Structural Effects of C6 Substitution in 6-(4-(Dimethylamino)phenyl)fulvenes

Article abstract of DOI:10.1021/jo990931x

The single-crystal X-ray structures of 6-(4-(dimethylamino)phenyl)-6-methylfulvene (2) and of two polymorphs of 6-(4-(dimethylamino)phenyl)-6-phenylfulvene (3(P2₁/c) and 3(Pca2₁)) have been determined and the structures of a series o

Full text of DOI:10.1021/jo990931x

J . Org. Chem. 1999, 64, 9067-9076
9067
Str u ctu r a l Effects of C6 Su bstitu tion in
6-(4-(Dim eth yla m in o)p h en yl)fu lven es
Matthew L. Peterson, J effrey T. Strnad, Thomas P. Markotan, Carl A. Morales,
Donald V. Scaltrito, and Stuart W. Staley*
Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, Pennsylvania 15213
Received J une 8, 1999
The single-crystal X-ray structures of 6-(4-(dimethylamino)phenyl)-6-methylfulvene (2) and of two
polymorphs of 6-(4-(dimethylamino)phenyl)-6-phenylfulvene (3(P21/c) and 3(Pca21)) have been
determined and the structures of a series of 6-arylfulvenes (1-8) have been optimized at the HF/
6-31G level. Analysis of these structures reveals how resonance and crystal lattice effects influence
the degree of coplanarity between the aryl and fulvene rings. The torsional angles at the aryl-
fulvene bonds are significantly larger in the optimized structures than in the X-ray structures.
Natural bond orbital π charges and dipole moments calculated for the X-ray and optimized structures
show that the crystalline environment enhances molecular polarization. Qualitative responses for
second harmonic generation in powder samples have been observed in 2, 3(Pca21), and 5. Compounds
2 and 3(Pca21) have similar packing motifs despite packing in different space groups.
In tr od u ction
to have alternating single and double bonds and to be
only weakly polar (µ ) 0.42 D).3g Substituted fulvenes
have been shown to maintain this basic structural
element, although structural perturbations suggestive of
an increased importance of an “aromatic” five-membered
ring have been observed.3e,f In addition, substitution
greatly increases the polarity of fulvene,8b a response
indicative of a highly polarizable moiety.2
The study of fulvene and substituted fulvenes1 encom-
passes many areas of chemistry, including structure,2,3
NMR4 and electronic5 spectroscopy, reactivity,6 materials
research,7 and the study of fundamental concepts such
as polarity8 and aromaticity.9 Fulvene has been studied
in the gas phase by microwave spectroscopy and found
Recent work has focused on the transmission of sub-
stituent effects through aryl spacers to the fulvene moiety
from the standpoint of structure, charge distribution and
chemical reactivity.2,4,6c,d,f,g,i,7b,8b The relative importance
of the resonance forms that describe the structure and
electron distribution in these compounds has been dis-
cussed frequently over the years.
(1) Reviews: (a) Yates, P. Adv. Alicycl. Chem. 1968, 2, 59. (b)
Bergmann, E. D. Chem. Rev. 1968, 68, 41. (c) Lloyd, D. Stud. Org.
Chem. 1984, 16 (Nonbenzenoid Conjugated Carbocyclic Compounds).
(2) Wingert, L. M.; Staley, S. W. Acta Crystallogr. 1992, B48, 782.
(3) (a) Duda, L.; Erker, G.; Fro¨hlich, R.; Zippel, F. Eur. J . Inorg.
Chem. 1998, 1153. (b) Bolte, M.; Amon, M. Acta Crystallogr. 1997, C53,
1354. (c) Teuber, R.; Ko¨ppe, R.; Linti, G.; Tacke, M. J . Organomet.
Chem. 1997, 545-546, 105. (d) Rau, D.; Behrens, U. J . Organomet.
Chem. 1990, 387, 219. (e) Ammon, H. L. Acta Crystallogr. 1974, B30,
1731. (f) Bo¨hme, R.; Burzlaff, H. Chem. Ber. 1974, 107, 832. (g) Baron,
P. A.; Brown, R. D.; Burden, F. R.; Domaille, P. J .; Kent, J . E. J . Mol.
Spectrosc. 1972, 43, 401. (h) Burzlaff, H.; Hartke, K.; Salamon, R.
Chem. Ber. 1970, 103, 156. (i) Chiang, J . F.; Bauer, S. H. J . Am. Chem.
Soc. 1970, 92, 261. (j) Norman, N.; Post, B. Acta Crystallogr. 1961,
14, 503.
(4) (a) Bircher, H.; Neuenschwander, M. Helv. Chim. Acta 1989, 72,
1697. (b) Bo¨nzli, P.; Otter, A.; Neuenschwander, M.; Huber, H.;
Kellerhals, H. P. Helv. Chim. Acta 1986, 69, 1052. (c) Sardella, D. J .;
Keane, C. M.; Lemonias, P. J . J . Am. Chem. Soc. 1984, 106, 4962. (d)
Otter, A.; Mu¨hle, H.; Neuenschwander, M.; Kellerhals, H. P. Helv.
Chim. Acta 1979, 62, 1626. (e) Pines, A.; Rabinovitz, M. J . Chem. Soc.
B 1971, 385.
(5) (a) Fabian, J .; J unek, H. Dyes Pigments 1985, 6, 251. (b) Griffiths,
J .; Lockwood, M. J . Chem. Soc., Perkin Trans. 1 1976, 48. (c) Weiss,
R.; Murrell, J . N. Tetrahedron 1970, 26, 1131.
A significant increase in the second-order nonlinear
optical response was found upon inclusion of a double-
bond spacer between the aryl and fulvene rings.7a,d
Because π-electrons are more easily perturbed then
σ-electrons, this work is indicative of long range π-elec-
tron communication in these compounds. However, a
limited understanding of the electronic interactions in
(6) (a) Nair, V.; Kumar, S. Tetrahedron 1996, 52, 4029. (b) Nair, V.;
Kumar, S.; Anilkumar, G.; Nair, J . S. Tetrahedron 1995, 51, 9155. (c)
Gugelchuk, M. M.; Chan, P. C.-M.; Sprules, T. J . J . Org. Chem. 1994,
59, 7723. (d) Gugelchuk, M.; Paquette, L. A. J . Am. Chem. Soc. 1991,
113, 246. (e) Chandrasekhar, S.; Venkatesan, V. J . Chem. Res., Synop.
1989, 276. (f) Gugelchuk, M.; Paquette, L. A. Isr. J . Chem. 1989, 29,
165. (g) Paquette, L. A.; Gugelchuk, M. J . Org. Chem. 1988, 53, 1835.
(h) Houk, K. N.; George, J . K.; Duke, R. E., J r. Tetrahedron 1974, 30,
523. (i) Kresze, G.; Rau, S.; Sabelus, G.; Goetz, H. Liebigs Ann. Chem.
1961, 648, 57.
(7) (a) Kawabe, Y.; Ikeda, H.; Sakai, T.; Kawasaki, K. J . Mater.
Chem. 1992, 2, 1025. (b) Kondo, K.; Goda, H.; Takemoto, K.; Aso, H.;
Sasaki, T.; Kawakami, K.; Yoshida, H.; Yoshida, K. J . Mater. Chem.
1992, 2, 1097. (c) Papadopoulos, M. G.; Waite, J . J . Chem. Soc.,
Faraday Trans. 1990, 86, 3525. (d) Ikeda, H.; Kawabe, Y.; Sakai, T.;
Kawasaki, K. Chem. Phys. Lett. 1989, 157, 576.
(8) (a) Replogle, E. S.; Trucks, G. W.; Staley, S. W. J . Phys. Chem.
1991, 95, 6908 and references cited; (b) Kresze, G.; Goetz, H. Chem.
Ber. 1957, 90, 2161.
(9) (a) Tomas, X.; Andrade, J . M.; Alvarez-Larena, A. Talanta 1999,
48, 781 (b) Krygowski, T. M.; Cyran˜ski, M. Tetrahedron 1996, 52, 1713.
