Interplay between steric and electronic factors in determining the strength of intramolecular N-H...O resonance-assisted hydrogen bonds in β-enaminones

Article abstract of DOI:10.1107/S0108768106036421

The crystal structures of five β-enaminones are reported: (2Z)-3-(benzylamino)-1,3-diphenyl-prop-2-en-1-one, (2Z)-3-(benzylamino)-3-(2- hydroxyphenyl)-1-phenyl-prop-2-en-1-one, (2Z)-3-(benzylamino)-3-(4- methoxyphenyl)-1-(3-nitrophenyl)-prop-2-en-1-one, 2

Full text of DOI:10.1107/S0108768106036421

research papers
Acta Crystallographica Section B
Structural
Science
Interplay between steric and electronic factors in
determining the strength of intramolecular N—
Hꢀ ꢀ ꢀO resonance-assisted hydrogen bonds in
b-enaminones
ISSN 0108-7681
Received 10 July 2006
Valerio Bertolasi,* Loretta
Pretto, Valeria Ferretti, Paola
Gilli and Gastone Gilli
The crystal structures of five ꢀ-enaminones are reported:
(2Z)-3-(benzylamino)-1,3-diphenyl-prop-2-en-1-one, (2Z)-3-
Accepted 21 September 2006
(benzylamino)-3-(2-hydroxyphenyl)-1-phenyl-prop-2-en-1-one,
(2Z)-3-(benzylamino)-3-(4-methoxyphenyl)-1-(3-nitrophenyl)-
prop-2-en-1-one, 2-{1-[(4-methoxyphenyl)amino]ethylidene}-
cyclohexene-1,3-dione and 2-{1-[(3-methoxyphenyl)amino]-
ethylidene}cyclohexene-1,3-dione. The structures were
analysed and compared with those of similar compounds in
order to establish which factors determine the range (2.53–
Dipartimento di Chimica and Centro di Strut-
`
turistica Diffrattometrica, Universita di Ferrara,
Via L. Borsari 46, I-44100 Ferrara, Italy
˚
Correspondence e-mail: m38@unife.it
2.72 A) of Nꢀ ꢀ ꢀO hydrogen-bond distances in intramolecularly
hydrogen-bonded ꢀ-enaminones. It has been shown that,
beyond electronic resonance-assisted hydrogen-bond effects
modulated by substituents, the necessary requirements to
produce very short N—Hꢀ ꢀ ꢀO hydrogen bonding are steric
intramolecular repulsions, including the embedding of an
enaminonic C—C or C—N bond in an aliphatic six-membered
ring. By considering the structural features it is possible to
expect the strength of N—Hꢀ ꢀ ꢀO hydrogen bonds adopted by
specific ꢀ-enaminones.
1. Introduction
ꢀ-Enaminones, ꢀ ꢀ ꢀO C—C C—NHꢀ ꢀ ꢀ, are well known for
their ability to form intramolecular resonance-assisted
hydrogen bonds (RAHBs; Gilli et al., 2000; Gilli & Gilli, 2000;
´
Cindric et al., 2005), which are strictly similar to those formed
by the homologous series of ꢀ ꢀ ꢀO C—C N—NHꢀ ꢀ ꢀ keto-
hydrazones (Bertolasi et al., 1993, 1994, 1999; Gilli et al., 2002,
2005, 2006). Their most attractive feature is that they can
display ꢀ ꢀ ꢀO C—C C—NHꢀ ꢀ ꢀ $ ꢀ ꢀ ꢀHO—C C—C
Nꢀ ꢀ ꢀ enaminone–iminoenol tautomerism and can then form,
according to their substituents, both N—Hꢀ ꢀ ꢀO and O—
Hꢀ ꢀ ꢀN RAHBs. In a previous detailed analysis (Gilli et al.,
2000) it has been shown that ꢀ-enaminones can be classified in
three different groups (Scheme 1).
(i) Group A: ꢀ-Enaminones (also called simple ꢀ-enam-
inones). They can form only N—Hꢀ ꢀ ꢀO hydrogen bonds which
are significantly weaker than the corresponding homonuclear
O—Hꢀ ꢀ ꢀO bonds in ꢀ-diketone enols ꢀ ꢀ ꢀO C—C C—
˚
OHꢀ ꢀ ꢀ [2.39 ꢁ d(Oꢀ ꢀ ꢀO) ꢁ 2.44 A] (Emsley, 1984; Gilli et al.,
1989; Gilli & Bertolasi, 1990; Bertolasi et al., 1991), a fact that
has been accounted for by the much higher proton affinity
(PA) of nitrogen with respect to oxygen, which substantially
hinders any valence bond N—Hꢀ ꢀ ꢀO $ Nꢀ ꢀ ꢀH—O resonance
mixing. Observed Nꢀ ꢀ ꢀO distances are found to be scattered in
˚
the range 2.53–2.72 A and to display strong dependence on
# 2006 International Union of Crystallography
the chemical nature of the substituents.
Printed in Great Britain – all rights reserved
1112 doi:10.1107/S0108768106036421
Acta Cryst. (2006). B62, 1112–1120

research papers
Scheme 1. Defining ÁE = E(O—Hꢀ ꢀ ꢀN) ꢂ E(N—Hꢀ ꢀ ꢀO) as
the energy difference between the two hydrogen-bond
minima, the calculated values of ÁE and d(Oꢀ ꢀ ꢀN) are
ꢂ1
33.49 kJ molꢂ1 and 2.713 A for A, ꢂ14.95 kJ mol
and
˚
ꢂ1
˚
˚
2.614 A for B, and 1.88 kJ mol and 2.571 A for C (Gilli et al.,
2000).
