was stirred at RT for 2–
12 h under nitrogen, then poured into water (30 mL) and extracted
with CH2Cl2 (2ꢄ10 mL). The combined organic layers were succes-
sively washed with aqueous saturated NaHCO3 (2ꢄ10 mL), water
(2ꢄ10 mL), dried over MgSO4, filtered, and evaporated to dryness.
The residue was washed with pentane and dried under vacuum.
Procedure B: A mixture of the corresponding acyl anhydride
(5.5 mmol), corresponding aniline derivative (5.0 mmol), and CH2Cl2
(10 mL) was stirred at RT for 1–2 h under nitrogen, then poured
into water (30 mL) and extracted with CH2Cl2 (2ꢄ10 mL). The com-
bined organic layers were successively washed with aqueous satu-
rated NaHCO3 (2ꢄ10 mL), water (2ꢄ10 mL), dried over MgSO4, fil-
tered, and evaporated to dryness. The residue was washed with
pentane and dried under vacuum.
Polymerization reactions
Under nitrogen, in a dry Schlenk tube, were successively intro-
duced organocatalyst (35 mmol), initiator (4-biphenylmethanol,
35 mmol), lactide (700 mmol), 4 ꢂ molecular sieves (five beads), an-
hydrous dichloromethane (1 mL), and cocatalyst (CyNMe2 or (À)-
sparteine, 35 mmol). The reaction mixture was stirred at 208C
under nitrogen for 24 h, then filtered and concentrated in vacuo.
Scheme 4. Two different binding modes of a-halogenoacetanilides depend-
ing on the nature of the halide.
Conclusion
1
Conversion was determined by H NMR spectroscopic analysis, by
For the first time, a-halogenoacetanilides have been shown to
be efficient activators of carbonyl bonds. Electron-withdrawing
properties of the substituents and their number are found to
be crucial. In the ROP of lactide, organocatalysts provided with
electron-deficient aromatic (m,m’-NO2) substituents induce the
highest conversion, irrespective of which H-bond acceptor co-
catalyst was used (Sp or CyNMe2). The effect of the halide sub-
stituents is more subtle; due their differing electron-withdraw-
ing nature and their impact on the acidity of compounds, the
catalytic properties increase in the following order: F<Cl<Br.
Importantly, XnCH halogenomethyl groups (n=1–2; X=Br, Cl)
are shown to be modular and complementary binding groups
to NH. For a-dichloro and a-dibromoacetanilides, we have
demonstrated the utility of nonconventional XnCÀH···O interac-
tions in the activation of C=O bonds even in the presence of
tertiary amines (Sp and CyNMe2). Halogenated acetanilides are
under investigation as organocatalysts in other reactions.
integrating the signals of the methine proton in the residual lac-
tide and the polymer. Polymer molar masses and the dispersity
index were measured by size-exclusion chromatography (SEC) with
a PL-GPC50 Plus apparatus equipped with RI and UV detectors and
Tosoh G4000HXL, G3000HXL and G2000HXL columns (eluent: THF;
flow rate 1.0 mLminÀ1; temperature: 408C; calibrated with poly-
styrene standards). MALDI-TOF analysis was conducted on the
same samples.
X-ray crystallography
Single-crystals of 3h, 3i, 6d, 6i, and 9i were grown by slow diffu-
sion of pentane in a solution of dichloromethane, containing each
catalyst. Detailed crystal structures, cell parameters, and R values
are reported in the Supporting Information. CCDC-944150 (3h),
944151 (3i), 944147 (6d), 944148 (6i), and 944149 (9i) contain the
supplementary crystallographic data for this paper. These data can
be obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Experimental Section
Materials
Molecular modeling
dl-Lactide was recrystallized three times in toluene and freshly
sublimed. Dichloromethane was dried over calcium hydride and
distilled. 4-Biphenylmethanol was purified by precipitation in pen-
tane. CyNMe2 was dried over CaH2 and distilled under argon
before use. Commercially available (À)-sparteine was used as re-
ceived.
