Syntheses and third-order NLO properties of η6 complexes of N-ethylcarbazole with Cr(CO)3 and Cr(CO)2PPh3 moieties

Article abstract of DOI:10.1039/b518346d

Carbazole mono- and bimetallic complexes [(N-ethylcarbazole){Cr(CO) ₃}₂] (3) and [(N-ethylcarbazole)Cr(CO)₂L] (L = CO (2) and PPh₃ (4)) have been synthesized, and X-ray crystal structure analyses performed for 2 and 4. Each complex possesses spectroscopically intense intraligand (IL) and metal-to-ligand charge-transfer (MLCT) excitations that are well separated in energy. The MLCT bands display a small to moderate negative solvatochromism. The third-order non-linear optical (NLO) properties, measured by the time-resolved forward degenerate four-wave mixing (DFWM) technique in solution at 532 nm, have been studied for these complexes, together with N-ethylcarbazole for comparison. It is found that the third-order NLO response can be tuned by chromium complexation. The molecular second-order hyperpolarizabilities, γ, of the synthesized complexes are determined to be in the order of 10^-30 esu and clearly depend on both the nature of the ligand and the extent of chromium tricarbonyl coordination by the carbazole ring. The carbazole-mediated metal-metal electronic interaction existing in 3 is believed to be a possible factor responsible for a large improvement in the non-linearity. the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006.

Full text of DOI:10.1039/b518346d

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Syntheses and third-order NLO properties of g6 complexes of
N-ethylcarbazole with Cr(CO)3 and Cr(CO)2PPh3 moietiesw
Yanchao Che, Xiaohui Tian,* Hui Chen, Zhenyu Tang and Jiaping Lin
Received (in Montpellier, France) 4th January 2006, Accepted 4th April 2006
First published as an Advance Article on the web 27th April 2006
DOI: 10.1039/b518346d
Carbazole mono- and bimetallic complexes [(N-ethylcarbazole){Cr(CO)3}2] (3) and
[(N-ethylcarbazole)Cr(CO)2L] (L = CO (2) and PPh3 (4)) have been synthesized, and X-ray
crystal structure analyses performed for 2 and 4. Each complex possesses spectroscopically intense
intraligand (IL) and metal-to-ligand charge-transfer (MLCT) excitations that are well separated in
energy. The MLCT bands display a small to moderate negative solvatochromism. The third-order
non-linear optical (NLO) properties, measured by the time-resolved forward degenerate four-wave
mixing (DFWM) technique in solution at 532 nm, have been studied for these complexes,
together with N-ethylcarbazole for comparison. It is found that the third-order NLO response can
be tuned by chromium complexation. The molecular second-order hyperpolarizabilities, g, of the
synthesized complexes are determined to be in the order of 10ꢀ30 esu and clearly depend on both
the nature of the ligand and the extent of chromium tricarbonyl coordination by the carbazole
ring. The carbazole-mediated metal–metal electronic interaction existing in 3 is believed to be a
possible factor responsible for a large improvement in the non-linearity.
act as useful NLO materials. On the other hand, the carbazole
Introduction
unit is known as a hole-transporting and electroluminescent
structure; its simple functionalization allowing the preparation
of attractive molecules with multi-functional properties6 such
as photoconductivity,7 optical non-linearity8 and photorefrac-
tivity.9 Nisoli et al. reported that uniaxially-oriented films of
poly[1,6-di(N-carbazolyl)-2,4-hexadiyne] exhibit a very large
resonant third-order NLO susceptibility, w(3), up to 2 ꢁ 10ꢀ7
esu.10 Recent results show that enhanced g values can be
readily obtained in carbazole derivatives by extending the
conjugation length, altering the substituted positions, or
choosing appropriate substituents.11 With the above consid-
erations in mind, we have designed a series of metallocarba-
zole chromophores with novel bonding patterns and
coordination geometries which permit spatial arrangements
of atoms that may not be easily accessible in other systems,
resulting in an increase in design flexibility. The Cr(CO)3
group is incorporated into the carbazole ring in order to allow
the fine-tuning of its electronic/optical properties, and more-
over, the realization of new properties resulting from syner-
gistic effects.12 By replacing one CO group by another ligand,
organometallic chromophores with different absorption bands
(linear and non-linear) can be prepared. We describe in this
paper the syntheses and structures of [(N-ethylcarbazole){Cr
(CO)3}2] (3) and [(N-ethylcarbazole)Cr(CO)2L], where L =
CO (2) and PPh3 (4) (Scheme 1), as well as the effects of the
metal and its ligand on their linear and non-linear optical
properties. Considering that the different substituents can be
easily introduced into the carbazole ring,13 it should be noted
that these complexes seem to be intriguing candidates as
chromophoric segments for the future development of orga-
nometallic NLO polymers.