Since the three groups have been studied in uneven detail,
we have undertaken a series of investigations on the less
studied groups A and C, generally aimed at establishing the
role played by the steric and electronic properties of the
substituents in determining both the strength of the intra-
molecular RAHBs formed and the shape of the proton-
transfer energy profile. In this first paper, these criteria are
applied to group A compounds carrying different substituents
and, for this purpose, we report here the crystal structures of
three simple ꢀ-enaminones, (1) (2Z)-3-(benzylamino)-1,3-
diphenyl-prop-2-en-1-one, (2) (2Z)-3-(benzylamino)-3-(2-
hydroxyphenyl)-1-phenyl-prop-2-en-1-one and (3) (2Z)-3-
(benzylamino)-3-(4-methoxyphenyl)-1-(3-nitrophenyl)-prop-
2-en-1-one, which are derived from diaryl ꢀ-diketones, and
two ꢀ,ꢀ0-diketo enamines, (4) 2-{1-[(4-methoxyphenyl)-
amino]ethylidene}cyclohexene-1,3-dione and (5) 2-{1-
[(3-methoxyphenyl)amino]ethylidene}cyclohexene-1,3-dione,
which are derived from 2-acetyl-1,3-cyclohexanedione
(Scheme 2). Experimental geometries will be compared with
those of other similar compounds endowed with a variety of
substituents retrieved from the Cambridge Structural Data-
base (CSD; Allen et al., 1979).
(ii) Group B: ꢀ-Enaminones characterized by the fusion of
the RAHB ring with a phenyl group. A number of chemical
variations have been investigated, such as N-salicylidene-
anilines (Inabe, 1991; Wozniak et al., 1995; Ogawa et al., 1998;
Burgess et al., 1999; Fukuda et al., 2003; Arod et al., 2005) and
2-iminophenols (Filarowski et al., 2002, 2005; Filarowski,
2005). Although few compounds are present in both tauto-
meric forms, these molecules prefer to form rather short O—
Hꢀ ꢀ ꢀN (instead of N—Hꢀ ꢀ ꢀO) hydrogen bonds [2.46 ꢁ
˚
d(Nꢀ ꢀ ꢀO) ꢁ 2.57 A] whose formation is to be imputed (Gilli et
al., 2000) to the large resonant energy of the phenyl ring,
which stabilizes the iminoenol (2-iminophenol) form against
the N/O PA difference; this energy would be lost in forming
the enaminone tautomer.
(iii) Group C: Probably the most interesting enaminones.
Here the RAHB ring is fused with a naphthalene group and,
because of the smaller resonance energy of this molecule, the
two N—Hꢀ ꢀ ꢀO and Nꢀ ꢀ ꢀH—O tautomers become nearly
isoenergetic, so as to be easily shifted from one state to the
other by the effect of properly chosen N-substituents. This
situation results in the formation of tautomeric N—Hꢀ ꢀ ꢀO/
Nꢀ ꢀ ꢀH—O bonds in a double-well potential with substituent-
modulated proton populations. Accordingly, their crystal
structures become dynamically or statically disordered
according to the lower or higher value of the proton transfer
energy barrier (Inabe et al., 1994; Kaitner et al., 1998; Nazir et
2. Experimental
2.1. Synthesis
Compounds (1), (2) and (3) were synthesized by the
condensation reaction, in a benzene solution, of a suitable 1,3-
diaryl ꢀ-diketone derivative with benzylamine. The reaction
mixture was refluxed and stirred for some hours. The slurry
was reduced in volume on a rotatory evaporator and allowed
´
al., 2000; Popovic et al., 2001).
The classification into three groups is strongly supported by
the density functional theory calculations carried out at the
B3LYP/6–31+G(d,p) level on the three sample molecules of
ꢃ
Acta Cryst. (2006). B62, 1112–1120
Valerio Bertolasi et al.
RAHBs in ꢀ-enaminones 1113

research papers
Table 1
Experimental details.
(1)
(2)
(3)
(4)
(5)
Crystal data
Chemical formula
Mr
C22H19NO
313.38
Triclinic, P1
C22H19NO2
329.38
Orthorhombic, Pbca
C23H20N2O4
388.41
ꢀ
Triclinic, P1
C15H17NO3
259.30
ꢀ
Triclinic, P1
C15H17NO3
259.30
Monoclinic, P21/n
ꢀ
Cell setting, space
group
Temperature (K)
150
8.9727 (2), 9.7552 (2),
10.0208 (2)
150
150
6.9599 (2), 8.4786 (3),
17.1616 (6)
150
7.7200 (2), 8.9566 (3),
9.5777 (3)
150
14.0608 (3), 6.6407 (1),
14.4581 (3)
˚
a, b, c (A)
14.2380 (3),
14.3109 (3),
16.8047 (3)
90.00, 90.00, 90.00
ꢁ, ꢀ, ꢃ (ꢄ)
89.081 (1), 71.002 (1),
85.786 (1)
827.08 (3)
2
1.258
