Effect of through-bond interaction on conformation and structure in rod-shaped donor - acceptor systems: Part 1. Crystal structures of five N-arylpiperidin-4-one derivatives

Article abstract of DOI:10.1002/hlca.200390081

The crystal structures of five N-arylpiperidin-4-one derivatives 2P2, 3P2, 5P2, 1P3, and 2P3 are presented (Fig. 2 and Tables 1-5) and discussed together with the derivatives 1P2 and 4P2 published previously. In all but one structure, 1P2, the aryl group is in an equatorial position. The piperidine ring adopts a normal chair conformation. In 1P2, the piperidine ring central C-C bonds are significantly elongated, which is consistent with the idea that through-bond interaction is more pronounced in the axial conformation. Through-bond interaction also influences the pyramidalization at the piperidine C(4)-atom in such a way that a strong interaction is directing the ethylene C-atom C(9) into the axial direction.

Full text of DOI:10.1002/hlca.200390081

Helvetica Chimica Acta ± Vol. 86 (2003)
799
Effect of Through-Bond Interaction on Conformation and Structure in
Rod-Shaped Donor± Acceptor Systems
Part 1
Crystal Structures of Five N-Arylpiperidin-4-one Derivatives
by Dirk J. A. De Ridder*, Kees Goubitz, and Henk Schenk
Laboratory for Crystallography, Institute of Molecular Chemistry, Faculty of Science, University of Amsterdam,
Nieuwe Achtergracht 166, NL-1018 WV Amsterdam
(
phone: (31)-20-5257039; fax: (31)-20-5256940; e-mail: dirkdr@science.uva.nl)
1
and Bert Krijnen ) and Jan W. Verhoeven
Laboratory of Organic Chemistry, Institute of Molecular Chemistry, Faculty of Science,
University of Amsterdam, Nieuwe Achtergracht 129, NL-1018 WS Amsterdam
The crystal structures of five N-arylpiperidin-4-one derivatives 2P2, 3P2, 5P2, 1P3, and 2P3 are presented
Fig. 2 and Tables 1 ± 5) and discussed together with the derivatives 1P2 and 4P2 published previously. In all but
(
one structure, 1P2, the aryl group is in an equatorial position. The piperidine ring adopts a normal chair
conformation. In 1P2, the piperidine ring central CÀC bonds are significantly elongated, which is consistent
with the idea that through-bond interaction is more pronounced in the axial conformation. Through-bond
interaction also influences the pyramidalization at the piperidine C(4)-atom in such a way that a strong
interaction is directing the ethylene C-atom C(9) into the axial direction.
1. Introduction. ± In 1968, Hoffmann et al. [1] introduced the term −through-bond
interaction× (TBI) to designate the intramolecular interaction between functional
groups via the connecting s-bonds. According to this concept, the orbitals of the
functional groups may interact with each other as a consequence of their mutual mixing
with the intervening s-bonds.
The study of TBI in rod-shaped bichromophoric compounds containing a one-
electron donor and acceptor has been a main topic in the research group of Verhoeven
and co-workers [2 ± 6]. Krijnen [7] investigated systematically whether through-bond
donor± acceptor interaction can be found in the rod-shaped donor± acceptor com-
pounds shown in Fig. 1. Both of these systems contain a substituted N-atom as the
potential one-electron donor and a substituted exocyclic CˆC bond as the potential
acceptor, whereas donor and acceptor are separated by three s-bonds in a well-defined
arrangement. Variations in the N-aryl group serve to modify the electron-donating
ability of the N-atom, whereas the R substituent will influence the electron affinity of
the acceptor.
1)
Permanent address: Unilever R&D Vlaardingen, Olivier van Noortlaan 120, 3133 AT Vlaardingen,
Netherlands

8
00
Helvetica Chimica Acta ± Vol. 86 (2003)
Fig. 1. Donor ± acceptor compounds investigated. Left: piperidin-4-one-like systems (this paper); right: tropan-
-one-like systems [9]. D ˆ Donor, A ˆ Acceptor; R ˆ CN, MeOCO; Arˆ Ph, 4-MeÀC , 3,5-Me ÀC
-MeOÀC , 4-FÀC , 2,4,6-Me ÀC
3
6
H
4
2
6 3
H ,
4
6
H
4
6
H
4
3
6 2
H .
In the first part of this series of three papers, the crystal structures of five N-
arylpiperidin-4-one derivatives are presented and discussed together with two N-
arylpiperidin-4-one derivatives already published [8]. In the second part of this series,
the crystal structures of some N-aryltropan-3-one derivatives will be discussed [9]. In
the crystal structures described in the first two papers, the exocyclic CˆC bond carries
two CNgroups, and one CNand one COOCH group, respectively. In the last paper of
3
this series, three N-arylpiperidin-4-one derivatives carrying only one non-H substituent
on the exocyclic CˆC bond, and two piperazine derivatives will be presented [10].
2
. Results and Discussion. ± The systematic code introduced by Krijnen [7] to
indicate the compounds studied will also be used here. Piperidine-like systems will be
designated by nPm, whereas tropane-like systems will be denoted by nTm (cf. Scheme).
The first index n indicates which aryl group is attached to the N-atom, whereas the
second index m codes for the substituent at C(4) of the central piperidine ring. m ˆ 0
means that there is no substituent attached to C(4), whereas m ˆ 1 means that the
CH (4) of the piperidine ring is converted to a C(4)ˆO function. The donor± acceptor
2
systems have been synthesized by Knoevenagel-type condensations [11] of the ketones
nP1 and nT1. The condensation products of the reaction with malononitrile will be
designated by nP2 and nT2, whereas the products of the condensation with methyl
cyanoacetate will be denoted as nP3 and nT3 (cf. Scheme).
X-Ray crystal structures have been determined for the compounds 1P2 ± 5P2, 1P3,
and 2P3; Krijnen et al. [8] already reported the synthesis and X-ray structures of 1P2
and 4P2. Several attempts to determine the X-ray structure of 6P2 were unsuccessful.
Crystallographic data of the new structures are given in Table 1, selected bond lengths
and angles of all compounds are compiled in Table 2. In Fig. 2, the ORTEP drawings
are presented, showing the atomic-numbering system.
The N-aryl group adopts the equatorial position, with the exception of 1P2, in
which the Ph group is found in the axial orientation. From temperature-dependent
absorption measurements [7], it has been concluded that, at room temperature in
solution, the axial conformation of 1P2 and 3P2 is populated by ca. 30% of the
molecules, whereas this is decreased in the other nPm compounds. The present
structures, therefore, give a strong indication that such an axial conformer is
thermodynamically feasible under normal conditions.
2
.1 Configuration of the Piperidine Ring. The strength of TBI is thought to be
sensitive to the orientation of the interacting donor and acceptor orbitals with respect