(c) Krygowski, T. M.; Ciesielski, A.; Cyran˜ski, M. Chem. Papers 1995,
49, 128.
10.1021/jo990931x CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/13/1999

9068 J . Org. Chem., Vol. 64, No. 25, 1999
Peterson et al.
Ta ble 1. X-r a y Da ta for 2, 3(P 21/c), a n d 3(P ca 21)
bond angles, and torsional angles for 1-8 are listed in
Tables 5-7, respectively.
2
3(P21/c)
C20H19
273.36
3(Pca21)
C20H19
formula
formula wt
(g/mol)
color and
habit
C
15H17
N
N
N
211.30
273.36
orange square orange
plates blades
orthorhombic monoclinic
7.426(2)
7.676(2)
21.694(4)
90
orange
needles
orthorhombic
30.890(6)
8.585(2)
5.850(1)
90
cryst syst
a (Å)
10.295(5)
12.257(5)
13.129(6)
90
b (Å)
c (Å)
R (°)
â (°)
90
90
110.42(2)
90
90
90
γ (°)
V (Å3)
Z
1236.7(4)
4
1552.6(12)
4
1551.4(5)
4
space grp
F(000)
P212121
456
P21/c
584
1.169
Pca21
584
1.170
D (calcd) (g/cm3) 1.135
cryst size (mm)
0.42 × 0.40
× 0.11
0.496
0.36 × 0.47
× 0.49
0.067
0 < h < 12
0 < k < 14
0.15 × 0.09
× 0.04
0.511
0 < h < 34
0 < k < 9
µ (mm-1
)
collection range -2 < h < 8
-2 < k < 8
-6 < l < 24
-15 < l < 14 -6 < l < 0
2737
unique data
measured
obsd data
no. of variables
R
1102
1295
1099
162
0.0536
0.1451
1920
248
0.0453
0.1275
1294
209
0.0491
0.0869
Frequency doubling of a 1064 nm laser beam was
qualitatively observed for powder samples of 2, 5, and
3(Pca21), although the latter compound showed only
weak second harmonic generation (SHG). These experi-
ments did not afford quantitative results because the
NLO response depends on the crystal sizes as well as on
the molecular response and crystal packing.14
wR2
these compounds makes the optimization of technologi-
cally important properties difficult.
In the present study, we have determined the single-
crystal X-ray structures of several strong donor-substi-
tuted fulvenes, 6-(4-(dimethylamino)phenyl)-6-methyl-
fulvene (2) and two polymorphs of 6-(4-(dimethylamino)-
phenyl)-6-phenylfulvene (3). Steric and crystal packing
effects have been elucidated by comparing the structures
of similarly substituted compounds and by comparing the
X-ray structures with HF/6-31G geometry-optimized
structures. Finally, macroscopic nonlinear optical effects
(second harmonic generation) have been evaluated quali-
tatively for several of these compounds.
Discu ssion
The major objective of this study is to elucidate the
interdependence of π delocalization (π polarization and
through resonance) and molecular structure in 6-phenyl-
and 6-(4-(dimethylamino)phenyl)fulvenes and to deter-
mine how these effects are modified on incorporation into
a crystal lattice. Because π delocalization is necessarily
inextricably linked with steric effects of the other sub-
stituent at C6, our approach to this problem is as follows.
We have optimized the structures of 1-8 by ab initio
molecular orbital theory at the Hartree-Fock (HF) level;
these structures are appropriate for the isolated (“gas
phase”) molecules. We have also determined the struc-
tures of 2 and 3 in the solid state by single-crystal X-ray
diffraction, and the crystal structures of 1, 4, and 8 were
Resu lts
Single-crystal molecular structures for 2, 3(P21/c), and
3(Pca21) were determined by X-ray diffraction. Synthetic
details and procedures for crystal growth are given in
the Experimental Section. Selected X-ray data for 1,2 2,
3(P21/c), 3(Pca21), and 43d are given in Table 1, and
complete data are collected in the Supporting Informa-
tion. Selected X-ray bond lengths, bond angles and
torsional angles for these compounds are listed in Tables
2-4, respectively. It should be noted that the X-ray
structure of 1 was determined from a low-temperature
data set while all of the other data collections were at
ambient temperature. Hydrogen atoms were placed at
idealized positions.
(10) Frisch; M. J .; Trucks, G. W.; Head-Gordon, M.; Gill, P. M. W.;
Wong, M. W.; Foresman, J . B.; J ohnson, B. G.; Schlegel, H. B.; Robb,
M. A.; Replogle, E. S.; Gomperts, R.; Andres, J . L.; Raghavachari, K.;
Binkley, J . S.; Gonzalez, C.; Martin, R. L.; Fox, D. J .; Defrees, D. J .;
Baker, J .; Stewart, J . J . P.; Pople, J . A. GAUSSIAN92, Revision A;
Gaussian, Inc., Pittsburgh, PA, 1992.
(11) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, B. G.; Robb, M. A.; Cheeseman, J . R.; Keith, T. A.; Petersson,
G. A.; Montgomery, J . A.; Raghavachari, K.; Al-Laham, M. A.;
Zakrzewski, V. G.; Ortiz, J . V.; Foresman, J . B.; Cioslowski, J .;
Stefanov, B. B.; Nanyakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andres, J . L.; Replogle, E. S.; Gomperts,
R.; Martin, R. L.; Fox, D. J .; Binkley, J . S.; Defrees, D. J .; Baker, J .;
Stewart, J . P.; Head-Gordon, M.; Gonzalez, C.; Pople, J . A. GAUSSIAN-
94W (Revision 4.1); Gaussian, Inc., Pittsburgh, PA, 1995.
(12) Hehre, W. J .; Ditchfield, R.; Pople, J . A. J . Chem. Phys. 1972,
56, 2257.
The molecular structures of 1-8 were optimized at the
Hartree-Fock level using GAUSSIAN9210 or GAUSSIAN-
94W11 and the 6-31G12 basis set and were checked by
analytical frequency analysis. Optimization of 2 using the
6-31G*13 basis set showed only minor changes in struc-
ture and polarity (dipole moment). Selected bond lengths,
(13) Hariharan, P. C.; Pople, J . A. Chem Phys Lett. 1972, 16, 217;
Hariharan, P. C.; Pople, J . A. Theor. Chim Acta 1973, 28, 213.
(14) Nicoud, J . F.; Twieg, R. J . Nonlinear Optical Properties of
Organic Molecules and Crystals; Chemla, D. S., Zyss, J ., Eds.; Academic
Press: New York, 1987; vol. 2, p 224 and references therein.

Effects of C6 Substitution in 6-(4-(Dimethylamino)phenyl)fulvenes
J . Org. Chem., Vol. 64, No. 25, 1999 9069
Ta ble 2. Selected Bon d Len gth s for th e X-r a y Str u ctu r es of 1-4a
bondsb
Dbl ) C1C2, C3C4
1Ac
1Bc
2
3(P21/c)
3(Pca21)
4
1.355(2)
1.461(2)
1.364(3)
1.451(3)
1.354(2)
1.467(2)
1.362(2)
1.456(2)
1.338(5)
1.458(2)
1.366(5)
1.465(5)
1.334(3)
1.454(2)
1.363(3)
1.472(3)
1.482(3)
1.370(2)
1.398(2)
1.379(2)
1.380(2)
1.372(2)
0.120
1.36(1)
1.44(1)
1.37(1)
1.46(1)
1.49(1)
1.39(1)
1.40(1)
1.37(1)
1.37(1)
1.40(1)
1.347(2)
1.448(2)
1.366(3)
1.472(2)
1.474(2)
1.377(2)
1.401(1)
1.376(2)
1.396(2)
1.369(2)d
0.101
Sngl ) C5C1, C2C3, C4C5
C5C6
C6C7
C6C13
b ) C8C9, C11C12
a ) C7C8, C9C10, C10C11, C12C7
b ) C14C15, C17C18
a ) C13C14, C15C16, C16C17, C18C13
1.372(2)
1.412(1)
1.377(2)
1.413(1)
1.377(4)
1.398(3)
C
10N
1.373(2)
0.106
0.040
1.376(2)
0.113
0.036
1.368(3)
0.120
0.021
∆(Sngl - Dbl)
∆(a - b)
0.028
0.01
0.024
0.020e
a
b
In Å. The value given where more than one bond is listed is the average value. c There are two crystallographically independent
molecules. Both CN bonds are the same length. e Value for the second p-(dimethylamino)phenyl ring.