There has been considerable interest in the non-linear optics
(NLO) of organometallic compounds that possess low-energy
metal-to-ligand charge-transfer (MLCT) or ligand-to-metal
charge-transfer (LMCT) excitations,1 because these effects
can exert additional control on the optoelectronic properties
through the oxidation state of the metal and the nature of the
attached ligands. Moreover, the interaction of d-orbitals of the
transition metal with p-orbitals of the organic moiety offers
new ways of enhancing the non-linear response. To date,
several useful reviews of the NLO behavior of organometallic
compounds have been published.2 Among the enormous
variety of metallic sub-units that can be attached to p-con-
jugated organics, the Cr(CO)3 moiety has been a target of
focus due to its high polarity, the amphoteric nature of its
electrons and its rich chemistry.3 Enhancement of the first-
order hyperpolarizability, b, of alkynylated and alkenylated
arenes upon Cr(CO)3 complexation has been theoretically
predicted and experimentally corroborated.2d,4 Cheng et al.
measured the second-order hyperpolarizability, g, of
Cr(CO)3–benzene (CBC) complexes and found that small
changes in the substituents attached to the benzene ring could
produce a major effect on the non-linear response.5 This
implies that correctly designed molecules of this type may
School of Materials Science and Engineering, East China University
of Science and Technology, Shanghai 200237, China. E-mail:
tianxh@263.net; Fax: þ86 21-54288175
w Electronic supplementary information (ESI) available: 1H NMR
and IR spectra of 2–4, and full crystallographic summary data for 2
and 4. For data in CIF or other electronic format see DOI: 10.1039/
b518346d
ꢂc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 883–889 | 883

View Article Online
1
1324. H NMR (500 MHz, CDCl3) d: 1.50 (t, 3 H, J = 14.6
Hz), 4.18 (m, 2 H), 5.10 (t, 1 H, J = 12.5 Hz), 5.62 (t, 1 H, J =
12.9 Hz), 5.89 (d, 1 H, J = 6.9 Hz), 6.62 (d, 1 H, J = 6.4 Hz),
7.27 (t, 1 H, J = 17.9 Hz), 7.32 (d, 1H, J = 8.2 Hz), 7.50 (t, 1
H, J = 15.4 Hz) and 7.91 (d, 1 H, J = 7.8 Hz). 13C NMR (125
MHz, CDCl3) d: 14.22, 38.80, 75.04, 85.83, 89.34, 92.66, 93.41,
110.06, 120.83, 121.55, 123.15, 123.76, 128.27, 142.76 and
234.72. MS-EI (70 eV, m/z (%)): 331 [M1] (7.0), 247 [M1
ꢀ
3CO] (100.0), 195 [M1 ꢀ Cr(CO)3] (12.6) and 52 [Cr1] (41.8).
Complex 3: Yield 0.21 g (6.0%). Mp 184–185 1C (dec.).
Anal. calc. for C20H13NO6Cr2: C, 51.39; H, 2.78; N, 3.00; Cr,
22.27; found: C, 51.31; H, 2.89; N, 2.93; Cr, 22.35%. UV-vis
(CH2Cl2, lmax/nm (104 e/cmꢀ1 molꢀ1 dm3)): 432 (0.7) and 346
(1.6). IR (KBr, n/cmꢀ1): 1939, 1902, 1853, 1530, 1506, 1452,
Scheme 1
Experimental
General procedures
1
Unless otherwise noted, all manipulations with air-sensitive
materials were performed under a nitrogen atmosphere using
standard Schlenk techniques and a Vacuum Atmospheres dry
box. Solvents were freshly distilled over appropriate drying
reagents, and all Schlenk flasks were flame-dried under va-
cuum prior to use. Cr(CO)6 (99%, Strem Chemistry) was
sublimed prior to use. N-ethylcarbazole (95%, Aldrich) was
used as received. Elemental analyses were carried out on
Elementar Vario EL III (C, H, N) and TJARIS 1000 (Cr)
analyzers, respectively. Electron impact mass spectra were
recorded on a Micromass GCT (EI, 70 eV) mass spectrometer.
1H NMR (500 MHz) and 13C NMR (125 MHz) spectra were
recorded on a Bruker AVANCE 500 MHz spectrometer using
CDCl3 as solvent at 298 K and Me4Si as an internal standard.
IR spectra were recorded on a Nicolet Magna-IR 550 instru-
ment using KBr pellets. Column chromatography was per-
formed on Matrix silica (60–90 mm). UV-vis spectra were
obtained on a Varian Cary 500 spectrophotometer rebuilt by
OLIS to incorporate computer control. Melting points were
measured using a Reichert Thermopan melting point micro-
scope and were uncorrected.
1433 and 1347. H NMR (500 MHz, CDCl3) d: 1.55 (t, 3 H,
J = 21.3 Hz), 3.98 (m, 2 H), 5.07 (t, 2 H, J = 12.3 Hz), 5.59 (t,
2 H, J = 12.7 Hz), 5.73 (d, 2 H, J = 6.8 Hz) and 6.42 (d, 2 H,
J = 6.3 Hz). 13C NMR (125 MHz, CDCl3) d: 13.91, 30.39,
74.65, 85.86, 87.54, 92.33, 93.50, 125.05 and 233.37. MS-EI (70
eV, m/z (%)): 467 [M1] (6.0), 383 [M1 ꢀ 3CO] (4.5), 299
[M1 ꢀ 6CO] (10.7), 195 [M1 ꢀ 2Cr(CO)3] (2.8) and 52.0 [Cr1]
(27.5).
Synthesis of [(N-ethylcarbazole)Cr(CO)2PPh3] complex (4)
2 (0.5 g, 1.5 mmol) and triphenylphosphine (0.59 g, 2.3 mmol,
recrystallized from ethanol) were irradiated in 75 mL of dry
benzene, in a Schlenk tube, under a Hanovia ultraviolet lamp,
with continuous stirring, under a nitrogen atmosphere, for 8 h.