Helvetica Chimica Acta ± Vol. 86 (2003)
801
Scheme. Explanation of the Codes Used to Indicate the Compounds Studied. At the bottom, the reaction
scheme of the Knoevenagel-type condensation [11] is shown.
to the s-relay connecting them, and also for the conformation of the s-relay itself
[
1][12]. Consequently, the conformation of the central piperidine ring has to be
examined first.
As evident from Fig. 2, the piperidine ring in the X-ray structures is found
in a chair conformation; least-squares planes have been calculated through
C(2)ÀC(3)ÀC(5)ÀC(6) (a), C(3)ÀC(4)ÀC(5) (b), and C(2)ÀN(1)ÀC(6) (c) to
define the ring. The sharp angles between these planes, and the distance of N(1) and
C(4) to the central C(2)ÀC(3)ÀC(5)ÀC(6) plane are listed in Table 3.
Throughout the nPm series, the orientation of the C(2) ± C(6) part of the molecule
is roughly the same, although a slight flattening is encountered in 1P3, which has a
lower angle between planes a and b.
In these structures, the orientation of the C(2)ÀN(1)ÀC(6) part of the molecule is
fairly constant, the angle between a and c having a small range.
A priori, it has to be expected that, in the nP2 structures, the piperidine ring displays
(
near) C symmetry, i.e., the bonds N(1)ÀC(2) and N(1)ÀC(6), C(3)ÀC(4) and
s
C(4)ÀC(5), and C(2)ÀC(6) and C(5)ÀC(6), should have (almost) equal lengths. In
3
P2, the piperidine-ring mirror plane coincides with the crystallographic mirror plane,
and, therefore, the ring has C symmetry. Deviations from this (quasi) C symmetry can
s
s

8
02
Helvetica Chimica Acta ± Vol. 86 (2003)
Table 1. Crystallographic Data of N-Arylpiperidin-4-ones 2P2, 3P2, 5P2, 1P3, and 2P3. For the methods used in the
structure determination, see Exper. Part. The crystallographic data of 1P2 and 4P2 have been published in [8].
2P2
3P2
5P2
1P3
2P3
Formula
C
15
H
15
N
3
C
16
H
17
N
3
C
14
H
12FN
3
C
15
H
16
N
2
O
2
16 18 2 2
C H N O
Formula weight
Wavelength/ä
237.3
0.71069
251.3
0.71069
241.3
1.5418
256.3
0.71069
270.3
0.71069
Source
MoK
a
MoK
a
CuK
a
MoK
a
MoK
a
T/K
293
293
293
293
293
a/ä
b/ä
c/ä
a/8
b/8
g/8
V/ä3
Crystal size/mm
Crystal system
Space group
6.695(1)
8.078(2)
12.679(2)
102.59(2)
90.81(2)
103.22(2)
650.0(2)
13.778(3)
12.761(2)
15.687(3)
10.6030(7)
15.5439(10)
7.5771(9)
14.526(5)
12.691(3)
7.246(1)
6.621(2)
11.004(5)
11.253(3)
64.37(2)
84.44(3)
81.89(3)
731.6(5)
94.081(9)
1245.6(2)
104.63(2)
1324.6(6)
2758.1(9)
3
0.35 Â 0.45 Â 0.50 0.38 Â 0.50 Â 0.50 0.20 Â 0.30 Â 0.45 0.20 Â 0.23 Â 0.30 0.23 Â 0.50 Â 0.75
Triclinic
Orthorhombic
Cmca
Triclinic
Monoclinic
Triclinic
≈
≈
P1
P2
1
/n
P2
1
/a
P1
Z
D
m/mm
F(000)
2
8
1.21
0.07
1072
4
1.29
0.73
504
4
1.29
0.09
544
2
1.23
0.08
288
À3
x
/g cm
1.21
0.07
252
À1
q Range/8
Measured refls.
1.6 ± 34.9
5659
2.5 ± 24.8
1263
5.1 ± 59.9
1840
2.2 ± 30.0
3855
2.0 ± 24.9
2547
Obs. refls (I > 2.5s(I)) 1955
976
1588
1263
1719
Refined parameters
164
95
164
173
182
G
R )
3845(156)
0.064
0.055
0.22, 0.01, 0.01
0.93
0.31, À0.35
30534(1063)
0.055
0.052
2.0, 0.01, 0.01
1.09
10483(341)
0.053
0.060
0.5,0.01,0.01
1.03
415(67)
0.041
0.038
0.5, 0.01, 0.01
1.02
5033(207)
0.055
0.050
0.25,0.01,0.01
1.01
a
R
w
b)
c
A, B, C )
Goodness-of-fit
D1 (max,min)/eä
À3
0.35, À0.25
0.18, À0.19
0.62, À0.39
0.39, À0.22
CCDC Deposition No. 190407
190408
190409
190410
190411
a) R ˆ S(jFobs j À k jFcalcj)/S(jFobsj). b) R
ˆ Sw((jFobs j À k jFcalcj) )/S( jFobs j ). c) wÀ1 ˆ (A‡B ¥ (s(Fobs)) ‡C/(s(Fobs)))
2
2
2
w
Table 2. SelectedBondLengths [ä] andAngles [8] with EstimatedStan da rdDeviations in Parentheses . S is the
sum of the bond angles at N(1). Data for 1P2 and 4P2 are taken from [8]. For 3P2, a value in italics indicates that
it is equal to that on the line above by symmetry.
1
P2
2P2
1.465(4)
3P2
1.466(3)
4P2
1.461(2)
5P2
1.454(4)
1P3
1.456(4)
2P3
1.475(4)
N(1)ÀC(2)
N(1)ÀC(6)
N(1)ÀC(10)
C(2)ÀC(3)
C(5)ÀC(6)
C(3)ÀC(4)
C(4)ÀC(5)
C(2)ÀN(1)ÀC(6)
1.459(2)
1.461(2)
1.412(2)
1.546(2)
1.542(2)
1.494(2)
1.495(2)
1.461(3)
1.429(4)
1.524(4)
1.526(4)
1.487(4)
1.491(4)
1.466(3)
1.422(5)
1.538(4)
1.538(4)
1.487(3)
1.487(3)
1.459(2)
1.423(2)
1.526(2)
1.529(2)
1.489(2)
1.489(2)
1.450(4)
1.412(4)
1.525(5)
1.522(5)
1.490(5)
1.481(5)
1.471(4)
1.402(4)
1.512(5)
1.521(5)
1.510(4)
1.504(4)
1.468(4)
1.426(4)
1.487(5)
1.511(4)
1.512(5)
1.507(4)
109.9(2)
C(2)ÀN(1)ÀC(10) 118.9(2)
111.1(2)
112.6(2)
111.3(1)
112.1(3)
110.5(3)
109.6(3)
116.3(2)
116.5(2)
343.9(3)
117.9(2)
117.9(2)
348.4(3)
115.9(1)
116.4(1)
343.6(2)
116.9(3)
117.1(3)
346.1(5)
120.3(2)
120.6(3)
351.4(5)
111.6(2)
115.8(2)
337.0(4)
C(6)ÀN(1)ÀC(10) 118.6(2)
S
347.4(3)

Helvetica Chimica Acta ± Vol. 86 (2003)
803
Fig. 2. Atomic numbering (arbitrary) scheme in nPm. ORTEP [20] drawings of 2P2, 3P2, 5P2, 1P3, and 2P3.
The shapes of the ellipsoids correspond to 30% probability contours of atomic displacement. The H-atoms have
been omitted for clarity.
occur when the aryl group is strongly twisted around the N(1)Àaryl bond, by an
asymmetry in the acceptor part of the molecule (e.g., in the structures nP3 containing
an asymmetrically substituted exocyclic CˆC bond), or by distortions of the solid state
structure due to crystal packing effects. The asymmetry parameters (Table 3)
introduced by Duax et al. [13] indicate that the deviations from (quasi) C symmetry
s
are very small, indicating that the piperidine ring adopts a normal chair conformation.

8
04
Helvetica Chimica Acta ± Vol. 86 (2003)
Fig. 2. (cont.)
Because of crystallographic symmetry in 3P2, the asymmetry parameter is equal to
zero.
The bond lengths (Table 2) indicate that the piperidine ring in 5P2 is not symmetric,
which can be related to the pronounced twist of the aryl group around the N(1)Àaryl
bond (see Sect. 2.2), which induces an asymmetry in the molecule. The piperidine ring
in the other nP2 structures is highly symmetric. In the structures containing the
asymmetric acceptor chromophore (1P3 and 2P3), the piperidine ring is asymmetric in
all cases (Table 2), especially regarding the central C(2)ÀC(3) and C(6)ÀC(5) bonds.
2
.2 Configuration at the Piperidine N-Atom. In all compounds, the piperidine N-
atom N(1) adopts a flattened pyramidal configuration, as indicated by the sum of the

Table 3. Angles [8] between CalculatedLeast-Squares Planes andDistances [ä] of N(1) andC(4) to the Central
Plane a Definedby C(2), C(3), C(5), andC(6); Plane b Definedby C(3), C(4), andC(5); Plane c Definedby
C(2), N(1), andC(6) . Data for 1P2 and 4P2 are taken from [8]; e.s.d.×s calculated with XTAL3.7 [18] from CIF.
s
Asymmetry parameter DC calculated according to Duax et al. [13].
1
P2
2P2
41.9(3)
3P2
50.0(3)
4P2
42.0(2)
5P2
44.3(3)
1P3
36.7(3)
2P3
44.2(3)
Angle between
planes a and b
Angle between
planes a and c
49.0(2)
55.1(2)
53.2(3)
51.9(3)
53.9(2)
53.8(3)
54.7(3)
53.6(3)
Distance of N(1) À 0.687(3) À 0.663(3) À 0.641(4) À 0.666(3) À 0.654(4) À 0.680(4) À 0.684(4)
to plane a
Distance of C(4)
to plane a
0.614(3)
1.4
0.537(4)
0.5
0.629(4)
0
0.538(3)
2.0
0.559(5)
0.8
0.492(5)
0.5
0.578(5)
3.2
DC
s
bond angles at N(1) (S in Table 2). This sum is fairly constant (343.6(2) ± 351.4(5)8),
with the exception of 2P3 in which the significantly smaller angles C(2)ÀN(1)ÀC(10)
and C(6)ÀN(1)ÀC(10) result in a smaller sum of the bond angles at N(1) (337.0(4)8).
AM1 Calculations performed on the N-arylpiperidines nP0 predicted this sum to be ca.
3458 [7], which is thus in good agreement with (most of) the aforementioned values.
From the torsion angles along the N(1)ÀC(10) bond, the orientation of the aryl
group can be derived (Table 4). When the N-atom lone pair and the aromatic p-system
are completely aligned, the torsion angles C(2)ÀN(1)ÀC(10)ÀC(11) (t ) and
1
C(6)ÀN(1)ÀC(10)ÀC(15) (t ) should be (almost) equal. The twist angle b, i.e., the
2
deviation from complete alignment of the lone pair and the p-system, can be calculated
1
as b ˆ ³
2
jj t jÀj t jj (see Table 4).
1
2
Table 4. Absolute Values of SelectedTorsion Angles [8] along the N(1)ÀC(10) BondandCalculatedValues of
the Twist Angles b. Data for 1P2 and 4P2 are taken from [8]; e.s.d.×s calculated with XTAL3.7 [18] from CIF. For
3
P2, t
1
ˆ t
2
by symmetry.
1P2
2P2
3P2
4P2
5P2
1P3
2P3
C(2)ÀN(1)ÀC(10)ÀC(11) (t
1
)
2
)
21.8(3)
22.7(4)
0.5(3)
36.0(4)
14.4(5)
10.8(3)
22.2(5)
22.2(5)
0
45.3(2)
4.7(3)
20.3(2)
49.7(4)
4.4(4)
22.7(3)
14.2(5)
25.1(5)
5.5(4)
63.3(5)
10.7(5)
26.3(4)
C(6)ÀN(1)ÀC(10)ÀC(15) (t
1
b ˆ ³
2
k t
1
j À j t
2
k
From both experiments and calculations [7], it has been concluded that in N-
arylpiperidines b ˆ 30 ± 508, independent of the conformation, i.e., axial or equatorial
orientation of the aryl group. Furthermore, AM1 calculations [7] performed on the
equatorial and axial conformation of 1P2 predict b ꢀ 248 for both conformers. In the
axial X-ray structure 1P2, however, this twist angle is only 0.5(3)8.
In the equatorial X-ray structures, the spread in the angle b (08 < b < 26.3(4)8) is
great, indicative for a less pronounced preference for alignment of the N-atom lone pair
and the aromatic p-system than in the axial X-ray structure. The values of b for 5P2 and
2P3 are, in fact, very close to those predicted by the AM1 calculations performed on
nP0 [7]. It seems, therefore, that the twist angle b in the X-ray structures of the donor±
acceptor systems depends on the equatorial or axial orientation of the aryl group, which

8
06
Helvetica Chimica Acta ± Vol. 86 (2003)
is in contrast with the predictions of the AM1 calculations performed on related
systems, although it should be noted that these latter results refer to the gas phase and
not to the solid state.
In general, a larger twist angle b is coupled to a small value of the sum of the bond
angles S at N(1) (Table 2). In 2P3, S is minimal (337.0(4)8) and b is maximal (26.3(4)8),
whereas, in the structures with a very small b, the sum is large (343.9(3) ± 351.4(5)8);
the correlation coefficient between S and b is À0.75. There is also a negative
correlation between S and bond length N(1)ÀC(10) (À0.73). It seems, therefore,
plausible that the observed flattening of N(1) is related to conjugative interaction
within the anilino chromophore, and, consequently, this flattening must be related to
the twist angle b and bond length N(1)ÀC(10). From the present data, however, no
conclusion can be drawn whether the strength of this conjugative interaction depends
on the type, position, and number of substituents on the phenyl ring.
In the equatorial compounds 2P2, 3P2, and 1P3, there are quite short intra-
molecular H ¥¥¥ H contacts (<1.9 ä) between the piperidine and phenyl ring H-atoms.
This correlates with the small b angles for these compounds. Apparently, the interaction
between the N-atom lone pair and the aromatic p-system overrides the steric strain
caused by the intramolecular H ¥¥¥ H contacts. Probably, this is not the case in the other
compounds because of crystal-packing effects.
2
.3 Elongation of the ×Central× CÀC Bondun de r the Influence of TBI. It has
been found, e.g., by Mislow, Dougherty, and co-workers [14] that, in the solid state,
TBI between two functionalities may result in a bond elongation of the central CÀC
bond in the s-frame connecting these functionalities. In the nPm compounds, TBI
could thus manifest itself in an elongation of the C(2)ÀC(3) and the C(5)ÀC(6)
bonds. In piperidine derivatives, the accepted value of the central CÀC bond is
1
.52 ± 1.53 ä.
The values found for the equatorial structures 2P2, 4P2, and 5P2 (1.522(5) ±
1
.529(2) ä) and nP3 (1.487(5) ± 1.521(5) ä) testify the absence of strong TBI in these
equatorial structures. It is evident that the axial Ph group in 1P2, thus allowing optimal
TBI, results in a significant bond elongation: the central CÀC bond has a mean length
of 1.544(1) ä in this structure. Comparing this value to those of 2P2, 4P2, and 5P2
reveals that TBI results in a bond elongation of at least 0.01 ä. The mean value
(
1.538(3) ä) of the central CÀC bond in 3P2 is close to the corresponding bond length
in 1P2, which may be indicative to a non-negligible TBI.
.4 Pyramidalization at C(4). To study the pyramidalization at the piperidine ring
2
C(4)-atom the torsion angles along the C(4)ÀC(9), C(3)ÀC(4), and C(5)ÀC(4) bonds
were calculated (Table 5). The flattening of the C(2) ± C(6) part of the molecule noted
previously for 1P3 (cf. Sect. 2.1) is nicely reflected in a decrease of the torsion angles
C(2)ÀC(3)ÀC(4)ÀC(5) and C(6)ÀC(5)ÀC(4)ÀC(3) and an increase of the torsion
angles H ÀC(3)ÀC(4)ÀC(9) and H ÀC(5)ÀC(4)ÀC(9) compared to the other
eq
eq
structures.
Substituted alkenes are known to show often slight deviations from planarity [15] if
the two alkene C-atoms, and the four attached atoms cannot define a plane of
molecular symmetry. Inspection of the torsion angles (Table 5) reveals a significant
non-coplanarity in the acceptor moiety: in the case of coplanarity
C(3)ÀC(4)ÀC(9)ÀC(16) and C(5)ÀC(4)ÀC(9)ÀC(17) should be equal to zero.

Helvetica Chimica Acta ± Vol. 86 (2003)
807
Table 5. Absolute Values of SelectedTorsion Angles [8] along the C(4)ÀC(9), C(3)ÀC(4), andC(5) ÀC(4)
Bonds. Data for 1P2 and 4P2 are taken from [8]; e.s.d.×s calculated with XTAL3.7 [18] from CIF. For 3P2, the
torsion angles along C(5)ÀC(4) (in italics) are equal to the equivalent ones along C(3)ÀC(4) by symmetry.
1P2
2P2
3P2
4P2
5P2
1P3
2P3
Along C(4)ÀC(9)
C(3)ÀC(4)ÀC(9)ÀC(16)
C(5)ÀC(4)ÀC(9)ÀC(17)
Along C(3)ÀC(4)
3.5(4)
2.0(4)
1.5(5)
0.1(5)
1.5(6)
1.5(6)
1.3(4)
2.9(4)
0.2(5)
3.3(5)
3.2(5)
0.2(6)
4.3(5)
8.0(6)
C(2)ÀC(3)ÀC(4)ÀC(5)
52.4(3)
3
63
119
182
46.1(4)
13.3(5)
74.0(4)
104.8(4) 116.9(4) 108
178.8(6) 182.4(5) 178
54.6(3)
2.8(5)
65.5(3)
46.7(3)
14
70
48.2(4)
8.8(5)
72.1(4)
109.3(3)
181.4(5)
40.8(4)
24.1(5)
79.9(4)
94.6(4) 105.6(4)
174.5(6) 177.8(6)
47.8(5)
11.0(5)
72.2(4)
H
H
H
eqÀC(3)ÀC(4)ÀC(9)
axÀC(3)ÀC(4)ÀC(5) (t
axÀC(3)ÀC(4)ÀC(9) (t
a
)
)
b
S
1
ˆj t
a
j‡j t j
b
Along C(5)ÀC(4)
C(6)ÀC(5)ÀC(4)ÀC(3)
53.3(3)
1
58
127
185
3.5
46.0(4)
13.6(5)
75.8(3)
102.9(3) 116.9(4) 107
178.7(4) 182.4(5) 178
54.6(3)
2.8(5)
65.5(3)
45.7(3)
12
71
48.7(4)
8.2(5)
72.3(4)
109.1(3)
181.4(5)
1.4(4)
40.7(5)
24.5(5)
80.0(4)
94.2(4) 107.5(4)
174.2(6) 177.6(6)
49.5(5)
10.7(6)
70.1(4)
H
H
H
eqÀC(5)ÀC(4)ÀC(9)
axÀC(5)ÀC(4)ÀC(3) (t
axÀC(5)ÀC(4)ÀC(9) (t
c
)
)
d
S
2
ˆj t
c
j‡j t
d
j
1
q ˆ ³
2
(S
1
‡ S
2
) À1808
À 1.2(4)
2.4(4)
À 2
À 5.6(4)
À 2.4(4)
There are two types of distortions from the ideal planar geometry: twisting around
the C(4)ÀC(9) bond and pyramidalization at C(4). In, e.g., the structures 4P2 and 2P3,
a slight twist around C(4)ÀC(9) is found, perhaps as the result of crystal-field effects.
More interesting, however, is the pyramidalization at C(4). A priori, this distortion
can direct C(9) in two different directions as indicated in Fig. 3. The angle q, defined in
Fig. 3, is a measure for the degree and direction of the pyramidalization at C(4). A
negative value for q indicates an equatorial bending of C(9) and a positive q bending
into the axial direction.
Fig. 3. Possible modes of pyramidalization at C(4) and the corresponding Newman projections along the
C(3)ÀC(4) bond; q ˆ (d ‡ e) À1808.

8
08
Helvetica Chimica Acta ± Vol. 86 (2003)
Theoretical studies on alkenes and carbonyls [15][16] predict that the C-atom will
pyramidalize toward a staggered geometry in order to relieve torsional interactions
between the allylic bonds and the two s-bonds and p-orbitals attached to the alkene C-
atom. Such a pyramidalization toward the bond most parallel with the p-system is
confirmed by a survey of neutron-diffraction crystal structures of amino acids and
dipeptides [16]. For our systems, this would correspond to a bending of C(9) into the
equatorial direction, thus toward H ÀC(3) and H ÀC(5) (q negative).
ax
ax
In Table 5, the values for q, obtained by averaging the q values determined from the
torsion angles along the C(3)ÀC(4) and C(5)ÀC(4) bonds, are listed.
Pyramidalization into the axial direction (q > 08) leads to a greater electron density
of the p-orbital on C(4) into the equatorial direction. Consequently, the p-lobe on C(4)
becomes more antiparallel to the central CÀC bond, thus allowing more effective
coupling of the acceptor with the N(1) lone pair. Such donor± acceptor interactions
result in an enthalpic stabilization of the molecular system. Therefore, the energetically
unfavorable distortion of the CˆC bond, i.e., the bending of C(9) resulting in a
pyramidalization at C(4) with q > 08, is probably compensated by a stabilization offered
by an increase of the donor± acceptor interaction upon such a pyramidalization.
If so, a pyramidalization with q > 08 has to be anticipated for 1P2 (axial Ph group),
whereas, for the other nPm structures, a negative pyramidalization is expected, because
of the much weaker TBI in structures with an equatorial aryl group. As evident from
Table 5, the pyramidalization in 1P2 is indeed positive (q ˆ ‡3.58), whereas the
structures 2P2, 4P2, 1P3, and 2P3 have negative q values. Although TBI is maximal in
the axial conformation of the donor± acceptor systems, non-negligible TBI also occurs
in the equatorial conformation, and this weak TBI might result in a small positive q as
found for 3P2 and 5P2. The assumption that, especially in the structure 3P2, some
minor manifestations of TBI are encountered, is also supported by the fact that the
central CÀC bond in this structure (mean value 1.538(3) ä) is somewhat elongated
compared to the other structures with an equatorial aryl group (Table 2). In the case of
5
P2, however, the central CÀC bond shows no elongation (mean value 1.528(4) ä),
indicating the absence of TBI. It should be stressed, however, that crystal-packing
effects could also result in the small positive q values observed for 3P2 and 5P2!
3
. Conclusions. ± In the available X-ray structures of the nPm compounds, the aryl
group is found in the equatorial orientation, with the exception of 1P2, in which an
axial preference of the Ph group is found in the solid state. In 1P2, the central CÀC
bond is significantly elongated, whereas, in the structures with an equatorial aryl group,
such elongation is virtually absent. These observations are consistent with the
assumption that in the equatorial conformation TBI is far less pronounced than in
the axial conformation.
Not only the central CÀC bond is influenced by the donor± acceptor interaction,
but also the pyramidalization at C(4). The X-ray structures indicate that, in general, the
degree and direction of this pyramidalization is mainly controlled by TBI. In structures
with strong TBI ± judged from the pronounced elongation of the central CÀC bond ±
the pyramidalization at C(4) directs C(9) into the axial direction. In structures with a
less pronounced TBI ± equatorial aryl group ± this pyramidalization is reduced or
directs C(9) into the sterically preferred equatorial direction.