d
Ta ble 3. Selected Bon d An gles for th e X-r a y Str u ctu r es
Ch a r t 1
of 1-4a
bond
angle
3-
3-
1Ab
1Bb
2
(P21/c) (Pca21)
4
C1C5C6
C4C5C6
C5C6C7
C5C6C13
C1C5C4
C6C13C14
C6C7C8
132.6(2) 132.9(2) 128.4(2) 126.8(2) 130(1) 127.7(2)
121.9(2) 121.8(2) 126.1(3) 128.3(2) 123(1) 127.1(2)
131.9(2) 132.5(2) 123.4(2) 122.9(2) 124(1) 121.7(2)
120.3(3) 121.0(2) 120(1) 122.8(2)
105.5(2) 105.2(2) 105.4(3) 104.9(2) 107(1) 105.1(2)
120.9(2) 119(1) 123.4(2)
125.8(2) 127.0(2) 121.9(2) 121.0(2) 123(1) 122.9(2)
C8C7C12 116.5(2) 116.1(2) 115.7(2) 116.5(2) 115(1) 115.9(2)
Molecu la r Str u ctu r es. Important structural param-
eters for HF/6-31G geometry-optimized structures and
for X-ray structures of 6-(4-(dimethylamino)phenyl)-
fulvenes 1-4 as well as for 6-phenylfulvenes 6-8 are
given in Tables 2-7. In each case, the bonds from the
central carbon atom (C6) to the attached rings are twisted
in the same sense so that 6,6-diarylfulvenes 3, 4, and 8
assume a propeller-like shape. (See Chart 1 for number-
ing.) The largest parameter changes are observed for the
average torsional angle at the C6C7 bond (ω(C6C7)),16
which varies over a range of nearly 30° in the calculated
structures and over 40° in the solid state (Tables 7 and
4, respectively) and represent a measure of the twisting
between the (dimethyamino)phenyl ring and the fulvene
ring. The value of ω(C6C7) increases in the order 6 < 8 )
3 < 7 for compounds with a 6-phenyl group and in the
order 1 < 3 ) 4 < 2 for those with a 6-p-(dimethylamino)-
phenyl group. Thus, the “steric size” of the C6 substituent
increases in the order H < Ph ) C6H4NMe2 < Me. We
attribute the slightly smaller steric size of phenyl and
(dimethylamino)phenyl relative to methyl in these com-
pounds to the ability of adjacent aryl groups to rotate in
the same sense in order to minimize repulsive interac-
tions.
Comparison of ω(C6C7) for phenyl in 6, 7, 3, and 8 with
(dimethylamino)phenyl in 1, 2, 3, and 4, respectively
(replacement of phenyl with (dimethylamino)phenyl),
indicates that the latter ring is less twisted by 5-7° in
the calculated structures and by 3-7° in the X-ray
structures. This is strong evidence for a greater resonance
effect for (dimethylamino)phenyl relative to phenyl.
Sardella et al. previously concluded that there is rela-
tively little resonance interaction in 6-aryl-6-methylful-
venes on the basis of 13C chemical shifts.4c However, close
examination of their data reveals a significant contribu-
tion for resonance.4c
116.4(2)c
C9C10C11 117.4(2) 117.3(2) 116.6(3) 116.6(3) 119(1) 117.0(2)
116.6(2)d
Te
10.7
11.1
2.3
-1.5
7
0.6
a
b
In degrees. There are two crystallographically independent
c
d
molecules.
C14C13C18
.
C15C16C17.
e Reference 17.
Ta ble 4. Selected Tor sion a l An gles for th e X-r a y
Str u ctu r es of 1-4a
torsional angle 1Ab 1Bb
2
3(P21/c) 3(Pca21)
4
ωC5C6
ωC6C7
ωC6C13
ωC10N1
3.1
2.5
7.2
11.9
37.4
49.2
4.2
10
28
60
10
11.0
36.0
45.3
2.4
13.2 13.8 32.7
7.0 3.5 3.7
7.6c
a
In degrees; the parameters ω(C6C7), ω(C5C6), ω(C6C13), ω(C10N),
and ω(C16N) are defined as the average of the torsional angles,
excluding those involving hydrogen, at a given bond. Angles larger
b
than 90° were subtracted from 180° before averaging. There are
two crystallographically independent molecules. c Value for the
second p-(dimethylamino)phenyl ring.
retrieved from the Cambridge Structural Database
(CSD).15 Comparison of compounds with different sub-
stituents at C6 and of phenyl vs p-(dimethylamino)phenyl
substitution provides insight into the relative contribu-
tions of steric and π delocalization effects.
We next analyze the structural changes that occur on
incorporating 1-4 into a crystal lattice by comparing
calculated structures with X-ray structures. We also have
been able to obtain and structurally characterize two
polymorphs of 3. Comparison of these polymorphs and,
more generally, of compounds possessing similar or
different crystal packing motifs provides additional in-
formation about the role of the crystal lattice. In the final
section, we qualitatively relate the solid-state structures
of these compounds with SHG, a potential technologically
important property of these compounds.
(16) The parameters ω(C6C7), ω(C5C6), and ω(C6C13) are defined as
the average of the torsional angles, excluding those involving hydrogen,
at a given bond. Angles larger than 90° were subtracted from 180°
before averaging.
(15) 3D Search and Research Using the Cambridge Structural
Database. Allen, F. H.; Kennard, O. Chem. Design Automation News
1993, 8, 1, 31.

9070 J . Org. Chem., Vol. 64, No. 25, 1999
Ta ble 5. Selected Bon d Len gth s for HF /6-31G-Op tim ized Str u ctu r es of 1-8a
Peterson et al.
bondsb
1
2
3
4
5
6
7
8
Dbl ) C1C2, C3C4
1.3431
1.4735
1.3420
1.3429
1.4746
1.3462
1.3434
1.4743
1.3507
1.3450
1.4721
1.3540
1.3437
1.47180
1.3433
1.3588c
1.4484
1.4634d
1.3412
1.4765
1.3384
1.3417
1.4764
1.3428
1.3420
1.4765
1.3474
Sngl ) C5C1, C2C3, C4C5
C5C6
C6C7
1.4651
1.4882
1.4851
1.4861
1.4733
1.4937
1.4922
C6C13
b ) C8C9, C11C12
1.4931
1.3815
1.3995
1.3866
1.3915
1.3821
0.1309
0.0181
0.0049e
1.4861
1.3818
1.3994
1.3817
1.3994
1.3829
1.1271
0.0176
0.0176e
1.4922
1.3865
1.3915
1.3865
1.3915
1.3805
1.4003
1.3823
1.3985
1.3801
1.4002
1.3858
1.3921
1.3868
1.3910
a ) C7C8, C9C10, C10C11, C12C7
b ) C14C15, C17C18
a ) C13C14, C15C16, C16C17, C18C13
C
10N
1.3814
0.1304
0.0198
1.3836
0.1317
0.0162
∆(Sngl - Dbl)
∆(a - b)
0.1281
0.0200
0.1353
0.0063
0.1347
0.0041
0.1345
0.0050
0.0050f
a
b
d
In Å. The value given where more than one bond is listed is the average value. c Length of the ethendiyl bond. Length of the single
bond between the ethendiyl group and the aryl ring. e Values for phenyl ring. f Value for the second p-(dimethylamino)phenyl ring.