The heat from the lamp allowed the reaction mixture to reflux.
The dark violet solution was filtered hot through a glass frit
and rinsed with n-hexane to obtain 0.30 g (35.2%) of a violet
solid which was stable to air. Recrystallization from dry, de-
gassed CH2Cl2/n-hexane (3 : 1 v/v) yielded large, dark-violet
crystals. Mp 160–161 1C (dec.). Anal. calc. for
C34H28NO2PCr: C, 72.21; H, 4.96; N, 2.48; Cr, 9.20; found:
C, 72.13; H, 4.88; N, 2.51; Cr, 9.25%. UV-vis (CH2Cl2, lmax
/
nm (103 e/cmꢀ1 molꢀ1 dm3)): 422 (3.6) and 351 (6.1). IR (KBr,
n/cmꢀ1): 1856, 1795, 1539, 1470, 1433 and 1329. 1H NMR (500
MHz, C6D6) d: 1.28 (t, 3 H, J = 14.2 Hz), 3.72 (m, 2 H), 4.58
(t, 1 H, J = 11.1 Hz), 4.96 (m, 2 H), 5.42 (t, 1 H, J = 4.4 Hz),
7.08 (d, 1 H, J = 7.9 Hz), 7.17 (d, 1 H, J = 8.2 Hz), 7.22 (s, 9
H) and 7.47 (m, 2 H) and 7.62 (m, 6 H). 13C NMR (125 MHz,
CDCl3) d: 14.43, 39.36, 71.57, 85.06, 89.46, 91.59, 109.24,
119.46, 120.41, 121.05, 126.35, 126.71, 128.34, 129.27, 133.69,
134.55, 139.97, 140.25 and 242.48. MS-EI (70 eV, m/z (%)):
565 [M1] (1.3), 509 [M1 ꢀ 2CO] (20.0), 314 [CrPPh31] (43.6),
262 [PPh31] (100.0), 195 [M1 ꢀ Cr(CO)2PPh3] (32.1) and 52
[Cr1] (26.9).
Syntheses of mono- (2) and bis-Cr(CO)3 (3) complexes of
N-ethylcarbazole
A mixture of N-ethylcarbazole (1.46 g, 7.5 mmol) and freshly
sublimed Cr(CO)6 (3.30 g, 15.0 mmol) in di-n-butylether (85
mL) and dioxane (12 mL) was heated under reflux (bath
temperature: 150 1C) for 30 h. After being cooled, the solution
was filtered through a Celite filter and the clear solution
concentrated until a solid appeared. The residue was extracted
into 10 mL of CH2Cl2 and passed through a 5 ꢁ 35 cm column
of silica gel, eluting with CH2Cl2/petroleum ether (1 : 3 v/v).
Two bands were observed: the first yellow band (complex 2;
MS (m/z): 331 [M]1) was collected with B500 mL of the
eluent, and the second orange band (complex 3; MS (m/z): 467
[M]1) collected with B300 mL of eluent. The products were
recrystallized twice from CH2Cl2/n-hexane (3 : 1 v/v), respec-
tively, to obtain light yellow crystals of 2 and orange micro-
crystals of 3.
Crystallography
Single crystals of 2 and 4 suitable for X-ray diffraction were
obtained by recrystallization from CH2Cl2/n-hexane at ꢀ5 1C
under an inert atmosphere, and were sealed in thin-walled
glass capillaries. Data were collected on a Bruker Smart Apex
CCD diffractometer with graphite-monochromated Mo-Ka
radiation (l = 0.71073 A) using the o-2y technique at
293 K. Both structures were solved by direct methods using
SHELXS-97 and refined by the full-matrix least-squares
Complex 2: Yield 1.33 g (53.7%). Mp 141–142 1C. Anal.