Mo Kꢁ
0.08
82.282 (2), 78.773 (2),
75.599 (2)
958.18 (6)
2
1.346
Mo Kꢁ
0.09
93.830 (2), 92.083 (2),
105.755 (2)
634.92 (3)
2
1.356
Mo Kꢁ
0.10
90.00, 107.312 (1), 90.00
3
˚
V (A )
3424.1 (1)
8
1.278
Mo Kꢁ
0.08
1288.85 (4)
4
1.336
Mo Kꢁ
0.09
Z
Dx (Mg mꢂ3
Radiation type
)
ꢄ (mmꢂ1
)
Crystal form, colour
Crystal size (mm)
Prism, colourless
0.44 ꢅ 0.24 ꢅ 0.21
Prism, colourless
0.32 ꢅ 0.31 ꢅ 0.25
Plate, colourless
0.35 ꢅ 0.33 ꢅ 0.09
Prism, colourless
0.50 ꢅ 0.38 ꢅ 0.17
Irregular, colourless
0.43 ꢅ 0.32 ꢅ 0.24
Data collection
Diffractometer
Data collection method
Absorption correction
No. of measured,
independent and
observed reflections
Criterion for observed
reflections
Nonius Kappa CCD
’ and ! scans
None
Nonius Kappa CCD
’ and ! scans
None
Nonius Kappa CCD
’ and ! scans
None
Nonius Kappa CCD
’ and ! scans
None
Nonius Kappa CCD
’ and ! scans
None
8302, 4703, 3866
10 220, 4138, 3631
8934, 4331, 3563
5780, 2862, 2554
11 215, 3749, 3285
I > 2ꢅ(I)
I > 2ꢅ(I)
I > 2ꢅ(I)
I > 2ꢅ(I)
I > 2ꢅ(I)
Rint
ꢆmax (ꢄ)
0.020
30.0
0.024
28.0
0.031
27.5
0.028
27.6
0.028
30.0
Refinement
Refinement on
R[F2 > 2ꢅ(F2)],
wR(F2), S
F2
0.049, 0.120, 1.06
F2
0.056, 0.132, 1.35
F2
0.056, 0.123, 1.19
F2
0.052, 0.145, 1.08
F2
0.053, 0.123, 1.15
No. of reflections
No. of parameters
H-atom treatment
4703
293
4138
303
4331
342
2862
232
3749
240
Refined independently
Mixture of independent
and constrained
refinement
Refined independently
Mixture of independent
and constrained
refinement
Refined independently
Weighting scheme
w = 1/[ꢅ2(Fo2) +
(0.0465P)2 +
0.3404P], where P =
(Fo2 + 2Fc2)/3
<0.0001
w = 1/[ꢅ2(Fo2) +
(0.0297P)2 +
2.1857P], where P =
(Fo2 + 2Fc2)/3
<0.0001
w = 1/[ꢅ2(Fo2) +
w = 1/[ꢅ2(Fo2) +
(0.0714P)2 +
0.3622P], where P =
(Fo2 + 2Fc2)/3
0.001
w = 1/[ꢅ2(Fo2) +
(0.0361P)2 +
0.7763P], where P =
(Fo2 + 2Fc2)/3
<0.0001
(0.033P)2 + 0.5384P],
where P = (Fo2
2Fc2)/3
+
(Á/ꢅ)max
Áꢇmax, Áꢇmin (e A
<0.0001
0.21, ꢂ0.27
ꢂ3
˚
)
0.24, ꢂ0.30
0.29, ꢂ0.25
0.46, ꢂ0.47
0.31, ꢂ0.22
Computer programs used: Kappa CCD Server Software (Nonius, 1997), DENZO–SMN (Otwinowski & Minor, 1997), SIR97 (Altomare et al., 1999), SHELXL97 (Sheldrick, 1997),
ORTEPIII (Burnett & Johnson, 1996), PARST (Nardelli, 1995), PLATON (Spek, 2003).
to stand overnight. The colourless products were isolated by
filtration and recrystallized from a mixture of methanol/ethyl
acetate.
The procedure for the synthesis of compounds (4) and (5)
was analogous, using, in a toluene solution, 2-acetyl-1,3-
cyclohexanedione and 4-methoxyaniline or 3-methoxyaniline.
The products were recrystallized from ethanol.
polarization effects. The structures were solved by direct
methods (Altomare et al., 1999) and refined using full-matrix
least squares. All non-H atoms were refined anisotropically
and the H atoms isotropically. In structure (2) the ortho
hydroxy group of the phenyl ring in position 3 is disordered
over both ortho, ortho0 positions. This O atom was refined over
both positions with partial occupancies of 0.8 and 0.2,
respectively. The O—H and C—H H atoms involved in this
disorder were included in calculated positions, riding on their
attached atoms. In structure (4) the C5(H2) group of the
cyclohexanedione ring was found to be disordered and was
refined over two positions with occupancies of 0.61 and 0.39,
respectively. Both the H atoms were included in calculated
positions, riding on their attached atoms. All calculations were
performed using SHELXL97 (Sheldrick, 1997) and PARST
2.2. Crystal structure determination
The diffraction data for all compounds were collected at
150 K using a Nonius KappaCCD diffractometer with
˚
graphite-monochromated Mo Kꢁ radiation (ꢂ = 0.7107 A).
Data sets were integrated using the DENZO–SMN package
(Otwinowski & Minor, 1997) and corrected for Lorentz and
ꢃ
1114 Valerio Bertolasi et al.
RAHBs in ꢀ-enaminones
Acta Cryst. (2006). B62, 1112–1120

research papers
Table 2
Selected bond distances (A), bond numbers (in italic) and angles ( ).
Table 3
Structural and spectroscopic parameters of hydrogen bonds.