The structures of all compounds were optimized at the B3LYP
level[31] in conjunction with the 6–31+G** basis set,[32] and the
lanl2dz potential[33] was used for the bromide atoms. Vibrational
frequency calculations were then performed at the same level of
calculation by using the standard approximations: rigid rotator and
harmonic approximation and we checked that all the frequencies
were positive, confirming the fact that theses structures are
minima of the potential energy surface. All the calculations were
performed by using Gaussian09.[34] Interaction energies were also
computed by using the counterpoise method and the values indi-
cated in the text are all corrected from BBSE (basis set superposi-
tion error).[35]
Synthesis of H-bonding catalysts
Depending upon the commercial availability of the reactants, two
procedures of condensation were undertaken. All reactions were
performed under an inert atmosphere.
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2857
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[9] a) I. Washington, K. N. Houk, Angew. Chem. 2001, 113, 4617–4620;
Angew. Chem. Int. Ed. 2001, 40, 4485–4488; b) O. Takahashi, K. Yasuna-
ga, Y. Gondoh, Y. Kohno, K. Saito, M. Nishio, Bull. Chem. Soc. Jpn. 2002,
75, 1777–1783; c) A. Hasegawa, Y. Naganawa, M. Fushimi, K. Ishihara, H.
Yamamoto, Org. Lett. 2006, 8, 3175–3178; d) E. J. Corey, T. W. Lee,
Chem. Commun. 2001, 1321–1329; e) B. W. Gung, X. Xue, W. R. Roush, J.
Am. Chem. Soc. 2002, 124, 10692–10697; f) R. S. Paton, J. M. Goodman,
Org. Lett. 2006, 8, 4299–4302; g) H. Yang, M. W. Wong, J. Am. Chem.
Soc. 2013, 135, 5808–5818; h) R. C. Johnston, P. H.-Y. Cheong, Org.
Biomol. Chem. 2013, 11, 5057–5064.
[10] a) Electronegativities of atoms on Pauling scale: H=2.1, Br=2.8, Cl=
3.0, and F=4.0; see: L. Pauling, The Nature of the Chemical Bond, 3rd
ed., Cornell University Press, Ithaca, 1960, pp. 93; b) Electronegativities
relative to H=2.176 (from J. March , Advanced Organic Chemistry, 4th
ed., Wiley, New York, 1992, pp. 15) for R/R1 group: CH3 =2.472,
CH2CH3 =2.482, CH2Cl=2.538, CBr3 =2.561, CHCl2 =2.602, CCl3 =2.666,
CF3 =2.985, NO2 =3.421.
Acknowledgements
Financial support was provided by the University of Bordeaux,
the French Ministry of Research (C.T. Fellowship) and the
French National Agency of Research (ANR-07-CP2D-15 Grant,
S.K. Fellowship). The authors are grateful to Mr Julien Morineau
for participation in the project. The CESAMO analytic center
(Claire Mouche) is thanked for mass spectrometry analysis.
Computer facilities were supplied by CECIC Center and
CIMENT (A.M.).
Keywords: hydrogen bonds
interactions · organocatalysis · ring-opening polymerization
·
lactide
·
noncovalent
[11] a) K. L. Dearfield, N. E. McCarroll, A. Protzel, H. F. Stack, M. A. Jackson,
M. D. Waters, Mutat. Res. 1999, 443, 183–221; b) Y. Yang, P. Shang, C.
Cheng, D. Wang, P. Yang, F. Zhang, T. Li, A. Lu, Y. Zhao, Eur. J. Med.
Chem. 2010, 45, 4300–4306.
[12] For a selection, see: a) T. Gunnlaugsson, D. A. Mac Donaill, D. Parker, J.
Am. Chem. Soc. 2001, 123, 12866–12876; b) E. J. Hennessy, S. L. Buch-
wald, J. Am. Chem. Soc. 2003, 125, 12084–12085; c) I. A. Schulte-
Wꢅlwer, J. Helada, R. Gçttlich, Synthesis 2003, 1886–1890; d) A. D.