calc. for C17H13NO3Cr: C, 61.63; H, 3.93; N, 4.23; Cr, 15.71;
found: C, 61.75; H, 3.87; N, 4.19; Cr, 15.82%. UV-vis
(CH2Cl2, lmax/nm (104 e/cmꢀ1 molꢀ1 dm3)): 395 (1.1) and
331 (2.1). IR (KBr, n/cmꢀ1): 1936, 1850, 1554, 1471, 1435 and
ꢂc
884 | New J. Chem., 2006, 30, 883–889 This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006

View Article Online
Table 1 Details of data collection and structure refinement for 2 and
4
into two pump beams, k1 and k2, and a probe beam, k3,
through reflecting beam splitters. They were then temporally
and spatially overlapped into the sample with a 205 mm focal
length lens. The signal beam, k4, was collected by a photo-
diode and then analyzed by a boxcar and computer. In the
experiment, the sample dissolved in CH2Cl2 was placed in a 1
mm quartz cell and the intensity of the DFWM response for
the solution samples, corresponding to the maximum of the
temporal profiles obtained by scanning the delay, was com-
pared to the signal obtained for a reference sample (CS2 was
employed in our case; its w(3) value was 6.8 ꢁ 10ꢀ13 esu at 532
nm) under the same light intensity conditions. The w(3) value
and the non-linear refractive index, n2, of the samples were
then calculated from eqn (1) and eqn (2), respectively.15
2
4
Empirical formula
Formula weight
Crystal system
Space group
Color, form
a/A
b/A
c/A
C17H13 Cr NO3 C35H30Cl2CrNO2P
331.28
650.47
Triclinic
P-1
Monoclinic
P2(1)/c
Yellow, blocks
8.7347(7)
11.5839(9)
14.6847(11)
90
90.8290(10)
90
1485.7(2)
4
1.481
Violet, blocks
10.0091(9)
10.7369(10)
14.6282(14)
90.880(2)
90.925(2)
101.742(2)
1538.7(2)
2
a/1
b/1
g/1
V/A3
Z
dcalc./g cmꢀ3
1.404
0.631
672
À Á
ꢀ
ꢁ
ꢀ
ꢁ
2
I4 1=2dr
n
Absorption coefficient/mmꢀ1 0.781
ad
2
ad exp
wð3Þ
¼
wðr3Þ
ð1Þ
F(000)
Crystal size/mm
680
I4r
d
nr 1 ꢀ expðꢀadÞ
0.489 ꢁ 0.325 ꢁ 0.486 ꢁ 0.432 ꢁ
0.237 0.120
12pwð3Þ
y range/1
2.24 to 27.00
8541
3212
1.39 to 27.00
9054
6528
n2 ¼
ð2Þ
No. of reflections collected
No. of unique data
No. of data/restraints/
parameters
Rint
GOF/F2
n
3212/0/217
6528/2/375
where a is the linear absorption coefficient, I4 refers to the
intensity of the conjugate signal, d is the thickness of the
sample, n is the linear refractive index of the medium and the
subscript ‘r’ refers to the reference sample counterpart. For
molecular systems, the second-order hyperpolarizability, g, is
related to w(3) by eqn (3).16
0.0701
1.021
0.0520
1.112
a,b
R indices [I > 2s(I)]
R1 = 0.0402
wR2 = 0.1098
R1 = 0.0469
wR2 = 0.1141
0.407/ꢀ0.550
R1 = 0.0699
wR2 = 0.2109
R1 = 0.0851
wR2 = 0.2249
1.948/ꢀ1.079
R indices (all data)
Largest differential peak
and hole (eAꢀ3
wð3Þ
L4N
)
g ¼
ð3Þ
1
a
b
2
2 2
R1 = SJFo| ꢀ |FcJ/S|Fo|. wR2 = {S[w(Fo ꢀ Fc2)2]/S[w(Fo ) ]}2.
Here, L (L = (n2 þ 2)/3) is the local field factor approximated
by the Lorentz expression, n is the linear refractive index of the
solvent and N represents the number density of the compound.
method on all F2 data using the program SHELXL-97.14
There was a CH2Cl2 solvate in the unit cell of complex 4.
All non-hydrogen atoms were refined anisotropically. Hydro-
gen atoms were inserted in idealized positions and refined
using a riding model. Relevant crystallographic information is
summarized in Table 1.
Results and discussion
Syntheses and structural characterization
The reaction of N-ethylcarbazole (1) with 2 equivalents of
Cr(CO)6 in di-n-butylether and dioxane (7 : 1 v/v) gives a
mixture of 2 and 3 in a 9 : 1 ratio. Attempts to improve the
yield of 3 using excess chromium reagent, a high boiling point
solvent and prolonged reaction time, failed. Heating pure 2 in
dioxane at temperatures up to 110 1C under an inert atmo-
sphere for several hours gives rise to a conversion to 3, but the
reverse process is not feasible for the conversion of 3 into 2
under the same conditions. The sequence of events indicates
that mononuclear complex 2 is the kinetic product, which
subsequently rearranges to the corresponding binuclear com-
plex 3 and free ligand 1 as the thermodynamically more stable
products.17 EI-mass spectra of 2 and 3 show peaks of highest
mass corresponding to their monomeric formulas, respec-
tively. The infrared spectra of 2 and 3 in the carbonyl stretch-
NLO measurement
The third-order NLO properties of the studied compounds
were measured using the technique of time-resolved forward
degenerate four-wave mixing (DFWM). Fig. 1 shows a sche-
matic diagram of the geometry of the light beams used in the
DFWM experiments.
The excitation source was a mode-locked Nd:YAG laser
system (wavelength 532 nm, pulse width 35 ps, repetition rate
10 Hz, pulse energy 1.5 mJ). The incident laser beam was split
ing region are very different: two bands are seen in 2 (nCO
=
1936 and 1850 cmꢀ1), while 3 has three bands (nCO = 1939,
1902 and 1853 cmꢀ1), reflecting some coupling of the CO
vibration of the Cr(CO)3 units in the latter case. In the proton
and carbon NMR spectra, the characteristic high-field shift of
the complexed arene carbon and proton resonances not only
facilitates the unambiguous assignment of the carbazole
Fig. 1 Schematic diagram of the geometry of the light beams in a
four-wave mixing experiment.
ꢂc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 883–889 | 885

View Article Online
signals but also clearly indicates the mutual electronic influ-
ence of both coordinated rings for the dichromium species.18
Further reaction of 2 with 1.5 equivalents of triphenylpho-
sphine in benzene gives 4 as a violet solid. We tried to replace 2
by
3 in this reaction, but no N-ethylcarbazole–bis(Cr
(CO)2PPh3) complex was found as a product; the reactions
involved appearing to be non-specific. Complex 4 is stable to
air and moisture in the solid state but decomposes slowly in
solution at room temperature. The EI mass spectrum of 4
shows the highest mass peak for the molecular ion, [M]1, with
a relative intensity of 1.3%. The IR spectrum reveals two CO
stretching bands (nCO = 1856 and 1795 cmꢀ1) that are shifted
to lower energies (compared to 2), suggesting a distinct mutual
influence of the individual ligands via the metal center. Similar
electronic effects are also noted in the carbon and proton
resonances of the coordinated ring in the NMR spectra, i.e.
shifted to high-field, while the carbonyl carbon resonances are
slightly shifted to low-field.