Ta ble 6. Selected Bon d An gles for HF /6-31G-Op tim ized
Ta ble 8. Ca lcu la ted Dip ole Mom en ts for 1-4
Str u ctu r es of 1-8a
dipole momenta
bond
angle
compd
optimized structureb X-ray structurec differencea,d
1
2
3
4
5
6
7
8
1A
1B
2
3(P21/c)
3(Pca21)
4
3.94
3.94
3.66
3.76
3.76
4.63
5.07
5.09
4.99
4.38
4.32
5.76
1.13
1.15
1.33
0.62
0.56
1.13
C1C5C6
C4C5C6
C5C6C7
C5C6C13
C1C5C4
C6C13C14
C6C7C8
130.7 127.4 127.6 127.5
124.1 127.8 127.4 127.5
129.4 122.4 122.5 121.8
122.5 121.3 121.8
129.1 130.3 127.3 127.5
125.4 124.5 127.5 127.5
126.6b 128.4 122.0 122.0
123.2 122.0
105.4 105.2 104.9 105.0
120.3
105.2 104.8 105.0 105.0
121.1 121.2
123.8 121.5 122.0 121.92 122.4c 122.6 120.8 121.2
a
b
d
In D. HF/6-31G//HF/6-31G. c HF/6-31G//X-ray. µ(X-ray) -
C8C7C12 116.8 116.8 116.9 116.92 116.6 118.4 118.5 118.5
µ(optimized).
116.92d
C9C10C11 117.2 117.1 117.2 117.22 117.1 119.6 119.6 119.6
by the solid state in general but also by the specific
crystal packing motif that is employed.
117.22e
Tf
6.6 -0.4
0.2
0.0
3.7
5.8 -0.2
0.0
Another parameter that can be employed to assess
relative steric size is T ) C1C5C6 - C4C5C6 (Tables
3 and 5), which is a measure of the “tilt” of the exocyclic
fulvene double bond toward the smaller of the two
substituents at C6.17 When the second substituent at C6
a
b
In degrees. Angle formed by atoms C5, C6 and the ethendiyl
carbon attached to C6. c Angle formed by the ethendiyl carbon
atoms and C7.
d
e
C14C13C18
.
C15C16C17.
f Reference 17.
Ta ble 7. Selected Tor sion a l An gles for HF /
6-31G-Op tim ized Str u ctu r es of 1-8a
is phenyl (3, 6-8) or (dimethylamino)phenyl (1-4),
T
torsional angle
1
2
3
4
5
6
7
8
in the calculated structures increases in the order H <
Ph e C6H4NMe2 < Me. This is the same order as
determined from ω(C6C7). In contrast, T in the X-ray
structures of 1-4 increases in the order H < Me < Ph <
C6H4NMe2 when the second substituent at C6 is (di-
methylamino)phenyl. This occurs in response to the
above-mentioned “flattening” of the (dimethylamino)-
phenyl ring, which causes an increase in its steric size.
The torsion at the C5C6 bond (ω(C5C6))16 increases in
the order 1 < 2 < 3 < 4 (H < Me < Ph < C6H4NMe2)
(Tables 4 and 7). Note that here also the effect of methyl
is less than that of aryl, in contrast to the order for ω-
(C6C7). We suggest that this is due to a greater stabiliza-
tion by aryl (relative to methyl) of resonance forms B and
C, which contribute single bond character to the C5C6
double bond. Strong evidence for the role of π delocal-
ization in these compounds is also given by (a) a very
good linear correlation (r2 ) 0.91) for a plot of the C6C7
bond length (r(C6C7)) vs ω(C6C7) (Figure 1), as would be
expected from a contribution of through-resonance form
B and (b) the NMR chemical shifts of C6, which increase
in the order 1 (139.9 in CDCl3)2 < 2 (150.6 in CDCl3) <
3 (153.4 in THF-d8), in accord with an increased contri-
bution of π polarization (resonance form C).
ωC5C6
ωC6C7
ωC6C13
ωC10N1
ωC16N2
4.4b
5.3c 9.9 11.4 0.0
3.5
3.4
8.4
30.4b 51.8c 45.3 45.5 0.0 37.0 58.9 50.0
50.2 45.5
0.6 0.3 0.0
0.5 0.0
50.0
0.9
0.6
a
In degrees; the parameters ω(C6C7), ω(C5C6), ω(C6C13), ω(C10N),
and ω(C16N) are defined as the average of the torsional angles,
excluding those involving hydrogen, at a given bond. Angles larger
b
than 90° were subtracted from 180° before averaging. Values of
ω(C5C6) and ω(C6C7) of 2.6° and 33.2°, respectively, were calculated
for 1 at HF/STO-3G (ref 2). Values of ω(C5C6) and ω(C6C7) of 13.1°
and 36.0°, respectively, were calculated for 2 at a DFT/J MW/double
numeric plus polarization level (ref 6c). However, these values
were not confirmed by analytical frequency analysis.
Comparison of the calculated structures for 3 and 8
with the corresponding X-ray structures indicates that
the twist of the phenyl group is essentially unchanged
on incorporation into the solid state. In contrast, the twist
of the (dimethylamino)phenyl group decreases in the solid
state by 17-19° in 1, 2, and 3(Pca21) and by 8-10° in
3(P21/c) and 4. These results strongly suggest an en-
hanced π-delocalization in the solid state. This is sup-
ported by HF/6-31G calculations of the dipole moments
(µ) for 1-4 at the optimized geometries and at the X-ray
structures (Table 8). In each case, µ increases by 0.56-
1.32 D on going to the X-ray structure. The crystal lattice
is expected to cause additional polarization. As discussed
in the next section, these changes are influenced not only
The value of r(C6C7) decreases by ca. 0.007 Å in the
“gas phase” and by 0.1-0.2 Å in the crystal on changing
(17) Bock, C. W.; Trachtman, M.; George, P. THEOCHEM 1985, 122,
155.

Effects of C6 Substitution in 6-(4-(Dimethylamino)phenyl)fulvenes
J . Org. Chem., Vol. 64, No. 25, 1999 9071
Ta ble 10. X-r a y Bon d An gles in P h en ylen e Rin gsa
compd
R
δ
compd
R
δ
∆Rb
∆δb
1
2
116.3 117.3
115.7 116.9
8d
9e
118.6 119.9
121.1 118.7 -0.5 -0.5
3(P21/c) 116.5 116.9
10f
11g
12h
118.3 117.1
119.2 118.6
116.7 118.1
1.1
0.6
1.0
0.4
0.1
0.4
3(P21/c)c 118.0 119.7
4
116.2 116.8
a
Angles R and â (defined in structure 9); in degrees. b Difference
between the predicted22a and average X-ray bond angles. c Values
d
for the phenyl ring in 3(P21/c). Reference 3b; no atomic coordi-
nates were available for the structure reported in ref 3c. e Average
values for three structures (ref 18). f Average values for five
g
structures (ref 19). Average values for two structures (ref 19a,
20). Reference 21.
F igu r e 1. Plot of ω(C6C7) vs r(C6C7) or ω(C6C13) vs r(C6C13
)
h
for calculated structures (9) of 1-8 and X-ray structures (])
of 1-4 (r2 ) 0.91). In cases where both phenyl and p-
(dimethylamino)phenyl substituents are present (3, 3(P21/c),
and 3(Pca21)), the labels Ph and Ar, respectively, are employed.
The angles R and δ (see 9) have previously been
employed to evaluate the effect of substituents on sub-
stituted benzenes in the solid state. Domenicano and co-
workers have compiled an extensive list of “angular
substituent parameters” (ASPs) that allow one to predict
the ability of a substituent to distort CCC angles from
the benzene value of 120°.22 These angles are primarily
determined by changes in the σ bonds due to the effect
of the electronegativity of the substituent on the hybrid-
ization of the ring carbon to which it is bonded. However,
as recognized by these workers, the contribution of
π-resonance effects will tend to shorten the C1-substitu-
ent bond and decrease R for both donors and acceptors.23
Thus, the deviation of the measured value of R from the
value predicted from ASPs can be taken to be an
indication of the contribution of through resonance to the
structure of a given para-disubstituted benzene.