Moorhouse, A. M. Santos, M. Gunaratnam, M. Moore, S. Neidle, J. E.
Moses, J. Am. Chem. Soc. 2006, 128, 15972–15973; e) E. Feng, H.
Huang, Y. Zhou, D. Ye, H. Jiang, H. Liu, J. Org. Chem. 2009, 74, 2846–
2849.
[1] a) P. R. Schreiner, Chem. Soc. Rev. 2003, 32, 289–296; b) P. M. Pihko,
Angew. Chem. 2004, 116, 2110–2113; Angew. Chem. Int. Ed. 2004, 43,
2062–2064; c) M. S. Taylor, E. N. Jacobsen, Angew. Chem. 2006, 118,
1550–1573; Angew. Chem. Int. Ed. 2006, 45, 1520–1543; d) X. Yu, W.
Wang, Chem. Asian J. 2008, 3, 516–532; e) Z. Zhang, P. R. Schreiner,
Chem. Soc. Rev. 2009, 38, 1187–1198; f) Hydrogen Bonding in Organic
synthesis (Ed.: P. M. Pihko), Wiley-VCH, Weinheim, 2009; g) Y. Takemoto,
Chem. Pharm. Bull. 2010, 58, 593–601; h) K. Etzenbach-Effers, A. Berkes-
sel, Top. Curr. Chem. 2010, 291, 1–27; i) F. Giacalone, M. Gruttadauria, P.
Agrigento, R. Noto, Chem. Soc. Rev. 2012, 41, 2406–2447, and referen-
ces therein.
[2] a) N. E. Kamber, W. Jeong, R. M. Waymouth, R. C. Pratt, B. G. G. Lohmeij-
er, J. L. Hedrick, Chem. Rev. 2007, 107, 5813–5840; b) D. Bourissou, S.
Moebs-Sanchez, B. Martin-Vaca, C. R. Chim. 2007, 10, 775–794; c) M. K.
Kiesewetter, E. J. Shin, J. L. Hedrick, R. M. Waymouth, Macromolecules
2010, 43, 2093–2107; d) A. P. Dove, ACS Macro Lett. 2012, 1, 1409–
1412.
[3] For H-bonding catalysts in multicomponent and sequential reactions,
see: a) C. de Graaff, E. Ruijter, R. V. A. Orru, Chem. Soc. Rev. 2012, 41,
3969–4009; b) C. M. Marson, Chem. Soc. Rev. 2012, 41, 7712–7722.
[4] a) S. Lin, L. Deiana, G.-L. Zhao, J. Sun, A. Cordova, Angew. Chem. 2011,
123, 7766–7772; Angew. Chem. Int. Ed. 2011, 50, 7624–7630; b) B. G. G.
Lohmeijer, R. C. Pratt, F. Leibfarth, J. W. Logan, D. A. Long, A. P. Dove, F.
Nederberg, J. Choi, C. Wade, R. M. Waymouth, J. L. Hedrick, Macromole-
cules 2006, 39, 8574–8583.
[5] a) S. Koeller, J. Kadota, A. Deffieux, F. Peruch, S. Massip, J.-M. Lꢀger, J.-P.
Desvergne, B. Bibal, J. Am. Chem. Soc. 2009, 131, 15088–15089; b) S.
Koeller, J. Kadota, F. Peruch, A. Deffieux, N. Pinaud, I. Pianet, S. Massip,
J.-M. Leger, J.-P. Desvergne, B. Bibal, Chem. Eur. J. 2010, 16, 4196–4205.
[6] For a selection, see: a) Z. S. Derewenda, L. Lee, U. Derewenda, J. Mol.
Biol. 1995, 252, 248–262; b) Y. Gu, T. Kar, S. Scheiner, J. Am. Chem. Soc.
1999, 121, 9411–9422, and references therein; c) R. K. Castellano, Curr.
Org. Chem. 2004, 8, 845–865; d) H. Park, J. Yoon, C. Seok, J. Phys. Chem.