ꢄ
˚
(1)
(2)
(3)
(4)
(5)
D—H Hꢀ ꢀ ꢀA Dꢀ ꢀ ꢀA
D—Hꢀ ꢀ ꢀA ꢊNH
ꢉNH
(ꢄ) (p.p.m.) (cmꢂ1
)
˚
(A)
˚
(A)
˚
(A)
O1—C1
C1—C2
C2—C3
N1—C3
1.254 (1)
1.62
1.426 (2)
1.32
1.386 (2)
1.57
1.337 (2)
1.44
0.39
1.272 (1)
1.52
1.412 (2)
1.40
1.388 (2)
1.56
1.330 (2)
1.39
0.43
1.256 (2)
1.61
1.418 (3)
1.37
1.388 (2)
1.56
1.338 (3)
1.43
0.41
1.252 (2)
1.64
1.445 (2)
1.21
1.429 (2)
1.30
1.328 (2)
1.50
0.44
1.248 (2)
1.66
1.446 (2)
1.21
1.424 (2)
1.35
1.328 (2)
1.50
0.42
(1)
N1—H1ꢀ ꢀ ꢀO1
0.92 (2) 1.94 (2) 2.693 (1) 137 (1)
11.56
3175
(2)
N1—H1ꢀ ꢀ ꢀO1
0.94 (2) 1.95 (2) 2.694 (2) 134 (2)
11.73
–
–
3288
–
–
O2—H20ꢀ ꢀ ꢀO1i 0.82†
O20—H200ꢀ ꢀ ꢀO1ii 0.82†
1.86
2.07
2.670 (2) 168
2.730 (8) 138
ꢂ
Del%
78
86
82
88
84
(3)
N1—H1ꢀ ꢀ ꢀO1
C1—C10
C3—C4
N1—C16
C2—C7
O2—C7
N1—C9
1.506 (1)
1.492 (2)
1.454 (1)
1.496 (2)
1.493 (2)
1.455 (2)
1.503 (2)
1.492 (3)
1.458 (2)
0.93 (3) 1.88 (2) 2.650 (2) 139 (2)
0.94 (2) 1.73 (2) 2.544 (2) 143 (2)
0.90 (2) 1.79 (2) 2.554 (2) 141 (2)
11.87
14.87
15.03
3261
2862
2880
(4)
N1—H1ꢀ ꢀ ꢀO1
1.465 (2)
1.230 (2)
1.433 (2)
1.465 (2)
1.229 (2)
1.430 (2)
(5)
N1—H1ꢀ ꢀ ꢀO1
O1—C1—C2
C1—C2—C3
N1—C3—C2
Æꢁ
123.1 (1)
124.0 (1)
122.2 (1)
369.3
122.2 (2)
124.4 (2)
122.8 (2)
369.4
123.2 (2)
122.9 (2)
122.1 (2)
368.2
123.1 (1)
120.1 (1)
119.0 (1)
362.2
123.0 (1)
120.3 (1)
119.5 (1)
362.8
Symmetry codes: (i) 1 ꢂ x; 21 þ y; 23 ꢂ z; (ii) x ꢂ 21 ; y; 23 ꢂ z. † Calculated H atoms.
C3—N1—C16
C3—N1—C9
126.8 (1)
125.5 (2)
126.5 (2)
127.4 (2)
127.2 (1)
3. Results and discussion
3.1. Compounds (1)–(3)
(Nardelli, 1995) as implemented in the WinGX (Farrugia,
1999) system of programs. The crystal data and refinement
parameters are summarized in Table 1.1
A selection of bond distances and angles is given in Table 2,
together with bond orders (in italic) and ꢈ-delocalization
parameters, ꢂ and Del%, of the heterodienic group.
1
Hydrogen-bond formation is associated with large H NMR
chemical shifts in the ranges 11.56ꢂ11.87 p.p.m. for (1)–(3)
and 14.82–15.03 p.p.m. for (4) and (5) (to be compared with
the 7–9 p.p.m. observed for weak hydrogen bonds) and with
red-shifted ꢉNH stretching frequencies from 3175 to 3288 cmꢂ1
for (1)–(3) and from 2862 to 2880 cmꢂ1 for (4) and (5) (to be
compared with 3400 cmꢂ1 for the free N—H bond).
Table 3 reports the hydrogen-bond parameters. ORTEPIII
(Burnett & Johnson, 1996) views are shown in Figs. 1–5. The
structural data retrieved from CSD files are given in Supple-
mentary Tables S1–S6.
2.3. Hydrogen-bond strength and p-delocalization indices
Resonance-assisted hydrogen bonding can be defined as a
positive synergism between hydrogen-bond strengthening and
ꢈ-delocalization of the interleaving resonant fragment, two
quantities that need to be precisely defined. In this paper,
hydrogen-bond strengths are evaluated through the Nꢀ ꢀ ꢀO
contact distances and the ꢈ-delocalization through the indices
ꢂ (0 ꢁ ꢂ ꢁ 1) and Del% = 100(1 ꢂ |2 ꢂ ꢂ 1|) (0 ꢁ Del% ꢁ 100)
(Gilli et al., 2004; Bertolasi et al., 2006), which have the
following meaning. With reference to the extreme ꢀ ꢀ ꢀHN—
Compounds (1), (2) and (3) are N-alkyl-1,3-diaryl-ꢀ-
enaminones derived from 1,3-diaryl-ꢀ-diketone enols. They
exhibit intramolecular N—Hꢀ ꢀ ꢀO hydrogen bonds with Nꢀ ꢀ ꢀO
C
C—C Oꢀ ꢀ ꢀ (EO = enaminone) and ꢀ ꢀ ꢀN C—C C—
OH (IE = iminoenol) forms, the two extreme geometries are
characterized by the same Del% = 0 and by ꢂ = 0 and 1 for EO
and IE, respectively, while the fully ꢈ-delocalized structure has
Del% = 100 for ꢂ = 0.5. The ꢈ-delocalization parameter, ꢂ, has
been evaluated from the Pauling bond numbers, n, of the
bonds involved as ꢂ = (1/4) [(2 ꢂ n1) + (n2 ꢂ 1) + (2 ꢂ n3) +
(n4 ꢂ 1)], where n1, n2, n3 and n4 are the bond orders of the
C1 O, C1—C2, C2 C3 and C3—N bonds, respectively.
Bond numbers are calculated from the Pauling (1960) equa-
tion d(1) ꢂ d(n) = clog10n, where the pure single (n = 1) and
˚
distances of 2.693 (1), 2.694 (2) and 2.650 (2) A, which are in
˚
agreement with the average value of 2.67 (2) A (Table 4)
calculated from a sample of 46 simple ꢀ-enaminones (Table S1)
retrieved from the CSD and carrying alkyl or aryl substituents
in positions 1 and 3.
In (1), where the 1,3-substituents are both phenyl groups,
the enaminone fragment undergoes an extended delocaliza-
tion (ꢂ = 0.39). In spite of this large delocalization the N—
Hꢀ ꢀ ꢀO bond remains moderately strong with respect to the
˚
˚
double (n = 2) bond distances are 1.49 and 1.33 A, 1.38 and
shortest Nꢀ ꢀ ꢀO distances of up to 2.53 A found in some
2
2
2
˚
˚
1.20 A, and 1.41 and 1.27 A for Csp —Csp , Csp —O (Allen et
al., 1987) and Csp2—Nsp2 (Ferretti et al., 1993) bonds,
respectively.