The deviations of the average X-ray values of R and δ
in 1-4 (for C6H4NMe2), 3 and 8 (for Ph) and in D/A-
substituted benzenes 9-12 from the values predicted
from the ASPs are given in Table 10. Note that the
deviations for phenyl are small and positive (i.e., in the
opposite direction expected for resonance). In contrast,
p-(dimethylamino)phenyl shows a negative deviation, but
one that is half or less than those shown by strong D/A-
benzenes 9-12. This suggests that 6-(4-(dimethylamino)-
phenyl)fulvenes have significant through-resonance con-
tributions, but less than found in 9-12. The twist
between the aryl and fulvene rings in 1-4 undoubtedly
accounts, at least in part, for the smaller ASP deviations
found in these compounds.
An analysis of the effects on molecular structure
brought about by incorporation into a crystal lattice was
previously presented for 1 (which had been optimized at
the HF/STO-3G level), and the following changes were
noted.2 (1) The interplanar angle between the fulvene and
aryl rings is reduced in the crystal with respect to its
calculated value. (2) The pyramidalization of the di-
methylamino group is largely (but not completely) elimi-
nated in the X-ray structure. (This is not true for the
compounds discussed here since the dimethyamino group
is not significantly pyramidalized in the structures
optimized using the 6-31G basis set.) (3) The “tilt angle”
( T) is increased. (4) The length of the C5C6 bond
(r(C5C6)) is increased while r(C6C7) is reduced by a similar
amount. (5) The average single bond length in the fulvene
ring is reduced in the crystal. (6) The bond lengths of
Ta ble 9. Aver a ge X-r a y Bon d Len gth s of P h en ylen e
Bon d s a a n d b a n d Th eir Differ en cea
compd
ra
rb
∆ (ra - rb) compd
ra
rb
∆ (ra - rb)
1
2
1.413 1.375
1.398 1.377
0.038
0.021
0.028
0.024
9b
10c
11d
12e
1.401 1.374
1.391 1.372
1.401 1.372
1.396 1.369
0.027
0.019
0.029
0.027
3(P21/c) 1.398 1.370
1.401 1.377
4
a
b
In Å; bonds a and b are defined in structure 9. Average
values for three structures (ref 18). c Average values for five
d
structures (ref 19). Average values for two structures (ref 19a,
20). e Reference 21.
from Ph to C6H4NMe2 (Tables 5 and 2, respectively),
again reflecting the greater resonance effect of the latter.
Further, r(C6C7) for C6H4NMe2 decreases by 0.011-0.023
Å in 1-4 on incorporation into the crystal, in contrast to
changes of -0.011 and +0.002 Å for Ph in 3 and 8,
respectively. This indicates that C6H4NMe2 is more
polarized in the solid state (relative to the gas phase)
than is Ph.
Another indicator of resonance is provided by the
quinoid character of the C6 aryl ring (as illustrated by
structures B and C). The difference (for X-ray structures)
between the average length of the “end” bonds (a) and
the “side” bonds (b) (see 9 and Table 9) is 0.038, 0.021,
0.028, and 0.024 Å for 1, 2, 3(P21/c) and 4, respectively.
This is about the same magnitude as found for the
average difference (0.026 Å) in four strong donor/strong
acceptor (D/A)-substituted benzenes, 9,18 10,19 11,19a,20
and 12.21
(18) (a) Tonogaki, M.; Kawata, T.; Ohba, S.; Iwata, Y.; Shibuya, I.
Acta Crystallogr. 1993, B49, 1031. (b) Colapietro, M.; Domenicano, A.;
Marciante, C.; Portalone, G. Z. Naturforsch. 1982, 37B, 1309. (c)
Trueblood, K. N.; Goldish, E.; Donohue, J . Acta Crystallogr. 1961, 14,
1009.
(19) (a) J ameson, G. B.; Sheikh-Ali, B. M.; Weiss, R. G. Acta
Crystallogr. 1994, B50, 703. (b) Heine, A.; Herbst-Irmer, R.; Stalke,
D.; Ku¨hnle, W.; Zachariasse, K. A. Acta Crystallogr. 1994, B50, 363.
(c) Gourdon, A.; Launay, J .-P.; Bujoli-Doeuff, M.; Heisel, F.; Miehe´, J .
A.; Amouyal, E.; Boillot, M.-L. J . Photochem. Photobiol. A: Chem. 1993,
71, 13.
(22) (a) Domenicano, A. Methods Stereochem. Anal. 1988, 10 (Ste-
reochem. Appl. Gas-Phase Electron Diff., Part B), 281. (b) Domenicano,
A.; Murray-Rust, P. Tetrahedron Lett. 1979, 2283.
(23) Domenicano, A.; Viciago, A.; Coulson, C. A. Acta Crystallogr
1975, B31, 221.
(20) Merlino, S.; Sartori, F. Acta Crystallogr. 1982, B38, 1476.
(21) Haisa, M.; Kashino, S.; Yuasa, T.; Akigawa, K. Acta Crystallogr.
1976, B32, 1326.

9072 J . Org. Chem., Vol. 64, No. 25, 1999
Peterson et al.
the (dimethylamino)phenyl ring change to reflect a more
quinoid shape for the ring (as in B and C). (7) The bonds
to N, particularly C10N, are shortened. It can be seen in
Tables 2-7 that all of these trends, except point 2 as
explained above, are also present in 2, 3(Pca21), and 4
when the X-ray structures are compared with the corre-
sponding HF/6-31G-optimized structures.
Cr ysta l P a ck in g. In most crystalline aromatic com-
pounds lacking strong hydrogen bond donors and accep-
tors, the major intermolecular binding forces are disper-
sion/electrostatic in nature. The latter are commonly
manifested as edge-to-face interactions between the
hydrogen periphery of one molecule and the π-cloud of
an adjacent molecule.24 It has been shown statistically
that the magnitude of the molecular dipole moment is
not related to bulk centrosymmetry (or noncentrosym-
metry) in the crystalline form.25 It has also been noted
that only slight misalignments of adjacent dipoles are
necessary to cause noncentrosymmetry.26 These mis-
alignments probably do not have a large influence on the
nature of the basic intermolecular interactions. The
molecules are still oriented so that their dipole moments
are aligned in nearly opposite directions and the impor-
tant electrostatic (edge-to-face) interactions are basically
the same. A study of the relationship between local dipole
moments and head-to-tail packing arrangements showed
that the latter become more important as the local dipole
moments are brought closer together.26
F igu r e 2. Packing motifs for 1-4.
We have previously discussed the relationship of
dipole-dipole alignment to crystal packing in 1.2 The
molecules align in a herringbone fashion with molecular
dipoles in adjacent layers pointing in opposite directions.
By packing in this way, molecular dipole moments are
enhanced through interactions with the dipole moments
of neighboring molecules. Both ωC5C6 and ωC6C7 are
smaller (by ∼1.5° and ∼16.8°, respectively) in the crystal
then in the optimized structures, suggesting that µ is
increased in the crystal.
A. 6-(4-(Dim eth yla m in o)p h en yl)-6-m eth ylfu lven e
(2). Compound 2 packs in space group P212121 with four
molecules per asymmetric unit. The molecules pack so
that the local dipole moments of the aryl rings are aligned
in a head-to-tail arrangement (type I, Figures 2 and 3)
with close contacts of 3.2 Å between H9 and C10 or N.
The aryl rings packing in this manner form a pentamer
with the angle between the central aryl ring and the
surrounding aryl rings being 58°. Two of the neighboring
molecules are offset more then the other two, the
distances between the centroids of the central aryl ring
and the corresponding centroids of the two pairs of
neighbors being 5.2 and 5.7 Å. The shorter centroid-
centroid distance involves the more offset aryl rings. This
arrangement leads to stacks of molecules along the b axis
with interleaved aryl rings, the distance between the
stacks being 7.67 Å, approximately the same as the sum
of the van der Waals radii for a methyl group and a
phenylene ring (7.7 Å27). Thus, the alignments of the local
dipole moments along the C10-N bond could enhance the
donation of π-electron density from the nitrogen to the
aryl ring.
F igu r e 3. Packing diagram for 2.
The fulvene ring in 2 also packs in a head-to-tail
manner (type III, Figure 3) with respect to neighboring
fulvene rings (intermolecular interplanar angle between
fulvene rings ) 75.6°). The contacts between fulvene
rings are edge-to-face in nature with C1‚‚‚H4, C2‚‚‚H4, and
C5‚‚‚H4 having intermolecular distances of 2.8, 3.0, and
3.2 Å, respectively.28 This arrangement aligns the dipole
moments of adjacent fulvene rings in opposite directions.
These two types of head-to-tail intermolecular interac-
tions between local dipoles act in concert to polarize the
molecule in the crystal and contribute to a decrease in
ωC6C7 of 19.1° relative to that in the optimized structure.
The fact that ωC5C6 is 1.9° larger in the crystal then in
(24) (a) Hunter, C. A. Chem. Soc. Rev. 1994, 101. (b) Hunter, C. A.;
Sanders, J . K. M. J . Am. Chem. Soc. 1990, 112, 5525.
(25) Whitesell, J . K.; Davis, R. E.; Saunders, L. L.; Wilson, R. J .;
Feagins, J . P. J . Am. Chem. Soc. 1991, 113, 3267.
(26) Gavezzotti, A. J . Phys. Chem. 1990, 94, 4319.
(27) Dunitz, J . D. X-ray Analysis and the Structure of Organic
Molecules; VCH: New York, 1995; p 339.
(28) It was recently pointed out that intermolecular interactions
need not be energetically stabilizing. The intermolecular interactions
shown in italics may be destabilizing: Dunitz, J . D.; Gavezzotti, A.
Acc. Chem. Res. 1999, 32, 677.

Effects of C6 Substitution in 6-(4-(Dimethylamino)phenyl)fulvenes
J . Org. Chem., Vol. 64, No. 25, 1999 9073
F igu r e 5. Packing diagram for 3(Pca21).
F igu r e 4. Packing diagram for 3(P21/c).
the optimized structure of 2 is indicative of a greater
importance of resonance forms B and/or C in the crystal
and/or an increase in steric effects arising from the
decrease in ω(C6C7). Increased molecular polarization in
the crystal was confirmed by an increase in µ (Table 8)
as well as the π charges at C1-6 (Supporting Information)
calculated at the HF/6-31G level and the X-ray geometry
(HF/6-31G//X-ray).
B. 6-(4-(Dim eth yla m in o)p h en yl)-6-p h en ylfu lven e
(3(P 21/c)). Several different packing interactions are
apparent in this compound (Figure 4). There are π-to-π
stacks between phenyl and fulvene rings (type VI, Figure
2) where the phenyl ring of one molecule sits over C1 and
C5-C8 of a neighboring molecule. H20 is involved in
intermolecular close contacts with C1, C5, and C6 of 2.8,
3.2, and 3.3 Å, respectively,28 and the phenyl and fulvene
rings are inclined by 21.7° degrees with respect to one
another. Since the two molecules are related by a center
of symmetry, the molecular µ’s are parallel and aligned
in opposite directions. In addition to the π-to-π stacks,
the fulvene ring of each molecule is sandwiched between
the dimethylamino groups of two neighboring molecules
such that the molecular dipole moments align in a
herringbone motif (type V, Figure 2). One methyl hydro-
gen (H19) is involved in three intermolecular close con-
tacts with C3, C4 and C5 of 3.1, 3.2, and 3.2 Å, respec-
tively, and another (H20) is involved in two intermolecular
close contacts with C1 and C5 of 2.8 and 3.2 Å, respec-
tively.28 This arrangement places the positive end of one
molecular dipole moment near the negative end of a
neighbor. There is one head-to-tail (type I, Figure 2)
interaction between aryl rings parallel to each other with
a C10‚‚‚N distance of 3.65 Å.
C. 6-(4-(Dim eth yla m in o)p h en yl)-6-p h en ylfu lven e
(3(P ca 21)). The packing here appears to be dominated
by type-I interactions. The nitrogen on one aryl group
sits over a neighboring aryl ring such that the local
dipoles of the (dimethylamino)phenyl rings are aligned
antiparallel to one another (Figure 5). H20 is in close
contact (2.8 Å)28 with both C1 and C2. The molecules
packed in this way form bilayers in the bc plane. The
planes of the aryl rings are tilted by 55° with respect to
one another and are also offset from one another.
By packing into bilayers, each aryl ring is surrounded
by four neighboring aryl rings aligned in antiparallel
directions similar to the packing motif found in 2. The
molecular bilayers are stacked with respect to a neigh-
boring bilayer so that the fulvene and phenyl rings of
one molecule are aligned head-to-head with those of a
F igu r e 6. Packing diagram for 4.
neighboring bilayer (type IV, Figure 2). The neighboring
fulvene rings are inclined by 6.1° relative to one another
while the intermolecular interplanar angle between
neighboring phenyl rings is 75.1°.
The most notable difference between the packing of
polymorphs 3(P21/c) and 3(Pca21) is that there is only one
type-I interaction per molecule in the former compared
to four such interactions in the latter. This difference is
undoubtedly related to the differences in molecular
conformation between the two polymorphs. The value of
ω(C6C7) is 10° smaller in 3(Pca21) than in 3(P21/c),
suggesting that the local µ of the former is larger.
D. 6,6-Bis(4-(d im eth yla m in o)p h en yl)fu lven e (4).3d
Another indication of the influence of crystal packing on
the molecular structure of these compounds can be seen
in 4. The two aryl rings in the calculated structure of 4
are equivalent. However, in the crystal one aryl ring is
more twisted relative to the fulvene ring then the other
(Table 7). The more coplanar aryl ring has one type I
interaction with the more coplanar aryl ring of an
adjacent molecule in the crystal (Figure 6). The inter-
molecular interplanar angle in this case is 0.0°. The other
aryl ring (larger ωC6CAr) is involved in a type-II interac-
tion (Figure 2) with the hydrogens of one aryl ring lying
over the π-system of the adjacent aryl ring (aryl-aryl
interplanar angle ) 51.6°) and offset so that one of the
dimethylamino methyls sits over the fulvene π-system
of a neighboring molecule.
E. Effect of Cr ysta l P a ck in g on Molecu la r Str u c-
tu r e. In all cases, ω(C6C7) is smaller in the crystal than
in the optimized structures, by 17-19° in 1, 2, and
3(Pca21) and by 8-10° in 3(P21/c) and 4, a clear indication
that the crystalline environment is more polarizing than
the “gas phase”. The type-I alignment of aryl rings
appears to be a more polarizing intermolecular interac-
tion then the other types of packing arrangements
encountered in 1-4. The number of type-I interactions
is also important. Thus, the four such interactions in

9074 J . Org. Chem., Vol. 64, No. 25, 1999
Peterson et al.