B 2008, 112, 1041–1048; e) C. R. Jones, M. K. N. Qureshi, F. R. Truscott,
S.-T. D. Hsu, A. J. Morrison, M. D. Smith, Angew. Chem. 2008, 120, 7207–
7210; Angew. Chem. Int. Ed. 2008, 47, 7099–7102; f) D. Sheppard, D.-W.
Li, R. Godoy-Ruiz, R. Brꢅschweiler, V. Tugarinov, J. Am. Chem. Soc. 2010,
132, 7709–7719; g) O. Takahashi, Y. Kohno, M. Nishio, Chem. Rev. 2010,
110, 6049–6076; h) J. Hsin, L. M. Lapointe, A. Kazy, C. Chipot, A. Senes,
K. Schulten, J. Am. Chem. Soc. 2011, 133, 14071–14081; i) D.-W. Zhang,
X. Zhao, J.-L. Hou, Z.-T. Li, Chem. Rev. 2012, 112, 5271–5316.
[13] a) K. H. Jensen, M. S. Sigman, Angew. Chem. 2007, 119, 4832–4834;
Angew. Chem. Int. Ed. 2007, 46, 4748–4750; b) K. H. Jensen, M. S.
Sigman, J. Org. Chem. 2010, 75, 7194–7201.
[14] a) J. W. Banks, A. S. Batsanov, J. A. K. Howard, D. O’Hagan, H. S. Rzepa, S.
Martin-Santamaria, J. Chem. Soc. Perkin Trans. 2 1999, 2409–2411;
b) C. R. S. Briggs, D. O’Hagan, J. A. K. Howard, D. S. Yufit, J. Fluorine
Chem. 2003, 119, 9–13; c) V. Peddie, R. J. Butcher, W. T. Robinson,
M. C. J. Wilce, D. A. K. Traore, A. D. Abell, Chem. Eur. J. 2012, 18, 6655–
6662; d) L. E. Zimmer, C. Sparr, R. Gilmour, Angew. Chem. 2011, 123,
12062–12074; Angew. Chem. Int. Ed. 2011, 50, 11860–11871.
[15] C. R. Jones, P. K. Baruah, A. L. Thompson, S. Scheiner, M. D. Smith, J. Am.
Chem. Soc. 2012, 134, 12064–12071.
[16] Synthetic procedures adapted from: a) V. K. Chintakunta, V. Akella, M. S.
Vedula, P. K. Mamnoor, P. Mishra, S. R. Casturi, A. Vangoori, R. Rajagopa-
lan, Eur. J. Med. Chem. 2002, 37, 339–347; b) S. Darvesh, R. S. McDonald,
K. V. Darvesh, D. Mataija, S. Mothana, H. Cook, K. M. Carneiro, N. Richard,
R. Walsh, E. Martin, Bioorg. Med. Chem. 2006, 14, 4586–4599.
[17] a) C. Thomas, F. Peruch, A. Deffieux, A. Milet, J.-P. Desvergne, B. Bibal,
Adv. Synth. Catal. 2011, 353, 1049–1054; b) C. Thomas, F. Peruch, B.
Bibal, RSC Adv. 2012, 2, 12851–12856.
[18] B. T. Gowda, J. Kozisek, M. Tokarcik, H. Fuess, Acta Crystallogr. Sect. E
2008, 64, o987.
[19] a) Y.-H. Wen, X.-M. Li, L.-L. Xu, X.-F. Tang, S.-S. Zhang, Acta Crystallogr.
Sect. E 2006, 62, o4427–o4428; b) B. T. Gowda, S. Foro, H. Fuess, Acta
Crystallogr. Sect. E 2007, 63, o2335.
[20] a) A. K. Pal, A. K. Bera, A. Banerjee, Z. Kristallogr. New Cryst. Struct. 1998,
213, 249–250; b) B. T. Gowda, H. Paulus, H. Fuess, Z. Naturforsch., A:
Phys. Sci. 2001, 56, 386–394.
[21] A. De, A. K. Basak, S. Chaudhuri, G. Poddar, Acta Crystallogr. Sect. C
1984, 40, 2104–2106.