Helvetica Chimica Acta ± Vol. 86 (2003)
809
Experimental Part
1. X-Ray Crystal Structure Determinations. The reflections were collected on an Enraf-Nonius CAD-4
diffractometer with graphite monochromated CuK or MoK radiation (Table 1). The unit-cell dimensions
a
a
resulted from a least-squares fit of the setting angles of 23 centred reflections. The structures were solved by
CRUNCH [17] and refined with XTAL3.7 [18]. Full-matrix least-squares refinement was used, anisotropic for
non-H-atoms and isotropic for H-atoms. The H-atoms were kept fixed at their calculated positions with fixed
2
U
eq ˆ 0.10 ä . An extinction correction [19] was applied, but no absorption correction.
The crystallographic data (excluding structure factors) of all new structures presented in Table 1 and Fig. 2
[20] have been deposited with the Cambridge Crystallographic Data Centre (CCDC); deposition numbers are
given in Table 1. Copies of the data can be obtained, free of charge, on application to the CCDC, 12 Union Road,
Cambridge, CB21EZ, UK (fax: ‡44-1223-336033; e-mail: deposit@ccdc.cam.ac.uk).
2. Syntheses. The compounds can be synthesized by Knoevenagel-type condensation [11] of malononitrile
(
for nP2 compounds) or methyl cyanoacetate (for nP3 compounds) and the appropriate ketones nP1 (Scheme).
.1. Synthesis of N-Arylpiperidin-4-ones nP1. 2.1.1. 1-(4-Methylphenyl)piperidin-4-one (2P1). A soln. of 1,5-
dichloropentan-3-one (1.00 g, 6.45 mmol) and a soln. of 1.1 equiv. of p-toluidine (0.76 g, 7.09 mmol), both in
0 ml of MeOH (dried over mol. sieves 4 ä), were added simultaneously to a stirred slurry of 1.40 g of Na CO
in 25 ml of dry MeOH during ca. 0.75 h. After the addition of the two solns., the mixture was refluxed for 1 ±
.5 h. After cooling to r.t., 100 ml of Et O was added. The mixture was dried (MgSO ), filtered over a fine glass
filter, and evaporated to give a light yellow oil. Purification by flash column chromatography (FC) (petroleum
ether 40 ± 60/Et O 1:1) yielded an almost colorless liquid, which solidified to give 2P1. White solid. Yield: 0.63 g
3.33 mmol, 47% calculated on p-toluidine). M.p. 25 ± 298. IR (CHCl ): 3000w, 2960s, 2920s, 2810m, 1708s, 1610s,
): 7.12 (d, J ꢀ 8.5, HÀC(3), HÀC(4) of Ph); 6.91 (d, J ꢀ 8.6, HÀC(2),
(2), CH (3), CH (5)); 2.29 (s, Me).
(6)); 2.55 (t, J ꢀ 6.0, CH
.1.2. 1-(3,5-Dimethylphenyl)piperidin-4-one (3P1) was prepared as described for 2P1 by using 3,5-
2
2
2
3
1
2
4
2
(
1
3
1
510s, 810s. H-NMR (200 MHz, CDCl
3
HÀC(6) of Ph), 3.55 (t, J ꢀ 6.0, CH
2
2
2
2
2
dimethylaniline (0.86 g, 7.10 mmol) instead of p-toluidine. The same workup and purification as for 2P1 yielded
an almost colorless oil, which solidified to 3P1. White solid. Yield: 0.71 g (3.49 mmol, 49%). M.p. 40 ± 448. IR
1
(
CHCl
HÀC(2), HÀC(6) of Ph); 6.57 (s, HÀC(4) of Ph); 3.59 (t, J ꢀ 6.0, CH
CH (5)); 2.30 (s, 2 Me).
.1.3. 1-(4-Fluorophenyl)piperidin-4-one (5P1) was synthesized as described for 2P1 by using 2.8 g of
CO in 50 ml of MeOH, 1,5-dichloropentan-3-one (2.00 g, 12.9 mmol) and 4-fluoroaniline (1.43 g,
2.9 mmol). The mixture was refluxed (1.5 h) and then evaporated to ca. 50 ml. After the addition of 150 ml
of a 0.1m KOH soln., the turbid mixture was extracted 3 times with CH Cl . The combined org. layers were dried
MgSO ), filtered, and evaporated to yield a brown oil. FC (silica gel; petroleum ether 40 ± 60/Et O 1:1) yielded
the products as an off-white solid. Recrystallization from petroleum ether 40 ± 60/Et O yielded 5P1. Cream-
): 3000w, 2995m, 2980m, 2900m, 2800m,
(2), CH (6)); 2.57
3
): 3000w, 2995s, 2960s, 2915s, 2815m, 1705s, 1590s, 1470s, 828s. H-NMR (200 MHz, CDCl
3
): 6.62 (s,
2
(2), CH
2
(6)); 2.55 (t, J ꢀ 6.0, CH (3),
2
2
2
Na
1
2
3
2
2
(
4
2
2
colored solid. Yield: 0.73 g (3.78 mmol, 29%). M.p. 77 ± 798. IR (CHCl
1
(
3
1
705s, 1500s, 822s, 809s. H-NMR (200 MHz, CDCl
3
): 6.97 (m, 4 arom. H); 3.50 (t, J ꢀ 6.1, CH
2
2
t, J ꢀ 6.1, CH
.2. Syntheses of the Donor ± Acceptor Systems nP2 and nP3. 2.2.1. 2-[1-N(4-Methylphenyl)pyrrolidin-4-
ylidene]propanedinitrile (2P2) was synthesized by stirring and refluxing the ketone 2P1 (0.21 g, 1.11 mmol),
malononitrile (83 mg, 1.26 mmol), 120 mg of AcONH , and 0.20 ml of AcOH in 10 ml of toluene for 1 h in a
Dean-Stark apparatus. After cooling, some toluene was added (ca. 20 ml), and the same workup as for 1P2 [8]
yielded a dark brown residue. FC (silica gel; petroleum ether 40 ± 60/Et O 3 :2) yielded 0.15 g of a yellow solid.
Recrystallization from cyclohexane/Et O (with a few drops of CH Cl ) yielded 2P2. Large yellow crystals,
2 2
(3), CH (5)).
2
4
2
2
2
2
suitable for X-ray analysis. Yield: 0.15 g (0.63 mmol, 57%). M.p. ca. 1118 (dec.). IR (CHCl
3
): 3020m, 2995m,
1
2
955m, 2910m, 2805m, 2210s, 1590s, 1505s, 804s. H-NMR (200 MHz, CDCl
3
): 7.12 (d, J ꢀ 8.3, HÀC(3), HÀC(5)
(2), CH (3),
(6)); 2.86 (t, J ꢀ 5.6, CH
): 180.6 (C(4)). 146.3 (C(1) of Ar); 130.3 (C(4) of Ar);
of Ar); 6.87 (d, J ꢀ 8.6, HÀC(2), HÀC(6) of Ar); 3.45 (t, J ꢀ 5.6, CH
2
2
2
1
3
CH
2
(5)); 2.29 (s, Me). C-NMR (50.3 MHz, APT, CDCl
3
1
30.0 (C(3), C(5) of Ar); 116.5 (C(2), C(5) of Ar); 111.3 (CN); 83.4 (C(4)ˆC); 50.3 (C(2), C(6)); 33.5 (C(3),
‡
C(5)); 20.3 (Me). HR-MS: 237.1280 (C15
H
15
N
3
; calc. 237.1266).
2
.2.2. 2-[1-(3,5-Dimethylphenyl)pyrrolidin-4-ylidene]propanedinitrile (3P2) was synthesized by refluxing
P1 (0.25 g, 1.23 mmol), malononitrile (95 mg, 1.44 mmol), 95 mg of AcONH , and 0.22 ml of AcOH in 10 ml of
toluene for 1 h in a Dean-Stark apparatus. After the same workup as for 2P2, the product was purified by FC
silical gel; petroleum ether 40 ± 60/Et O 1:1) and subsequent recrystallization from Et O/CH Cl 1:1 yielded
P2. Light-yellow crystals. Yield: 0.20 g (0.80 mmol, 65%). M.p. ca. 1508 (dec.). IR (CHCl ): 3030w, 3000m,
3
4
(
2
2
2
2
3
3