The structures of 2 and 4 have been determined by X-ray
crystallography.zz The molecular structures and selected bond
lengths and angles are shown in Fig. 2 and Table 2, respec-
tively. Attempts to grow X-ray quality crystals of 3 in various
solvents at different temperatures have been unsuccessful to
date. Complexes 2 and 4 are intrinsically asymmetric mole-
cules and crystallize in the monoclinic and triclinic crystal
systems, with space groups P2(1)/c and P-1, respectively.
There is a CH2Cl2 solvate in the unit cell of complex 4. To
probe the effect of ligand exchange, some useful comparisons
between the complexes can be made concerning the coordina-
tion geometry around the metal center. Complex 2 exhibits an
expected three-legged piano stool structure, having a staggered
conformation, with one carbonyl group being disposed above
the central heterocycle.3c Replacing one CO with a triphenyl-
phosphine ligand, as in 4, causes a rotation about the Cr–c(0)
[c(0) = coordinated ring centroid] axis that forces the Cr–P
bond into a position opposite the N-ethyl group in response to
the steric requirements of the PPh3 group, one phenyl ring of
which is less susceptible to the steric bulk and electronic
influences of the coordinated ring than the other two. The
phosphine ligand, which enhances the electron density along
the Cr–PPh3 bond axis, strengthens the Cr–CO bonds and
weakens the C–O bonds, as shown by the variation in the
corresponding bond lengths. The distance of the chromium
from the mean plane in 2 is 1.752 A, which is comparable to
that in 4 (1.749 A), suggesting that the metal–carbazole
bonding seems to be less sensitive to the nature of the ligand.
The carbazole ring and its attached N-methene carbon are
nearly coplanar, although slight deviations from planarity are
observed in the coordinated aryl ring (planar to within 0.032 A
for 2 and 0.055 A for 4). The C–C distances within the
coordinated ring are routinely longer than those in the un-
coordinated aryl ring (average 1.406/1.388 A for 2 and 1.408/
1.393 A for 4). This is likely to reflect a slight loss in
aromaticity upon coordination, owing to back-donation of
electron density from the metal into the p* orbitals of the ring.
A similar situation has been found for several (stilbene)Cr
Fig. 2 ORTEP drawings of 2 (top) and 4 (bottom) with thermal
ellipsoids at the 30% probability level.
(CO)3 compounds, in which the electron density is localized on
the coordinated ring rather than spread out over the entire
molecular configuration.3a In complexes 2 and 4, the differ-
ences between the two aryl rings of the carbazole can lead to
an increase in the magnitude of the permanent dipolar
moment along the carbazole plane.
Table 2 Selected bond lengths (A) and angles (1) for 2 and 4
2
N(1)–C(4)
N(1)–C(16)
C(4)–C(9)
C(11)–C(12)
C(10)–C(15)
Cr–C(2)
1.373(2)
1.456(2)
1.432(2)
1.392(3)
1.396(3)
1.828(2)
1.142(3)
1.149(2)
N(1)–C(11)
C(4)–C(5)
C(8)–C(9)
C(10)–C(11)
Cr–C(1)
1.392(2)
1.392(3)
1.406(3)
1.400(2)
1.839(2)
1.826(2)
1.149(3)
Cr–C(3)
C(2)–O(2)
C(1)–O(1)
C(3)–O(3)
C(4)–N(1)–C(11)
C(8)–C(9)–C(4)
C(14)–C(15)–C(10)
108.70(14)
119.28(17)
118.4(2)
C(5)–C(4)–C(9)
C(15)–C(10)–C(11)
N(1)–C(16)–C(17)
121.39(17)
119.58(18)
112.80(18)
4
N(1)–C(3)
N(1)–C(15)
C(9)–C(10)
C(3)–C(4)
C(7)–C(8)
Cr–C(2)
1.387(5)
1.454(5)
1.419(5)
1.415(6)
1.394(6)
1.806(4)
1.176(5)
N(1)–C(10)
C(10)–C(11)
C(9)–C(14)
C(3)–C(8)
Cr–C(1)
1.379(5)
1.411(6)
1.403(5)
1.404(6)
1.820(4)
2.3209(9)
1.170(5)
Cr–P(1)
C(2)–O(2)
C(1)–O(1)
C(17)–P(1)–Cr
C(29)–P(1)–Cr
C(28)–C(23)–C(24)
116.38(11)
119.63(11)
118.5(3)
C(23)–P(1)–Cr
C(18)–C(17)–C(22)
C(34)–C(29)–C(30)
112.33(11)
118.3(3)
118.1(3)
z CCDC 291883 (2) and CCDC 291884 (4). For crystallographic data
in CIF or other electronic format see DOI: 10.1039/b518346d
ꢂc
886 | New J. Chem., 2006, 30, 883–889 This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006

View Article Online
Fig. 4 Temporal DFWM signals of 1–4 in solution at 532 nm.