[22] S. Dou, B. T. Gowda, H. Paulus, A. Weiss, Z. Naturforsch., A: Phys. Sci.
1994, 49, 1136–1144.
[23] B. T. Gowda, H. Paulus, H. Fuess, Z. Naturforsch., A: Phys. Sci. 2000, 55,
711–720.
[24] Tests with the M06–2X functional and the same basis set were per-
formed. The optimized structures show the same interacting features,
especially concerning the nonconventional CÀH···O H-bonds (see the
Supporting Information), see: Y. Zhao, D. G. Truhlar, Theor. Chem. Acc.
2008, 120, 215–241.
[7] For quaternary ammonium ions as organocatalysts in asymmetric CÀC
bond formation (N+ÀCH···O interaction), see: a) C. E. Cannizzaro, K. N.
Houk, J. Am. Chem. Soc. 2002, 124, 7163–7169; b) T. Ooi, K. Maruoka,
Acc. Chem. Res. 2004, 37, 526–533; c) J. L. Vicario, D. Badia, L. Carrillo,
Synthesis 2007, 2065–2092; d) T. Tozawa, H. Nagao, Y. Yamane, T. Mu-
kaiyama, Chem. Asian J. 2007, 2, 123–134; e) Y. Park, E. Park, H. Jung, Y.-
J. Lee, S.-S. Jew, H.-G. Park, Tetrahedron 2011, 67, 1166–1170, and refer-
ences therein; f) S. Tarꢆ, R. Chinchilla, C. Nꢇjera, Tetrahedron: Asymmetry
2010, 21, 2872–2878.
[25] a) E. R. Johnson, S. Keinan, P. Mori-Sanchez, J. Contreras-Garcia, A.
Cohen„ W. Yang, J. Am. Chem. Soc. 2010, 132, 6498–6506; b) J. Contre-
[8] C. Thomas, A. Milet, F. Peruch, B. Bibal, Polym. Chem. 2013, 4, 3491–
3498.
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2858
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ras-Garcia, E. R. Johnson, S. Keinan, R. Chaudret, J.-P. Piquemal, D. N. Be-
ratan, W. Yang, J. Chem. Theory Comput. 2011, 7, 625–632.
[26] G. R. Desiraju, T. Steiner, The Weak Hydrogen Bond in Structural Chemistry
and Biology, Oxford University Press Inc., New York, 1999.
[27] K. M. Lippert, K. Hof, D. Gerbig, D. Ley, H. Hausmann, S. Guenther, P. R.
Schreiner, Eur. J. Org. Chem. 2012, 5919–5929.
[28] Calculations conducted on 2h/lactide/CH2Cl2 tertiary complex corrobo-
rated the impact of CÀCl···HÀC secondary interactions between CH2Cl2
and lactide on the complex geometry (see the Supporting Information).
[29] a) L. Skulski, J. Org. Chem. 1963, 28, 3565–3567; b) I. D. Rae, Can. J.
Chem. 1968, 46, 2589–2592; c) J. R. Bartels-Keith, R. F. W. Cieciuch, Can.
J. Chem. 1968, 46, 2593–2600.
[30] X. Li, H. Deng, B. Zhang, J. Li, S. Luo, J.-P. Cheng, Chem. Eur. J. 2010, 16,
450–455. This study on thiourea showed that the position of an aro-
matic substituent controlled the acidity of the NH group (pKA: 4-Cl>3-
Cl>2-Cl) and thus their efficiency in H-bonding.
[34] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A., Jr., Montgomery, J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢈ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[35] S. F. Boys, F. Bernardi, Mol. Phys. 1970, 19, 553–566.
[31] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) C. T. Lee, W. T.
Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
Received: September 17, 2013
[32] P. C. Hariharan, J. A. Pople, Theor. Chim. Acta 1973, 28, 213–222.
[33] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 270–283.
Published online on February 12, 2014
Chem. Eur. J. 2014, 20, 2849 – 2859
www.chemeurj.org
2859
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim