- Email: [email protected]
Optical poling of several halogen derivatives of pyrazoloquinoline
Optics Communications 242 (2004) 401–409 www.elsevier.com/locate/optcom
Optical poling of several halogen derivatives of pyrazoloquinoline E. Kos´cien´ a, J. Sanetra b, E. Gondek b, B. Jarosz c, I.V. Kityk J. Ebothe e, A.V. Kityk a,*
d,*,1
,
a
e
Institute for Computer Science, Technical University of Czestochowa, Armii Krajowej 17, 42-200 Czestochowa, Poland b Institute of Physics, Technical University of Krakow, Podhorazych 1, 30-084 Krakow, Poland c Department of Chemistry, Hugon Kołłotaj Agricultural University, Al. Mickiewicza 24/28, 30-059 Krakow, Poland d Institute of Biology and Biophysics, Technical University of Czestochowa, PL-42200 Czestochowa, Poland Universite de Reims, UFR Sciences, Laboratoire de Microscopie et dÕEtude de Nanostructures, E.A.n 3799, Boite Postale 138, 21 Rue Clement Ader, F-51685 Reims Cedex 02, France Received 5 June 2004; received in revised form 31 August 2004; accepted 2 September 2004
Abstract Paper deals with optical absorption and second-order optical poling effect in a new synthesized halogen derivatives of 1H-pyrazolo[3,4-b]quinoline. The experimental study and quantum chemical simulations are presented. In optical poling experiment (fundamental wavelength k = 1.76 lm) we have found that the maximal output of second-order susceptibility (up to 1.53 pm/V) is observed for the chromophore possessing two methyl groups and fluor. The quantum chemical analysis reveals similarity in the absorption spectra of methyl-containing halogen derivatives which are characterized by four or five strong absorption bands in the spectral range 200–500 nm. A substitution of the methyl groups by the phenyl group causes the substantial changes of the absorption spectra mainly in the spectral range 240–370 nm. The comparison of measured and calculated absorption spectra manifests rather good agreement, namely in the part regarding the spectral positions of the first oscillator (absorption threshold). At the same time, the measured spectra reveal the considerable broadening almost of all absorption bands and even complete damping some of them in the case of phenyl derivatives. Semi-empirical PM3 method demonstrate substantially better agreement with the experimental values compared to the AM1 method. The lowest magnitude of the nonlinear optical susceptibility is revealed for Br-containing [PQ]-derivative. It is assumed that Br leads here to a suppressing of the charge transfers mechanism what is the reason for a relatively low nonlinear optical efficiency. 2004 Elsevier B.V. All rights reserved.
*
Corresponding authors. Tel.: +48601504268; fax: +48223612228. E-mail addresses: [email protected] (I.V. Kityk), [email protected] (A.V. Kityk). 1 Present address: Solid State Department, Institute of Physics WSP Czestochowa, Al.Armii Krajowej 1315, Czestochowa 42201, Poland. 0030-4018/$ - see front matter 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.optcom.2004.09.012
402
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
PACS: 33.20.Lg; 33.20.Kf; 33.70.w; 42.70.Ky Keywords: Organic materials; Visible and ultraviolet spectra; Optical poling; Optical second harmonic generation
1. Introduction 1H-pyrazolo[3,4-b]quinoline and its derivatives are known as a class of highly fluorescent materials in the blue spectral range [1] as well as promising materials for electroluminescent applications [2,3]. The fundamental optical properties of these materials significantly depend on the pattern of substitution as well as on the type of substituents. In particular, the results of experimental studies and quantum chemical simulation of the absorption spectra of methyl- and/or phenyl-containing derivatives 1H-pyrazolo[3,4-b]quinoline have been presented in several our recent publications (see e.g. [4,5]). It is shown that a substitution of the methyl groups by at least one phenyl group causes the drastic changes of the absorption spectra mainly in the spectral range 240–370 nm. These differences are attributed to additional molecular double bounded segments C@C of the substituted phenyl groups i.e. to p ! p* conjugated transitions. A comparison of measured and the calculated absorption spectra manifests quite satisfactory agreement for all the compounds in the part regarding the spectral position of the first oscillator (absorption threshold). At the same time, the measured spectra demonstrate the considerable broadening almost of all the absorption bands and even complete damping some of them in the case of phenyl derivatives. From the other hand, the experiments performed with highly and weakly polar organic solvents [5] show that the solvatochromic effect on the absorption spectra is indeed small. For this reason, the discrepancies between the calculated and the measured spectra have been attributed mainly to electron-vibronic coupling as well as to rotational dynamics of aromatic groups. In this paper, we present the results of experimental studies and semi-empirical quantum chemical simulations (PM3 and AM1 methods) performed on a new synthesized halogen derivatives of 1H-pyrazolo[3,4-b]quinoline (referred