ꢀ-enaminones carrying peculiar substituents (Gilli et al., 2000).
The internal angles at atoms C1, C2 and C3 are significantly
greater than 120ꢄ (on average 123ꢄ), indicating that the
attractive force of the intramolecular N—Hꢀ ꢀ ꢀO hydrogen
bond is counterbalanced by repulsive interactions inside the
ꢀ-enaminonic cycle (Table 2).
1 Supplementary data for this paper are available from the IUCr electronic
archives (Reference: BK5040). Services for accessing these data are described
at the back of the journal.
ꢃ
Acta Cryst. (2006). B62, 1112–1120
Valerio Bertolasi et al.
RAHBs in ꢀ-enaminones 1115

research papers
In (2), the O2H hydroxy group, on the C4–C9 phenyl ring, is
disordered, with the two opposed ortho positions in the ratio
80:20. Both are in a suitable orientation to form inter-
molecular O2—Hꢀ ꢀ ꢀO1 hydrogen bonds with the same
C1 O1 carbonyl group. These bonds seem responsible for a
further polarization of the C1 O1 bond, which exhibits a
˚
remarkable lengthening [to 1.272 (1) A] with respect to the
˚
same bonds in (1) and (3) [1.254 (1) and 1.256 (2) A, respec-
tively]. As a consequence, the delocalization within the
heterodienic group also slightly increases, up to ꢂ = 0.43, while
the interatomic interactions within the heterodienic ring
Figure 1
ORTEPIII (Burnett & Johnson, 1996) view of (1), showing displacement
ellipsoids at the 40% probability level.
Figure 4
ORTEPIII view of (4), showing displacement ellipsoids at the 40%
probability level.
Figure 2
ORTEPIII view of (2), showing displacement ellipsoids at the 40%
probability level.
Figure 3
ORTEPIII view of (3), showing displacement ellipsoids at the 40%
probability level.
Figure 5
ORTEPIII view of (5), showing displacement ellipsoids at the 40%
probability level.
ꢃ
1116 Valerio Bertolasi et al.
RAHBs in ꢀ-enaminones
Acta Cryst. (2006). B62, 1112–1120

research papers
remain unchanged, the internal angles at atoms C1, C2 and C3
and the Nꢀ ꢀ ꢀO hydrogen-bond distance being comparable to
the values for compound (1).
Scheme 3, where the enaminonic system is differently fused, at
the C1—C2, C2—C3 or C3—N bonds, with a six-membered
aliphatic cycle. Structures where the enamine group is fused
with an aromatic ring, such as salicylideneanilines, have been
excluded for the reasons given in x1; those where at least two
bonds of the heterodienic system are included in six-
membered aliphatic cycles have also been excluded. Average
geometric and delocalization parameters for each class are
given in Table 4.
In (3) the substituents at the 1- and 3-positions are
m-nitrophenyl and p-methoxyphenyl groups, respectively. In
spite of a comparable delocalization of the heterodienic
system (ꢂ = 0.41), this structure shows a small narrowing of the
internal angles at atoms C1, C2 and C3 associated with a
significant shortening of the intramolecular N—Hꢀ ꢀ ꢀO
hydrogen bond. This shortening may be ascribed to an
increased ꢈ-polarizability of the two phenyl rings carrying the
m-NO2 and p-OCH3 substituents, which increases the flux of
ꢈ-electrons inside the RAHB ring. In this respect, it is inter-
esting to note that the corresponding (i.e. equally substituted)
ꢀ-diketone enol displays one of the shortest observed Oꢀ ꢀ ꢀO
˚
distances of 2.434 (1) A (Bertolasi et al., 1991; Gilli et al.,
2004).
3.2. Compounds (4) and (5)
Compounds (4) and (5) show intramolecular N—Hꢀ ꢀ ꢀO
˚
hydrogen-bond distances of 2.544 (2) and 2.554 (2) A, which
are much shorter than those observed in (1)–(3) [on average
˚
2.66 (2) A]. The shortening has been attributed to the elec-
tronic effect produced by the second ꢀ-diketone groups (Gilli
et al., 2000). In addition, in the present structures, this second
ꢀ-enaminonic group, HN1—C3 C2—C7 O2, is seen to
affect the RAHBs through a further delocalization of the C2—
C7 O2 group, which displays a small shortening of the C2—
2
2
˚
C7 (Csp —Csp ) bond (1.46 A) and a lengthening of the
Simple ꢀ-enaminones (Ia) display bond distances within the
heterodienic system corresponding to an almost complete ꢈ-
delocalization, in agreement with what was found for (1)–(3),
and a widening of the three internal angles that sum to an
average value of 369ꢄ. The average Nꢀ ꢀ ꢀO hydrogen-bond
˚
C7 O2 double bond (1.23 A; Table 1). However, other
factors of simple steric nature seem to be important, in
particular (i) the inclusion of the C1—C2 bond in a
cyclohexanedione ring and (ii) the short intramolecular
contact between atom O2 and the methyl group in position 3
(Scheme 2). These structural features are consistent with the
observed narrowing of the enaminone internal angles C1—
C2—C3 and N1—C3—C2, the anomalous lengthening of the
C1—C2 and C2—C3 bonds, and, as a consequence, the
increased approach of the NH group to atom O1. The slightly
different Nꢀ ꢀ ꢀO distances in (3) and (4) can be ascribed to the
different electronic effects produced by the methoxy groups in
the para and meta positions on the phenyl ring linked to the
N1 atom, in agreement with what was previously observed in
the analogous series of diketohydrazones (Bertolasi et al.,
1993).
˚
distance of 2.67 (2) A is not much shorter than the same
˚
intramolecular distance in non-resonant systems [2.72 (2) A;
Gilli et al., 2000; Chisolm et al., 2000], indicating that the
repulsions inside the ꢀ-enaminonic systems efficiently coun-
teract the electronic attractive interactions due to RAHB
effects.