8
10
Helvetica Chimica Acta ± Vol. 86 (2003)
1
2
CH
(
960m, 2920m, 2820m, 2225s, 1590s, 825m. H-NMR (200 MHz, CDCl
3
): 6.59 (s, 3 arom. H); 3.48 (t, J ꢀ 5.6,
1
3
2
(2), CH
2
(6)); 2.84 (t, J ꢀ 5.6, CH
2
(3), CH
2
(5)); 2.30 (s, 2 Me). C-NMR (62.9 MHz, APT, CDCl ): 180.7
3
C(4)); 148.5 (C(1) of Ar); 139.2 (C(2), C(5) of Ar); 122.6 (C(4) of Ar); 114.1 (C(2) or C(6) of Ar); 111.4 (C(2)
or C(6) of Ar); 111.4 ppm (CN); 83.4 (C(4)ˆC); 50.0 (C(2), C(6)); 33.6 (C(3), C(5)); 21.6 (Me). HR-MS:
‡
2
51.1383 (C16
H
17
N
3
; calc. 251.1422).
2
.2.3. 2-[1-(4-fluorphenyl)pyrrolidin-4-ylidene]propanedinitrile (5P2) was synthesized by refluxing 5P1
193 mg, 1.00 mmol), malononitrile (86 mg, 1.30 mmol), and 89 mg of AcONH in 5 ml of toluene for 1.5 h in a
Dean-Stark apparatus. The usual workup yielded a yellow solid, which was purified by FC (silica gel; petroleum
ether 40 ± 60/Et O 1:1). Recrystallization from Et O/CH Cl 1:1 yielded 5P2. Pale-yellow crystals. Yield:
33 mg (0.55 mmol, 55%). M.p. 134 ± 1388. IR (CHCl ): 3020m, 2960m, 2900w, 2810m, 2225s, 1595s, 1505s, 825s,
): 6.94 (m, 4 arom. H); 3.37 (t, J ꢀ 5.7, CH (2), CH (6)); 2.86 (t, J ꢀ 5.6, CH (3),
(5)). C-NMR (50.3 MHz, APT, CDCl
(
4
2
2
2
2
1
3
1
8
10s. H-NMR (250 MHz, CDCl
3
2
2
2
1
3
CH
2
3
): 180.0 (C(4)); 157.7 (d, J(C,F) ꢀ 240.7, C(4) of Ar); 145.5 (d,
J(C,F) ꢀ 2.5, C(1) of Ar); 118.5 (d, J(C,F) ꢀ 7.7, C(2), C(6) of Ar); 116.0 (d, J(C,F) ꢀ 22.4, C(3), C(5) of Ar);
‡
1
2
11.2 (CN); 83.9 (C(4)ˆC); 50.9 (C(2), C(6)); 33.7 (C(3), C(5)). HR-MS: 241.1002 (C14
H
12FN
3
; calc.
41.1015).
2
.2.4. Methyl 2-Cyano-2-(1-phenylpyrrolidin-4-ylidene)acetate (1P3) was obtained from Hermant et al. [21].
.2.5. Methyl 2-Cyano-2-[1-(4-methylphenyl)pyrrolidin-4-ylidene]acetate (2P3) was prepared as described
, and
.57 ml of AcOH. After the same workup as for 2P2, the product was purified by FC (silica gel; petroleum ether
0 ± 60/Et O 1:) to yield 0.44 g of a yellow solid. Recrystallization from petroleum ether 40 ± 60/Et O 1:1
):
2
for 2P2 by using 2P1 (0.61 g, 3.22 mmol), methyl cyanoacetate (358 mg, 3.61 mmol), 328mg of AcONH
4
0
4
2
2
yielded 2P3 as crystals suitable for X-ray analysis. Yield: 0.44 g (1.63 mmol, 51%). M.p. 85.5 ± 86.58. IR (CHCl
3
1
3
030m, 3000m, 2950m, 2920m, 2810m, 2220m, 1725s, 1605s, 1508s, 808m. H-NMR (200 MHz, CDCl
3
): 7.12 (d,
J ꢀ 8.3, HÀC(3), HÀC(5) of Ar); 6.87 (d, J ꢀ 8.6, HÀC(2), HÀC(6) of Ar); 3.85 (s, MeO); 3.44 (t, J ꢀ 5.8,
1
3
CH
CDCl
2
(2); 3.35 (m, CH
2
(6)); 3.27 (m, CH
2
(5)); 2.90 (t, J ꢀ 5.7, CH
2
(3)); 2.29 (s, Me). C-NMR (50.3 MHz, APT,
3
): 175.9 (C(4)); 162.0 (CO); 147.0 (C(1) of Ar); 129.8 (C(3), C(5) of Ar); 129.6 (C(4) of Ar); 116.3 (C(2),
C(6) of Ar); 115.0 (CN); 102.7 (C(4)ˆC); 52.5 (MeO); 50.3 (C(2)); 50.0 (C(6)); 35.2 (C(3)); 30.6 (C(5)); 20.3
‡
(
Me). HR-MS: 270.1350 (C16
H
18
N
2
O
2
; calc. 270.1368).
REFERENCES
1] R. Hoffmann, A. Imamura, W.J. Hehre, J. Am. Chem. Soc. 1968, 90, 1499.
2] A. W. J. D. Dekkers, J. W. Verhoeven, W. N. Speckamp, Tetrahedron 1973, 29, 1691.
3] P. Pasman, J. W. Verhoeven, Th. J. de Boer, Tetrahedron 1976, 32, 2827; P. Pasman, J. W. Verhoeven, Th. J.
de Boer, Tetrahedron Lett. 1977, 207; P. Pasman, J. W. Verhoeven, Th. J. de Boer, Chem. Phys. Lett. 1978, 59,
[
[
[
381; P. Pasman, N. W. Koper, J. W. Verhoeven, Recl. Trav. Chim. Pays-Bas 1982, 101, 363; P. Pasman, F. Rob,
J. W. Verhoeven, J. Am. Chem. Soc. 1982, 104, 5127; P. Pasman, G. F. Mes, N. W. Koper, J. W. Verhoeven, J.
Am. Chem. Soc. 1985, 107, 5839; P. Pasman, Ph. D. Thesis, University of Amsterdam, 1980.
[
4] G. F. Mes, H. J. van Ramesdonk, J. W. Verhoeven, Recl. Trav. Chim. Pays-Bas 1983, 102, 55; G. F. Mes, H. J.
van Ramesdonk, J. W. Verhoeven, J. Am. Chem. Soc. 1984, 106, 1335; G. F. Mes, B. de Jong, H. J. van
Ramesdonk, J. W. Verhoeven, J. M. Warman, M. P. de Haas, L. E. W. Horsman-van den Dool, J. Am.
Chem. Soc. 1984, 106, 6524; G. F. Mes, Ph. D. Thesis, University of Amsterdam, 1985.
[5] N. S. Hush, M. N. Paddon-Row, E. Cotsaris, H. Oevering, J. W. Verhoeven, M. Heppener,Chem. Phys. Lett.
1985, 117, 8; J. M. Warman, M. P. de Haas, M. N. Paddon-Row, E. Cotsaris, N. S. Hush, H. Oevering, J. W.
Verhoeven, Nature 1986, 320, 615; J. M. Warman, M. P. de Haas, H. Oevering, J. W. Verhoeven, M. N.
Paddon-Row, A. M. Oliver, N. S. Hush, Chem. Phys. Lett. 1986, 128, 95; J. W. Verhoeven, M. N. Paddon-
Row, N. S. Hush, H. Oevering, M. Heppener,Pure Appl. Chem. 1986, 58, 1285; H. Oevering, M. N. Paddon-
Row, M. Heppener, A. M. Oliver, E. Cotsaris, J. W. Verhoeven, N. S. Hush, J. Am. Chem. Soc. 1987, 109,
3258, H. Oevering, J. W. Verhoeven, M. N. Paddon-Row, J. M. Warman, Tetrahedron 1989, 45, 4751; H.
Oevering, Ph. D. Thesis, University of Amsterdam, 1988.
[
6] J. Kroon, A. M. Oliver, M. N. Paddon-Row, J. W. Verhoeven, Recl. Trav. Chim. Pays-Bas 1988, 107, 509;
A. M. Oliver, D. C. Craig, M. N. Paddon-Row, J. Kroon, J. W. Verhoeven, Chem. Phys. Lett. 1988, 150, 366.
7] B. Krijnen, Ph. D. Thesis, University of Amsterdam, 1990.
[
[
8] B. Krijnen, H. B. Beverloo, J. W. Verhoeven, C. A. Reiss, K. Goubitz, D. Heijdenrijk, J. Am. Chem. Soc.
1989, 111, 4433.
[9] D. J. A. De Ridder, K. Goubitz, H. Schenk, B. Krijnen, J. W. Verhoeven, Helv. Chim. Acta 2003, 86, 812.