Ta ble 11. Mu llik en Ch a r ges for Mon om er a n d P en ta m er Str u ctu r es of 2(P 212121) a n d 3(P ca 21)a
F(2(P212121))
monomer
F(2(P212121))
∆F(pentamerb
-
F(3(Pca21))
monomer
F(3(Pca21))
∆F(pentamerb
-
atom
pentamerb
monomer)
pentamerb
monomer)
N
-0.264
-0.131
0.070
-0.264
-0.145
0.072
-0.000
-0.014
0.003
-0.263
-0.119
0.040
-0.262
-0.129
0.041
0.001
-0.010
0.001
Σ(C1-C5)
C6
Σ(C1-C6)
Σ(C7-C12
C13
-0.062
-0.073
-0.012
-0.004
0.001
-0.079
0.076
-0.088
0.070
-0.009
-0.006
)
0.061
0.057
0.004
0.005
Σ(C13-C18
)
0.015
-0.013
0.079
0.015
0.004
0.089
0.001
-0.001
0.016
0.096
0.000
Σ(NMe2)
-0.004
0.057
-0.001
0.011
0.170
0.000
0.014
0.018
-0.001
Σ(C7-NMe2)
c
Ftot
-0.001
a
b
Charges include those of any attached hydrogen atoms. These values are for the central molecule of the pentamer. c Total molecular
charge.
3(Pca21) are more polarizing then the single interaction
trosymmetric crystal packing does not guarantee SHG
since molecular hyperpolarizabilities may accidentally
cancel one another.29
in 3(P21/c), as evidenced by a smaller value of ωC6C7 in
the former. Note that there are also four antiparallel
intermolecular interactions in the X-ray structure of 2.
To investigate whether type-A packing interactions
between aryl rings do, in fact, shift electron density into
the fulvene ring in 2 and 3(Pca21), we calculated Mulliken
charges at the HF/STO-3G level for both a single mol-
ecule and a pentamer consisting of a molecule surrounded
by four neighbors arranged exactly as in the crystal
lattice and with molecular geometries taken from the
crystal structure. The Mulliken charges calculated for the
monomer were then subtracted from the corresponding
charges for the central molecule of the pentamer (Table
11). The total Mulliken electron density in the fulvene
moiety (C1-C6) of the central molecule of the pentamer
is increased by 0.0116 and 0.0093 electrons for 2 and
3(Pca21), respectively, relative to the monomer. Thus,
electron density is shifted from the (dimethylamino)-
phenyl group into the fulvene ring of the central molecule
of the pentamer with only a minimal basis set calculation
and without any significant intermolecular charge trans-
fer. Stabilization of the molecules in the pentamers of 2
and 3(Pca21) of 0.63 and 0.89 kcal/mol, respectively, is
found on comparing the total molecular energy of the
monomers with the average total molecular energies for
the corresponding pentamers.
SHG was observed qualitatively in 2, 3(Pca21), and 5.
We made no attempt to quantify the NLO response in
these materials, but it appears that the response in 5 is
the largest and that in 3(Pca21) is the smallest. The
former result is in accord with previous measurements
of µâ.7a,d Unfortunately, we were unable to grow crystals
of 5 large enough to obtain a single crystal X-ray
structure. It is likely that the fulvene and aryl rings are
nearly coplanar based on comparison of the HF/6-31G-
optimized steric environment at C6 in 5 with that in 1.
We also measured the UV-vis spectra of 3 and 5 as
solutions in cyclohexane and as thin films cast from
solution. For 3, the long-wavelength absorption band
(λmax) appears at 393, 407, 404, 417-422, and 410-427
nm for a cyclohexane solution, benzene solution, metha-
nol solution, and for multiple films cast from benzene and
methanol, respectively.30 (Recall that 3(P21/c) and 3(Pca21)
were crystallized from benzene and methanol, respec-
tively.) The shift to longer wavelength in the films
relative to solution suggests less twist and an increase
in π delocalization in the former. When contrasted with
the greater value of λmax for 3 in benzene relative to
methanol, the small increase in λmax for some films of 3
cast from methanol compared to benzene might reflect
the observation that the fulvene and (dimethylamino)-
phenyl rings of 3(Pca21) are more nearly coplanar than
those of 3(P21/c). The later point is supported by ZINDO31
calculations using molecular geometries taken from the
crystal structures of 3(P21/c) and 3(Pca21), where λmax was
calculated to be 338 and 342 nm, respectively. No SHG
response was observed in any of the thin films.
It is of interest to consider the direction of polarization
in the noncentrosymmetric crystals of 2 and 3(Pca21). The
fulvene rings of all of the molecules of 2 are tilted up
along the b axis. In 3(Pca21), the fulvene rings are tilted
up along the c axis (Figures 3 and 5, respectively). We
suggest that the polarization directions are along the b
and c axes for 2 and 3(Pca21), respectively, based solely
on these structural observations.
NLO P r op er ties. Compounds 1 and 5 have been
shown to have exceptionally large molecular hyperpo-
larizability (µâ) values.7a,d Since a large molecular hy-
perpolarizability is a prerequisite for efficient second
harmonic generation (SHG) in bulk materials, 1-5
represent an interesting series of compounds for solid-
state studies.
Efficient SHG of bulk materials also depends on the
orientation of the molecules in the crystal with respect
to one another.14 If the molecules pack centrosymmetri-
cally, as observed for 1, 3(P21/c), and 4, the hyperpolar-
izabilities of the individual molecules cancel each other.
However, if there is a unique axis resulting from non-
centrosymmetric crystal packing, as in 2 and 3(Pca21),
then the hyperpolarizabilities of individual molecules
may reinforce one another and lead to SHG. Noncen-
Further, λmax for 5 appears at 418, 415, 429 and 433
nm for solutions in cyclohexane, hexanes, benzene and
a film cast from benzene, respectively.30 (We were unable
to cast a film of 5 from hexanes, the recrystallization
solvent.) Because the absorbance of these films is small
at 532 nm (the wavelength of the green light generated
by doubling the frequency of the infrared laser output at
1064 nm), these compounds might be attractive candi-
dates for NLO materials.
(29) Nicoud, J . F.; Twieg, R. J . Nonlinear Optical Properties of
Organic Molecules and Crystals; Chemla, D. S., Zyss, J ., Eds.; Academic
Press: New York, 1987; Vol. 1, pp 246-7.
(30) The films cast for 3 and 5 were not characterized other then
by UV-vis spectroscopy. The differences observed in λmax may be due
to differences in molecular order or to variations in film thickness.
(31) Zerner, M. C. Rev. Comput. Chem. 1991, 2, 313.

Effects of C6 Substitution in 6-(4-(Dimethylamino)phenyl)fulvenes
J . Org. Chem., Vol. 64, No. 25, 1999 9075
and concentrated to afford a red solid. The crude product was
chromatographed on silica gel with 3:1 methylene chloride/
hexanes to afford 150 mg (6%) of 5 as fine red needles, mp )
142-143 °C; TLC (3:1 methylene chloride/hexanes) Rf ) 0.70;
1H NMR (CDCl3) δ 7.5-6.2 (m, 11 H), 3.02 (s, 6 H); 13C NMR
(CDCl3) δ 150.9, 143.4, 141.5, 139.4, 131.8, 130.1, 128.9, 125.0,
124.6, 121.6, 118.1, 112.0, 111.5, 108.3, 40.2; IR (KBr) 2856,
1581, 1364 cm-1; UV-vis (cyclohexane) λmax ) 418 nm (ꢀ )
72 500). Anal. Calcd for C16H17N: C, 86.05; H, 7.67; N, 6.27.
Found: C, 85.75; H, 7.65; N, 6.24.
Su m m a r y
We have shown by comparison of geometry-optimized
and X-ray structures that 1-4 are polarized on incorpo-
ration into the crystal lattice. Head-to-tail alignment of
neighboring p-(dimethylamino)phenyl rings polarize π
charge into the fulvene moiety to a greater extent than
with other packing motifs. This effect is enhanced in the
pentamer packing arrangement in 2 and 3(Pca21) where
a central p-(dimethlamino)phenyl ring is surrounded by
four others aligned in the antiparallel direction. The
increased polarization caused by pentamer-type packing
is probably responsible for a decrease in ω(C6C7) in
3(Pca21) relative to that in 3(P21/c). Macroscopic nonlin-
ear quadratic susceptibility (SHG) is reported for 2,
3(Pca21) and 5, indicating the potential technological
importance of this class of compounds.