Compounds of subclasses (Ib), (Ic) and (Id) are indicative
of the effects induced by fusion of the C1—C2, C2 C3 or
C3—N bonds with the aliphatic six-membered cycle. In spite
of the fact that few compounds have been retrieved for
subclass (Ib), some general conclusions can be drawn. In (Ia)
the Nꢀ ꢀ ꢀO distance cannot be shortened by a narrowing of the
C1—C2 C3 angle as long as the C2 C3—N angle can
undergo an analogous enlargement, so the sum of the internal
angles remains unchanged, as does the Nꢀ ꢀ ꢀO distance (Table
4). Conversely, in subclasses (Ic) and (Id) the fusion of the
enaminonic C2 C3 or C3—N bonds with an aliphatic six-
membered ring gives rise to an equalization of the C1—
C2 C3 and C2 C3—N internal angles to about 121ꢄ. As a
consequence, the sum of the internal angles decreases to 366ꢄ
3.3. Comparison with CSD-retrieved structures
The relative importance of the factors responsible for the
N—Hꢀ ꢀ ꢀO hydrogen-bond shortening can be appreciated by a
systematic survey of the CSD files of structures containing the
ꢀ-enaminone fragment. According to previous indications
(Gilli et al., 2000), these compounds have been divided into
two classes: (I) (simple ꢀ-enaminones) (Scheme 3) and (II) (ꢀ,
ꢀ0-diketoenamines) (Scheme 4).
˚
while the Nꢀ ꢀ ꢀO distances is shortened to 2.59–2.61 A. Note
that the ꢈ-delocalization indices remain fairly constant in all
compounds of class (I), indicating that the electronic RAHB
3.3.1. Structures of class (I). The structures of class (I) have
been further divided in the four subclasses (Ia)–(Id) of
ꢃ
Acta Cryst. (2006). B62, 1112–1120
Valerio Bertolasi et al.
RAHBs in ꢀ-enaminones 1117

research papers
Table 4
ꢄ
˚
Average structural data (A and ) of ꢀ-enaminones of the classes reported in Schemes 3 and 4 retrieved from CSD files.
Bond numbers are reported in italics. ꢂ and Del% are delocalization parameters (see text); Æꢁ is the sum of the internal angles at the ꢀ-enaminone group.
Subclass
(Ia)
n
hOꢀ ꢀ ꢀNi hO1 C1i hC1—C2i hC2 C3i hC3—Ni
hO1 C1—C2i hC1—C2 C3i hC2 C3—Ni Æꢁ
hꢂi
hDel%i
46 2.67 (2)
1.25 (1)
1.65
1.41 (1)
1.41
1.38 (1)
1.61
1.34 (1)
1.41
123 (1)
124.4 (9)
121.1 (9)
369
369
366
366
366
367
362
367
364
0.39 78
0.34 68
0.36 72
0.39 78
0.38 76
0.40 80
0.48 96
0.39 78
0.40 80
(Ib)
2
5
2.650
1.256
1.61
1.25 (1)
1.65
1.26 (1)
1.59
1.418
1.37
1.432 (5)
1.29
1.42 (1)
1.35
1.362
1.74
1.385 (7)
1.58
1.39 (1)
1.54
1.354
1.32
1.345 (3) 123.7 (3)
1.38
1.35 (1)
1.35
122.4
121.6
125.5
(Ic)
2.59 (2)
120.5 (3)
122.5 (7)
119.8 (4)
122.1 (6)
120.6 (3)
122.6
121.8 (6)
121.0 (8)
126 (1)
123 (2)
118.1 (7)
118.4
(Id)
17 2.61 (2)
17 2.61 (3)
122.7 (6)
120.5 (5)
122 (2)
(IIa)
(IIb)
(IIc)
(IId)
(IIe)
1.240 (5)
1.71
1.450 (9)
1.19
1.390 (9)
1.54
1.32 (1)
1.56
3
2.63 (2)
1.24 (2)
1.71
1.257 (5)
1.61
1.244
1.69
1.240
1.71
1.43 (2)
1.30
1.442 (6)
1.23
1.426
1.32
1.450
1.19
1.40 (1)
1.48
1.428 (9)
1.31
1.409
1.42
1.412
1.40
1.33 (3)
1.49
1.315 (6) 123.6 (4)
1.60
1.350
1.35
14 2.55 (1)
1
2
2.644
2.584
125.9
121.7
1.328
1.50
120.2
122.0
effects are essentially constant. The reason for the Nꢀ ꢀ ꢀO
shortening must be steric in nature and there is, in fact,
evidence that the Nꢀ ꢀ ꢀO distance is strictly related to the sum
of the values of the internal ꢀ-enaminone angles.
3.3.2. Structures of class II. The structures of class II have
been divided in the five subclasses (IIa)–(IIe), as displayed in
Scheme 4. All these compounds are ꢀ,ꢀ0-diketoenamines, that
is ꢀ-enaminones substituted by a further ꢀ0-keto group.
associated with the supplementary ꢈ-delocalization within the
second HN—C3 C2—C4 O2 ꢀ0-enaminonic group,
complemented by the steric effects generated by the intra-
molecular short repulsive interactions between atom O2 and
the Me substituent in position 1, which are made possible by
the planarity of the ꢀ,ꢀ0-diketoenamine group. This can
explain why the internal angles at atoms C1 and C2 undergo a
significant narrowing, while the angle at atom C3, which
carries a simple H atom, can become wider, so keeping the
sum of the internal angles around 366ꢄ. These data are in
agreement with those reported for ꢀ,ꢀ0-diketohydrazones by
Bertolasi et al. (1993, 1994, 1999) and Gawinecki et al. (2001).