hereafter as [PQ]): 6-fluoro-1,3-dimethyl-1H-pyrazolo[3,4-b]quinoline (6-fluoro-1,3-dimethyl-[PQ]), 6-bromo-1,3-dimethyl-1H-pyrazolo [3,4-b]quinoline (6-bromo-1,3-dimethyl-[PQ]), 1,3-dimethyl-7trifluoromethyl-1H-pyrazolo[3,4-b]quinoline (1,3dimethyl-7-trifluoromethyl-[PQ]), 6-fluoro-3-methyl1-phenyl-1H-pyrazolo[3,4-b]quinoline (6-fluoro-3methyl-1-phenyl-[PQ]), 7-trifluoromethyl-3-methyl1-phenyl-1H-pyrazolo[3,4-b]quinoline (7-trifluoromethyl-3-methyl-1-phenyl-[PQ]). Optical absorption and optical poling have been chosen as appropriate methods to study the basic opto-electronic properties of these materials. The absorption spectra in UV and visible regions give an important information about interacting p-electrons which obviously are also responsible for the relevant luminescent and electroluminescent properties of the organic materials. Optical poling is a relatively new and quite effective technique providing an important information about second-order optical properties of disordered materials like organic solutions or organic molecules incorporated into a polymer matrix (see e.g. [6,7]). Two coherent optical beams possessing different frequencies (x and 2x) lead to the appearance of reversible long-lived static polarization inside the medium [6]. The magnitude of medium polarization is determined [6,8,9] by the formation of a spatially periodic electrostatic field. Such effect is actually called as optical poling or v(2) optical grating. Due to the loss of the center of symmetry there appear the properties of a non-centrosymmetric uniaxial crystal with a photoinduced optical axes determined by light polarization of the pumping beam. As a result, the optically induced second harmonic generation (SHG) and parametric amplification become possible in such perturbed medium. All optical poling of organic materials is well studied procedure [10]. However particular interest present the chromophore possessing simultaneously good luminescent properties and which are incorporated in the host polymer matrices. This
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
is very crucial for creation of the multifunctional materials possessing simultaneously v(2)-grating as well as the luminescent properties. Contrary to the organic liquid solvents which are used usually for the optical poling in this case a substantial role begin to play intermolecular interaction between the chromophore and polymer. So, comparing the evaluations of the hyperpolarizability performed for the individual molecule and incorporated inside the polymer could give an answer concerning the role of the matrix (in our case PMMA) in the properties observed. The goal of the present work is to study how the halogen substituents influence the photoinduced SHG output in 1H-pyrazolo[3,4-b]quinoline derivatives. The main goal here is to search correlation between the experimental data and quantum chemical simulation, in particular regarding the correlation between the measured macroscopic second-order optical susceptibility and calculated static molecular dipoles as well as calculated second-order nonlinear optical molecular susceptibilities. The experimental results are therefore supplied by quantum-chemical semi-empirical modeling for molecules in vacuum. Dealing with this approach, the quantum chemical PM3- and AM1-method have been found as the most appropriate for simulation of the experimental data.
2. Synthesis and experimental Depending on lateral substituents, several chemical compounds mentioned below were used for the synthesis of a particular [PQ]-derivative. All reagents were used as received from Aldrich or Fluka without further purification. Column chromatography was performed on Merck silica gel 60 (230–400 mesh) using toluene/ethyl acetate (3:1) as eluent. All the [PQ]-derivatives 3 were prepared according to procedures described in [11–14] by condensation of substituted aromatic amines 2 with 5-chloro-1,3-dimethyl- and 5-chloro-3-methyl-1-phenyl-1H-pyrazole-4-carbaldehydes 1 (see Scheme 1 and Table 1). As the result, the following compounds were synthesized for further optical studies: 6-fluoro-1,3-dimethyl-[PQ], 6-bromo-1, 3-dimethyl-[PQ], 1,3-dimethyl-7-trifluoromethyl-
403
Scheme 1. Synthesis of the 1H-pyrazolo[3,4-b]quinoline ([PQ]-) derivatives. Conditions: at the temperature of 140–190 C within 30–60 min. The radicals R1, R2 and R3 are mentioned in the Table 1.