Cr ysta l Gr ow th . Compound 2 was crystallized as dark
orange square plates by slow evaporation of an isopropylamine
solution. The crystal used for the X-ray structure determina-
tion of 3(P21/c) was grown as orange blades by slow evapora-
tion of a solution in acetone. Polymorph 3(P21/c) was also
crystallized from benzene and isopropylamine and found to
have the same crystal habit and unit cell dimensions as the
crystals grown from acetone. Polymorph 3(Pca21) was grown
by very slow evaporation of a methanol solution. A saturated
solution of 3 in methanol was prepared and placed into a test
tube that was then capped with a rubber septum and a syringe
needle inserted. The test tube was left to stand for several
months during which time very fine needles no thicker than
0.1 mm were formed.
Exp er im en ta l Section
Gen er a l Meth od s. IR spectra were recorded from KBr
windows. UV-vis spectra were measured using spectropho-
tometric grade benzene and HPLC grade methanol and
hexanes from Fischer Chemical Co. NMR spectra were re-
corded at 300 MHz for 1H and 75 MHz for 13C. NMR solvents
were used as received from Cambridge Isotope Laboratories.
1H spectra were referenced to TMS, and 13C spectra were
referenced to the solvent peak. Melting points are uncorrected.
Elemental analyses were performed by Midwest Microlabs,
Indianapolis, IN.
Da ta Collection . Data for 2 and 3(P21/c) were collected on
a Siemens P3 diffractometer, in the ω scan mode with ω scan
widths ) 1°, variable ω scan speed 4-20 deg min-1 and
graphite-monochromated Mo-KR radiation (0.71073 Å). Data
for 3(Pca21) were collected on a Rigaku AFC5R diffractometer
in the ω/2 Å mode with ω scan widths ) 1°, variable ω scan
speed 4-20 deg min-1 and nickel-filtered Cu-KR radiation
(1.541 78 Å).
Syn th esis. 6-(4-(Dim eth yla m in o)p h en yl)-6-m eth ylfu l-
ven e (2) (mp 107-108 °C) was prepared by the method of
Gugelchuk et al.6c
6-(4-(Dim eth yla m in o)p h en yl)-6-p h en ylfu lven e (3).32 Po-
tassium hydroxide (1.883 g, 33.56 mmol) and 18-crown-6 (400
mg, 1.51 mmol) were dissolved in 50 mL of anhydrous THF
and stirred in a foil-covered 250-mL round-bottomed flask
under an argon atmosphere while 1.85 mL (1.5 g, 23 mmol) of
freshly distilled cyclopentadiene was added dropwise. The
reaction mixture was stirred for 30 min, and then 4.5 g (20
mmol) of 4-(dimethylamino)benzophenone in 15 mL of THF
was added dropwise and the reaction mixture was stirred at
room temperature for 22 h. The reaction mixture was then
quenched with 75 mL of 3 M HCl and extracted with 3 × 50
mL of ethyl ether, and the combined organic layers were
washed with 50 mL of brine and dried (Na2SO4). A solid black
residue obtained on removal of the solvent was crushed and
washed with ether. The crude product was then dissolved in
methylene chloride and chromatographed on 80 g of silica gel
with 5:1 pentane/methylene chloride. The product was crystal-
lized from the eluant to afford 96 mg (5%) of 3 as orange
crystals: mp 118-119.5 °C; 1H NMR (CDCl3) δ 7.3-7.5 (m, 5
H), 7.24 (m, 2 H), 6.69 (apparent d, 2 H, J ) 9.0 Hz), 6.63-
6.60 (m, 1 H), 6.57-6.54 (m, 1 H), 6.49-6.46 (m, 1 H), 6.22-
6.17 (m, 1 H), 3.03 (s, 6 H); 13C NMR (CDCl3) δ 153.4, 150.8,
141.9, 141.5, 134.2, 132.4, 130.9, 130.3, 128.8, 128.5, 127.4,
124.4, 124.1, 111.0, 40.2. IR (KBr) 2891, 1526, 1360 cm-1; UV-
vis (cyclohexane) λmax ) 393 nm (ꢀ ) 9700), 298 (5600), 267
(7500). Anal. Calcd for C20H19N: C, 87.87; H, 7.01. Found: C,
87.72; H, 7.04.
Str u ctu r e An a lysis a n d Refin em en t. All structures were
solved by direct methods. Full-matrix least-squares refine-
ments were carried out with all nonhydrogen atoms anisotropic
and hydrogens in calculated positions. Structure solutions and
refinements were performed on either a Silicon Graphics
workstation or a Gateway PC computer using SHELXS34 and
SHELXL35 in the Siemens crystal structure solution package.
Atomic positions for 1, 4, and 9-12 were obtained from the
CSD.15 The coordinate files were read into ORTEP-3.36 Packing
diagrams for 2, 3(P21/c), and 3(Pca21) were generated using
Ortep-3,36 whereas those for 4 were generated using PLUTO.37
Molecu la r Or bita l Ca lcu la tion s. Ab initio molecular
orbital geometry optimizations of 1-8 were performed at the
Hartree-Fock level using the GAUSSIAN9210 and GAUSSIAN-
94W11 series of programs with the 6-31G12 basis set. All
geometry optimizations were checked by analytical frequency
analysis in order to confirm that a true minimum had been
found.
NLO Stu d ies. Crystals of 2, 3(Pca21), or 5 were ground with
an agate mortar and pestle and a sample was placed between
two glass slides that were then held in a Nd:YAG laser beam.
Thin films were prepared by placing a drop of a solution of 3
or 5 onto a quartz slide and allowing the solvent to evaporate.
Ack n ow led gm en t. We thank Dr. Steven Gieb for
his assistance with data collection for the X-ray struc-
tures reported herein. This work was supported by the
Petroleum Research Fund, administered by the Ameri-
can Chemical Society. Most of the calculations were
performed on a computer donated to Carnegie Mellon
University by the Intel Corp. We also express our
thanks to Prof. Gary Patterson at Carnegie Mellon
(E)-6-(2-(4-(Dim eth ylam in o)ph en yl)eth en yl)fu lven e (5).33
A 500-mL three-necked round-bottomed flask fitted with a
condenser and an addition funnel was charged with 3.0 mL
(2.4 g, 36 mmol) of freshly distilled cyclopentadiene, 1.0 g (43
mmol) of Na, and 30 mL of MeOH under N2. (E)-4-N,N-
Dimethylaminocinnamaldehyde (2.055 g, 11.72 mmol) in 400
mL of MeOH was added dropwise, and the stirred solution
turned orange. After 2 h, the reaction mixture was quenched
with water, extracted with 3 × 100 mL of ether, dried (MgSO4),
(34) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467.
(35) Sheldrick, G. M. SHELXL-93, A Program for Crystal Structure
Refinement, University of Go¨ttingen, 1993.
(36) ORTEP-3 for Windows: Farrugia, L. J . J . Appl. Cryst. 1997,
30, 565.
(32) Nesmeyanov, A. N.; Sazonova, V. A.; Drozd, V. N.; Rodionova,
N. A.; Zudkova, G. I. Bull. Acad. Sci. USSR, Div. Chem. Sci. (Engl.
Trans.) 1965, 2029.
(33) No synthetic details for 5 were reported by Kawabe et al. (ref
7a).
(37) PLUTO is part of the package of programs provided with the
Cambridge Structural Database software.

9076 J . Org. Chem., Vol. 64, No. 25, 1999
Peterson et al.
the calculated structures of 1-8. π-Charges (HF/6-31G//HF/
6-31G) for calculated structures of 1-8 and also for 2 (HF/6-
31G//X-ray). This material is available free of charge via the
Internet at http://pubs.acs.org.
University and Prof. David Pratt at the University of
Pittsburgh for assistance with the NLO experiments.
Su p p or tin g In for m a tion Ava ila ble: Tables of atomic
coordinates, bond lengths, and bond angles for the X-ray
structures of 2, 3(P21/c), and 3(Pca21). Atomic coordinates for
J O990931X