Subclasses (IIb) and (IIc) include compounds where the
C1—C2 bond is embedded in an aliphatic six-membered ring,
as in compounds (4) and (5) of this paper. The structures of
subclass (IIb) do not exhibit significant intramolecular short
repulsive interactions because the substituent at C3 is an H
atom and, accordingly, the structures display an average
˚
Nꢀ ꢀ ꢀO distance of 2.63 (2) A, which is significantly shorter
than in simple ꢀ-enaminones (Ia), but slightly longer than in
ꢀ,ꢀ0-diketoenamines (IIa). Therefore, the presence of a cycle,
such as cyclohexanedione, that includes atoms C1 and C2 is
˚
not sufficient to shorten Nꢀ ꢀ ꢀO distances below 2.63 A. To
obtain a greater Nꢀ ꢀ ꢀO shortening, the ꢀ-enaminone group
must be involved in stronger intramolecular steric effects, a
requirement satisfied by the structures of subclass (IIc)
because of the presence of an alkyl or aryl R3 group in posi-
tion 3. Owing to the O2ꢀ ꢀ ꢀR3 steric repulsion the Nꢀ ꢀ ꢀO
˚
distance is shortened, on average, to 2.55 (1) A, while the sum
of the internal angles is decreased to 362ꢄ, which is the
smallest average value reported in Table 4. This very short
hydrogen bond is associated with an almost complete delo-
calization (ꢂ = 0.48, Del% = 96) of the heterodienic unit.
A similar effect can be predicted for subclasses (IId) and
(IIe) where the C2 C3 double bond is included in an
aliphatic six-membered ring. In spite of the fact that very few
The structures of subclass (IIa) exhibit an average Nꢀ ꢀ ꢀO
˚
distance of 2.61 (3) A, which is significantly shorter than that
observed in simple ꢀ-enaminones (Ia). This shortening has
been imputed (Gilli et al., 2000) to the electronic effects
ꢃ
1118 Valerio Bertolasi et al.
RAHBs in ꢀ-enaminones
Acta Cryst. (2006). B62, 1112–1120

research papers
structures were retrieved from the CSD, there is no doubt that
the trend observed here is analogous to that found for
subclasses (IIb) and (IIc). The only structure of subclass (IId),
which displays an H atom as substituent in position 1, exhibits
We thank MIUR (Rome) for COFIN 2004 financial support
as part of our project ‘Smart Hydrogen Bonds in Nature and
Functional Materials’.
˚
the longer Nꢀ ꢀ ꢀO distance of 2.64 A and a sum of internal
angles of 367ꢄ, while the structures of subclass (IIe), because
of the strong intramolecular steric hindrance between O2 and
References
˚
Allen, F. H., Bellard, S., Brice, M. D., Cartwright, B. A., Doubleday,
A., Higgs, H., Hummelink, T., Hummelink-Peters, B. G., Kennard,
O., Motherwell, W. D. S., Rodgers, J. & Watson, D. G. (1979). Acta
Cryst. B35, 2331–2339.
R1, exhibit a shorter Nꢀ ꢀ ꢀO distance of 2.58 A, with a
concomitant narrowing of the enaminonic internal angles,
whose sum is 364ꢄ.
Allen, F. H., Kennard, O., Watson, D. G., Brammer, L., Orpen, A. G.
& Taylor, R. (1987). J. Chem. Soc. Perkin Trans. 2, pp. S1–19.
Altomare, A., Burla, M. C., Camalli, M., Cascarano, G. L.,
Giacovazzo, C., Guagliardi, A., Moliterni, A. G., Polidori, G. &
Spagna, R. (1999). J. Appl. Cryst. 32, 115–116.
Arod, F., Pardon, M., Pattison, P. & Chapuis, G. (2005). Acta Cryst.
C61, o317–o320.
Bertolasi, V., Ferretti, V., Gilli, P., Gilli, G., Issa, Y. M. & Sherif, O. E.
(1993). J. Chem. Soc. Perkin Trans. 2, pp. 2223–2228.
Bertolasi, V., Gilli, P., Ferretti, V. & Gilli, G. (1991). J. Am. Chem. Soc.
113, 4917–4925.
Bertolasi, V., Gilli, P., Ferretti, V., Gilli, G. & Vaughan, K. (1999). New
J. Chem. 23, 1261–1267.
Bertolasi, V., Nanni, L., Gilli, G., Ferretti, V., Gilli, P., Issa, Y. M. &
Sherif, O. E. (1994). New J. Chem. 18, 251–261.
Bertolasi, V., Pretto, L., Gilli, G. & Gilli, P. (2006). Acta Cryst. B62,
850–863.
Burgess, J., Fawcett, J., Russell, D. R., Gilani, S. R. & Palma, V. (1999).
Acta Cryst. C55, 1707–1710.
4. Conclusions
All the reported structural data of intramolecularly hydrogen-
bonded ꢀ-enaminones suggest that the Nꢀ ꢀ ꢀO bond length is
determined by two main factors: the ꢈ-conjugation of the
resonant fragment and the intramolecular steric interactions
involving the substituents at the heterodienic system. Simple
ꢀ-enaminones (Ia) display extended ꢈ-conjugation of the
heterodienic group, but the RAHB effect alone does not
produce a significant shortening of Nꢀ ꢀ ꢀO hydrogen-bond
˚
distances below 2.67 A because of the different proton affi-
nities of the N and O atoms and because of the absence of
strong intramolecular steric compression. Moderate short-
˚
enings down to 2.61 A are, however, produced in subclasses
(Ib)–(Id) by the introduction of aliphatic six-membered rings
fused with the enaminonic moiety, which contribute to stif-
fening the overall structure.
Burnett, M. N. & Johnson, C. K. (1996). ORTEPIII. Report ORNL-
6895. Oak Ridge National Laboratory, Tennessee, USA.
Chisholm, G., Kennedy, A. R., Wilson, S. & Teat, S. J. (2000). Acta
Cryst. B56, 1046–1053.