Fig. 3 Electronic absorption spectra of 1–4 in CH2Cl2 at 298 K.
in the case of 4, the solvatochromic behavior in various solvent
environments appears to be comparatively less significant.
Clearly, such spectroscopic behavior is not a simple attribute
of the electronic coupling between the carbazole unit and the
metal atom, but is also associated with the nature of the other
ligand. The charge-transfer from metal-to-ligand will give rise
to a significant electronic asymmetry in the ground state,
despite the lack of a good coincidence between the charge-
transfer and dipole moment directions.2d,4 The delocalization
of p-electrons, on the other hand, will always increase the
electronic symmetry and reduce the dipole difference between
the ground and the excited states. This can help to explain the
above result, where the red-shifts in the positions of the
corresponding absorption peaks and reduced negative solva-
tochromism concur when a CO group is replaced by a PPh3
ligand, as in 4. Overall, a correct elaboration of the coordina-
tion environment and its solvatochromism will allow the
tuning of the position of the absorption band to a desired
wavelength, so as to create an optimal balance between near
resonance and low a, as required in NLO devices.
Linear optical properties
The electronic absorption spectra of molecules 1–4 were
recorded in a variety of solvents with different polarities (i.e.
cyclohexane, CH2Cl2 and DMF). Fig. 3 displays the spectra in
CH2Cl2. Complexes 2–4 show two characteristic absorption
bands in the 300–525 nm region.3a–c,4,12 The absorption band
at lower energy is assigned to MLCT arising from a charge
transition from the chromium center to the p- and s-bound
ligands. The next band, at higher energy, is the most intense
absorption, and arises from intraligand (IL) transitions with
high oscillator strengths, mostly from p–p* transitions. The
change from 2 to 3 causes a subtle red-shift, both in the lower
energy band by 37 nm and in the higher energy band by 15 nm,
indicating that a significant electron density transfer takes
place from one metal center to the other through the carbazole
conjugated system. This inference is also supported by the
assumption that the chromium–carbonyl–arene fragment be-
haves electron-amphoterically to achieve charge redistribution
within the dichromium system. Replacing a CO group in 2
with a phosphine ligand, as in 4, is also found to lower the
transition energy and to enhance the MLCT intensity, with a
long tail extending far into the visible. We attribute the red-
shifts to an increased delocalization of p-electrons into the
carbazole unit by virtue of the metal-to-ligand p back-bonding
interaction, because the PPh3 ligand is electron-donating in
nature and renders the chromium center more electron-rich.
As can be seen from Table 3, the absorption bands of 2–4 are
governed by a negative solvatochromism, which is typically
more pronounced for the low energy absorption, indicating
that the electronic ground state is more polar than the first
excited state, and therefore an increase in solvent polarity
causes a hypsochromic shift. This is also in accordance with
the highly polar nature of these complexes, which can be
attributed to dominant mz-dipole vector component directed
from the carbazole plane to the metal center. To our surprise,
Non-linear optical properties
Insert results from DFWM are given in Fig. 4, in which the
solid line is the theoretical fitted curve with a Gaussian
function. It is generally believed that the presence of any
population grating induced by one-photon or two-photon
absorption will result in an asymmetrically-temporal DFWM
signal.19 For each studied molecule, even with complex 4,
which has weak absorption at the measured wavelength of 532
nm, the measured signal did not show appreciable asymmetry
either. This implies that the one-photon or two-photon con-
tribution to the DFWM process is negligible. The third-order
susceptibility, w(3), the non-linear refractive index, n2, and the
second-order hyperpolarizability, g, were obtained, as listed in
Table 3. The results reflect the fact that there is a significant
Table 3 MLCT band maxima and third-order non-linear optical coefficients for 1–4 in CH2Cl2, unless otherwise stateda
max/nm (umax/cmꢀ1
)
w(3)/10ꢀ13 esu
g/10ꢀ30 esu
n2/10ꢀ11 esu
Comments
Compound
l
1
2
3
4
0.51
6.41
9.49
8.24
b
0.26
2.54
3.81
3.27
0.14
1.70
2.51
2.18
Off-resonance
Off-resonance
Off-resonance
Near-resonance
409 (24450)b; 395 (25316); 387 (25840)c
434 (23041)b; 432 (23148); 417 (23981)c
424 (23585)b; 422 (23697); 416 (24038)c
g and w(3) values were normalized to a concentration of 1.0 ꢁ 10ꢀ4 mol Lꢀ1
.
a
c
Measured in cyclohexane. Measured in DMF.
ꢂc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 883–889 | 887

View Article Online
contribution to the magnitude of the third-order non-linear
response from chromium complexation, and that the mea-
sured w(3) and g values are highly dependent on both the nature
of the ligand and the extent of chromium tricarbonyl coordi-
nation by the carbazole ring. Cheng et al. reported that the g
values for several CBC complexes, measured by the third
harmonic generation (THG) technique, are 2–21 ꢁ 10ꢀ36
esu.5 The g value of 2 is five or six orders of magnitude higher
than that of the CBCs, confirming the potential advantage of
the well-conjugated, planar and rigid ring of carbazole for
enhancing the third-order response. Recently, Qian et al.
investigated the third-order optical non-linearity of six 3,6-
di-acceptor-substituent carbazole multi-polar chromophores
using the single-beam Z-scan technique.11a The g value of these
organic molecules was estimated to be 1.5–2.9 ꢁ 10ꢀ33 esu,
three orders of magnitude smaller than that of 2, revealing the
vital role of metal complexation. Although the g values
measured by the different techniques do not, in general,
produce a quantitatively reliable comparison, since the experi-
mental conditions and techniques used in each laboratory vary
significantly, a qualitative trend for a systematically varied
structure can still be deduced. A very large g value up to 3.8 ꢁ
10ꢀ30 esu was obtained for complex 3, in which two metal
centers are simultaneously coordinated to the carbazole ring.