Helvetica Chimica Acta ± Vol. 86 (2003)
811
[
[
10] D. J. A. De Ridder, K. Goubitz, H. Schenk, B. Krijnen, J. W. Verhoeven, in preparation
11] G. Jones, in −Organic Reactions×, Eds. R. Adams, A. H. Blatt, V. Boekelheide, T. L. Cairns, A. C. Cope, D. J.
Cram, H. O. House; J. Wiley & Sons Inc., New York, 1967, Vol. 15, Chapter 2, p. 204.
12] R. Hoffmann, Acc. Chem. Res. 1971, 4, 1; R. Gleiter, Angew. Chem. 1974, 86, 770; M. N. Paddon-Row, Acc.
Chem. Res. 1982, 15, 245.
[
[
[
13] W. L. Duax, C. M. Weeks, D. C. Rohrer, Top. Stereochem. 1976, 9, 271.
14] D. A. Dougherty, W. D. Hounshell, H. B. Schlegel, R. A. Bell, K. Mislow, Tetrahedron Lett. 1976, 3479;
D. A. Dougherty, H. B. Schlegel, K. Mislow, Tetrahedron 1978, 34, 1441.
[15] N. G. Rondan, M. N. Paddon-Row, P. Caramella, K. N. Houk, J. Am. Chem. Soc. 1981, 103, 2436; K. N.
Houk, N. G. Rondan, F. K. Brown, W. L. Jorgensen, J. D. Madura, D. C. Spellmeyer, J. Am. Chem. Soc.
1983, 105, 5980.
[16] G. A. Jeffrey, K. N. Houk, M. N. Paddon-Row, N. G. Rondan, J. Mitra,J. Am. Chem. Soc. 1985, 107, 321.
[17] R. de Gelder, R. A. G. de Graaff, H. Schenk, Acta Crystallogr., Sect. A 1993, 49, 287.
[18] −XTAL3.7 System×, Eds. S. R. Hall, D. J. du Boulay, R. Olthof-Hazekamp, University of Western Australia,
Lamb, Perth, Australia, 2000.
[
[
[
19] W. H. Zachariasen, Acta Crystallogr. 1967, 23, 558.
20] A. L. Spek, Acta Crystallogr., Sect. A 1990, 46, C-34.
21] R. M. Hermant, N. A. C. Bakker, T. Scherer, B. Krijnen, J. W. Verhoeven, J. Am. Chem. Soc. 1990, 112,
1214.
ReceivedSeptember 24, 2002