˚
Significant Nꢀ ꢀ ꢀO shortenings down to 2.55 A can be
´
ˇ
Cindric, M., Novak, T. K. & Uzarevic, K. (2005). J. Mol. Struct. 750,
´
achieved only in compounds (II) having, at the same time, the
ꢀ-enaminone system included in a six-membered aliphatic
ring, a further ꢀ0-keto group substituted in position 2 and
substituents able to produce strong intramolecular repulsive
interactions.
135–141.
Emsley, J. (1984). Struct. Bonding, 57, 147–191.
Farrugia, L. J. (1999). J. Appl. Cryst. 32, 837–838.
Ferretti, V., Bertolasi, V., Gilli, P. & Gilli, G. (1993). J. Phys. Chem. 97,
13568–13574.
Filarowski, A. (2005). J. Phys. Org. Chem. 18, 686–698.
Filarowski, A., Kochel, A., Cieslik, K. & Koll, A. (2005). J. Phys. Org.
Chem. 18, 986–993.
Filarowski, A., Koll, A. & Glowiak, T. (2002). J. Mol. Struct. 615, 97–
108.
Fukuda, H., Amimoto, K., Koyama, H. & Kawato, T. (2003). Org.
Biomol. Chem. 1, 1578–1583.
Gawinecki, R., Kolehmainen, E., Janota, H., Kauppinen, R., Nissinen,
M. & Osmialowski, B. (2001). J. Phys. Org. Chem. 14, 797–803.
Gilli, G., Bellucci, F., Ferretti, V. & Bertolasi, V. (1989). J. Am. Chem.
Soc. 111, 1023–1028.
Gilli, G. & Bertolasi, V. (1990). The Chemistry of Enols, edited by Z.
Rappoport, ch. 13, pp. 713–764. Chichester, UK: J. Wiley and Sons.
Gilli, P., Bertolasi, V., Ferretti, V. & Gilli, G. (2000). J. Am. Chem. Soc.
112, 10405–10417.
Gilli, P., Bertolasi, V., Pretto, L., Antonov, L. & Gilli, G. (2005). J. Am.
Chem. Soc. 117, 4943–4953.
Gilli, P., Bertolasi, V., Pretto, L., Ferretti, V. & Gilli, G. (2004). J. Am.
Chem. Soc. 126, 3845–3855.
It seems interesting that the ꢈ-delocalization index (Table 4)
remains essentially fixed to the value of 78% of simple
ꢀ-enaminones (Ia), irrespective of the shortening of the Nꢀ ꢀ ꢀO
distance, the only exception being the strongest hydrogen
˚
bonds in (IIc) (Nꢀ ꢀ ꢀO of 2.55 A), for which this index suddenly
rises to 96%. This can be interpreted on the grounds of the
observation that resonance-assisted hydrogen bonding simply
attests the existence of an intercorrelation between Nꢀ ꢀ ꢀO
shortening and increased ꢈ-delocalization (Gilli et al., 2005),
but tells nothing about the original cause of the shortening (i.e.
the ꢈ-delocalization itself or the steric compression on the
hydrogen-bonded ring). The present data are consistent with
the idea that the ꢈ-delocalization is certainly the driving force
˚
of Nꢀ ꢀ ꢀO contraction up to 2.67 A [the value observed for
simple ꢀ-enaminones (Ia)]. Below this limit the steering
factors become steric and are controlled by the relative
encumbrances of the different substituents. Only when this
compression reaches a maximum [and the Nꢀ ꢀ ꢀO distance a
Gilli, P., Bertolasi, V., Pretto, L. & Gilli, G. (2006). J. Mol. Struct. 790,
40–49.
ˇ
Gilli, P., Bertolasi, V., Pretto, L., Lycka, A. & Gilli, G. (2002). J. Am.
Chem. Soc. 124, 13554–13567.
Gilli, G. & Gilli, G. (2000). J. Mol. Struct. 552, 1–15.
Inabe, T. (1991). New J. Chem. 15, 129–136.
Inabe, T., Luneau, I., Mitani, T., Maruyama, Y. & Takeda, S. (1994).
Bull. Chem. Soc. Jpn, 67, 612–621.
˚
minimum of 2.55 A in (IIc)] can the RAHB mechanism induce
a further delocalization of the resonant fragment, delocaliza-
tion which, however, is not the driving force but the driven
factor of the process.
ꢃ
Acta Cryst. (2006). B62, 1112–1120
Valerio Bertolasi et al.
RAHBs in ꢀ-enaminones 1119

research papers
ˇ
´ ´
Kaitner, B., Mestrovic, E. & Pavlovic, G. (1998). J. Chem. Crystallogr.
28, 77–82.
Nardelli, M. (1995). J. Appl. Cryst. 28, 659.
Pauling, L. (1960). The Nature of the Chemical Bond, 3rd ed. Ithaca,
NY: Cornell University Press.
ˇ
Popovic, Z., Roje, V., Pavlovic, G., Matkovic-Calogovic, D. & Giester,
´
G. (2001). J. Mol. Struct. 597, 39–47.
´
´
´
¨
¨
Nazir, H., Yildiz, M., Yilmaz, H., Tahir, M. N. & Ulku, D. (2000). J.
Mol. Struct. 524, 241–250.
¨
Sheldrick, G. M. (1997). SHELXL97. University of Gottingen,
Germany.
Spek, A. J. (2003). J. Appl. Cryst. 36, 7–13.
Wozniak, K., He, H., Klinowski, J., Jones, W., Dziembowska, T.
& Grech, E. (1995). J. Chem. Soc. Faraday Trans. 91, 77–
85.
Nonius (1997). KappaCCD Server Software. Nonius BV, Delft, The
Netherlands.
Ogawa, K., Kasahara, Y., Ohtani, Y. & Harada, J. (1998). J. Am.
Chem. Soc. 120, 7107–7108.
Otwinowski, Z. & Minor. W. (1997). Method Enzymol. 276, 307–326.
ꢃ
1120 Valerio Bertolasi et al.
RAHBs in ꢀ-enaminones
Acta Cryst. (2006). B62, 1112–1120