In this case, the Cr(CO)3 fragment functions as either a ground
state acceptor or an excited state donor.3a,4 These factors
influence the nature of the charge-transfer processes between
the metal and the carbazole ligand, or even inside the carba-
zole itself, producing a reciprocal push–pull organometallic
architecture with carbazole-mediated metal–metal electronic
interactions. The g value of 4 appears to be larger than that of
2, suggesting that the effect of ligand exchange on the orga-
nometallic moiety has a more subtle impact on non-linearity.
The strong electron-donating ability of the PPh3 group can
permit more metal electrons in 4 to be involved in charge-
transfer excitation than in 2. The enhanced g value of the
former is considered to originate from an additional contribu-
tion of the MLCT excitation. 4 exhibits a strong charge-
transfer absorption band, occurring in a relatively longer
wavelength region, even though this band displays a small
negative solvatochromism. Apart from the concern about
making a judicious choice of ligands and metals to obtain
the highest g values, we are also interested in structural
features of the carbazole ring and their ease of substitution
by polymerizable groups. For this reason, these complexes can
form polymers or be incorporated in polymers, either as
pendant chains or into the polymer backbone, affording the
possibility of introducing more polarizable groups into poly-
mer chains than may be accessible in a solely organic system.
Furthermore, the very bulky organometallic moiety may be
helpful in reducing chromophore–chromophore electrostatic
interactions, and in optimizing the chromophore loading level
and the order parameter for the polymer matrix.20
third-order optical non-linearities. Chromium complexation
can be one of the most effective tools for modulating and
improving the non-linear response. It has been observed that
the w(3) and g values of these complexes are highly dependent
on both the nature of the ligand and the extent of chromium
tricarbonyl coordination by the carbazole ring. The corre-
sponding non-linearity undergoes a strong enhancement when
two metal centers are simultaneously coordinated to one
carbazole molecule. A carbazole-mediated metal–metal elec-
tronic interaction is proposed to be associated with the
amphoteric behavior of the Cr(CO)3 fragment. Ligand ex-
change on the organometallic moiety also exerts a more subtle
impact on the electronic structure of the molecule. The
electron-donating ability of the PPh3 ligand renders the car-
bazole ring more electron-rich, leading to a substantial im-
provement in the NLO response. These results suggest that the
studied complexes are a particular class of compounds offering
a promising route for third-order NLO studies. To gain more
insight into the factors that affect the non-linear optical
properties, we are continuing our studies on the rational
modification of the organic structure, such as introducing
functional groups onto the coordinated carbazole, in particu-
lar, by extending the carbazole conjugated system to macro-
molecular level.
Acknowledgements
We are grateful to the National Natural Science Foundation
of China (grant no. 20274011) for generous financial support.
We thank Dr Jie Sun for assistance in the X-ray data collection
and refinement of complexes 2 and 4.
References
1 (a) I. R. Whittall, A. M. McDonagh and M. G. Humphrey, Adv.
Organomet. Chem., 1999, 43, 349; (b) U. Behrens, H. Brussaard, U.
Hagenau, J. Heck, E. Hendrickx, J. Kornich, J. G. M. van der
¨
Linden, A. Persoons, A. L. Spek, N. Veldman, B. Voss and H.
´ ´
Wong, Chem.–Eur. J., 1996, 2, 98; (c) G. Rojo, F. Agullo-Lopez, J.
A. Campo, J. V. Heras and M. Cano, J. Phys. Chem. B, 1999, 103,
11020; (d) M. P. Cifuentes, J. Driver, M. G. Humphrey, I.
Asselberghs, A. Persoons, M. Samoc and B. Luther-Davies, J.
Organomet. Chem., 2000, 607, 72; (e) W. F. Sun, C. C. Byeon, C.
M. Lawson and D. Y. Wang, Appl. Phys. Lett., 1999, 74, 3254.
2 (a) J. Heck, S. Dabek, T. Meyer-Friedrichsen and H. Wong,
Coord. Chem. Rev., 1999, 190–192, 1217; (b) G. de la Torre, P.
Vazquez, F. Agullo-Lopez and T. Torres, Chem. Rev., 2004, 104,
´ ´ ´
3723; (c) H. S. Nalwa, Appl. Organomet. Chem., 1991, 5, 349; (d) D.
R. Kanis, M. A. Ratner and T. J. Marks, J. Am. Chem. Soc., 1992,
114, 10338.
3 (a) T. M. Gilbert, F. J. Hadley, C. B. Bauer and R. D. Rogers,
Organometallics, 1994, 13, 2024; (b) T. J. J. Muller, M. Ansorge
¨
and H. J. Lindner, Chem. Ber., 1996, 129, 1433; (c) T. J. J. Muller
¨
and J. Blumel, J. Organomet. Chem., 2003, 683, 354; (d) C. A.
¨
Merlic, B. N. Hietbrink and K. N. Houk, J. Organomet. Chem.,
2001, 66, 6738; (e) C. A. Merlic, J. C. Walsh, D. J. Tantillo and K.
N. Houk, J. Am. Chem. Soc., 1999, 121, 3596.
4 T. J. J. Muller, A. Netz, M. Ansorge, E. Schmalzlin, C. Brauchle
¨
¨
and K. Meerholz, Organometallics, 1999, 18, 5066.
¨
5 L. T. Cheng, W. Tam, G. R. Meredith and S. R. Marder, Mol.
Cryst. Liq. Cryst., 1990, 189, 137.
6 G. H. Wang, C. W. Yuan, H. W. Wu and Y. Wei, J. Appl. Phys.,
1995, 78, 2679.
7 (a) V. S. Mylnikov, Adv. Polym. Sci., 1994, 115, 1–88; (b) J. M.
Pearson and M. Stolka, Poly(N-vinylcarbazole), Gordon and
Breach, New York, 1981.
Conclusions
In summary, we have synthesized three novel Z6-complexes of
N-ethylcarbazole with Cr(CO)3 and Cr(CO)2PPh3 moieties,
and investigated their linear spectroscopic properties and
ꢂc
888 | New J. Chem., 2006, 30, 883–889 This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006

View Article Online
8 (a) B. Kippelen, K. Tamura, N. Peyghambarian, A. B. Padias and
H. K. Hall, Phys. Rev. B: Condens. Matter, 1993, 48, 10710; (b)
Z.-H. Peng, Z.-N. Bao and L.-P. Yu, J. Am. Chem. Soc., 1994,
116, 6003.
Sheldrick, SHELXL-97, Program for refinement of crystal struc-
tures, University of Gottingen, Germany, 1997.
15 D. V. G. L. N. Rao, F. J. Aranda, J. F. Roach and D. E. Remy,
Appl. Phys. Lett., 1991, 58, 1241.
¨
9 (a) W. E. Moerner and S. M. Silence, Chem. Rev., 1994, 94, 127; (b)
G. G. Mallioras, V. V. Krasnikov, H. J. Balink and G. Hadzi-
laonnou, Appl. Phys. Lett., 1994, 65, 262.
10 M. Nisoli, V. Pruneri, V. Magni, S. De Silverstri, G. Dellenpiane,
D. Comoretto, C. Cuniberti and J. Le Moigne, Appl. Phys. Lett.,
1994, 65, 590.
11 (a) Q. Ying, B.-P. Lin, Y.-M. Sun, G. Wang, Y.-P. Cui and
C.-W. Yuan, Acta Chim. Sin., 2005, 63, 2141; (b) X.-W.
Zhan, Y.-Q. Liu, D.-B. Zhu, X.-C. Liu, G. Xu and P.-X. Ye,
Chem. Phys. Lett., 2001, 343, 493; (c) X.-W. Zhan, Y.-Q. Liu,
D.-B. Zhu, X.-C. Liu, G. Xu and P.-X. Ye, Chem. Phys. Lett.,
2002, 362, 165.
16 K. Kandasamy, S. J. Shetty, P. N. Puntambekar, T. S. Srivastava,
T. Kundu and B. P. Singh, Chem. Commun., 1997, 13, 1159.
17 (a) K. H. Dotz, J. Stendel, Jr, S. Muller, M. Nieger, S. Ketrat and
¨
M. Dolg, Organometallics, 2005, 24, 3219; (b) K. H. Dotz, N.
¨
¨
Szesni, M. Nieger and K. Nattinen, J. Organomet. Chem., 2003,
¨
S. Spichiger, Organometallics, 1987, 6, 1173; (d) E. P. Kundig, C.
¨
671, 58; (c) E. P. Kundig, V. Desobry, C. Grivet, B. Rudolph and
¨
Perret and S. Spichiger, J. Organomet. Chem., 1985, 286, 183.
18 (a) S. C. Jones, T. Hascall, S. Barlow and D. O’Hare, J. Am. Chem.
Soc., 1998, 124, 11610; (b) S. Barlow and D. O’Hare, Chem. Rev.,
1997, 97, 637; (c) Y. Oprunenko, I. Gloriozov, K. Lyssenko, S.
Malyugina, D. Mityuk, V. Mstislavsky, H. Gunther, G. von Firks
¨
12 K.-Y. Shen, X.-H. Tian, J.-H. Zhong, J.-P. Lin, Y. Shen and P.-Y.
Wu, Organometallics, 2005, 24, 127.
13 J. V. Grazulevicius, P. Strohriegl, J. Pielichowski and K. Pieli-
chowski, Prog. Polym. Sci., 2003, 28, 1297.
14 (a) G. M. Sheldrick, SHELXS-97, Program for solution of crystal
and M. Ebener, J. Organomet. Chem., 2002, 656, 27.
19 M. T. Zhao, M. Samoc, P. N. Prasad, B. A. Reinhardt, M. R.
Unroe, M. Prazak, R. C. Evers, J. J. Kane, C. Jariwala and M.
Sinsky, Chem. Mater., 1990, 2, 670.
20 P. Wang, P. W. Zhu, C. Ye, A. E. Asato and R. S. Liu, J. Phys.
Chem. A, 1999, 103, 7076.
structures, University of Gottingen, Germany, 1997; (b) G. M.
¨
ꢂc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 883–889 | 889