Table 1 Supplement to the Scheme 1: chemical radicals R1, R2 and R3 Compound
R1
R2
R3
6-fluoro-1,3-dimethyl-[PQ] 7-trifluoromethyl-3-methyl-1-phenyl-[PQ] 6-bromo-1,3-dimethyl-[PQ] 6-fluoro-3-methyl-1-phenyl-[PQ] 1,3-dimethyl-7-trifluoromethyl-[PQ]
Me Ph Me Ph Me
Me Me Me Me Me
6-F 7-CF3 6-Br 6-F 7-CF3
Here, [PQ] „ 1H-pyrazolo[3,4-b]quinoline.
[PQ], 6-fluoro-3-methyl-1-phenyl-[PQ] and 7-trifluoromethyl-3-methyl-1-phenyl-[PQ]. The purity of these compounds was checked by TLC (Merck). The optical absorption spectra were recorded in tetrahydrofuran solution (mass concentration of about 0.1%) using Shimadzu UV–VIS 2101 scanning spectrophotometer in range 200–500 nm. The measurements were performed using standard 1 cm path length quartz cuvette for absorption spectrometry. The measurements of the optical second-order susceptibility have been performed in standard optical poling experiments (see setup shown in the Fig. 1). As a fundamental laser source we have used Gd:YAG-laser generating at k = 1.76 lm with pulse power P = 21 MW and pulse duration about 30 ps. The laser beam is consequently split into two coherent beams by the beamsplitter BS. The second beam is converted by b-BBO crystal to a doubled-frequency signal (k = 0.88 lm). Using two-channel scheme we achieved the interaction of two coherent waves. We have found that coherence length is equal to about 28–36 lm. The observed output signal is detected by photomultiplier PM. Evaluation of the effective second-order
404
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
Fig. 1. The setup in optical poling experiment. BS is the beamsplitter; P1, P2 and P3 are the polarizers; S is the sample; F is the filter; PM is the photomultiplier; BC is the electronic boxcar.
susceptibility was done using a simple procedure described in the [15]. The investigated chromophores were incorporated into the PMMA polymer matrices with weight concentration of about 5.5% and their space homogeneity was equal about 2.1%.
3. Calculation procedure The calculation of optical absorption spectra, static dipole moments and molecular second-order nonlinear constants have been performed using the molecular dynamics and quantum chemical package Hyperchem 7.5. The geometrical optimization has been carried out by (MM+) force field method. Together with semi-empirical AM1 and PM3 methods this model is known to be as the most general and frequently used for molecular mechanics calculations developed particularly for organic molecules. The calculations of optical absorption spectra and nonlinear susceptibility were consequently performed within several semi-empirical quantum chemical models available in this software. However, the results will be presented for AM1 and parameterized PM3 methods, only. These methods indeed give the best agreement with the experiment. The optical spectra were calculated considering only the singly excited configuration interactions (CI); the excitation energies have been limited in both methods by the orbital criterion, i.e. to 12 occupied and 12 unoccupied orbitals. The intensity of spectral optical absorption I(x) is determined by the following expression (for details see e.g. [4,5]):
IðxÞ x
n X X
~r j Wki ij2 j hWji j ihr
k¼1 i¼x;y;z
h2 ðx xjk Þ2 þ ðC=2Þ2
;
ð1Þ
where Wj and Wk are wave functions of jth ground ~r is state and kth excited state energy levels and ihr a transition dipole momentum operator of molecule, hxjk = (EjEk) is the energy difference between the ground and excited states, x is the frequency of incident electromagnetic wave. Taking into account that the integration is performed over the space volume the intensity of spectral absorption may be expressed as: IðxÞ x
i
2
n X X
j ðlÞjk j
k¼1 i¼x;y;z
ðx xjk Þ þ ðC=2hÞ
2
2
;
ð2Þ
i ~r j Wki i j are the correspondwhere ðlÞjk ¼j hWji j r ing transition dipole moments, parameter C is related inversely with relaxation times. The resonance frequencies xjk and transition dipole momentums ðlÞijk have been obtained directly within PM3 and AM1 procedures, as described above. The specific features of these two methods consists in the fitting of semi-empirical parameters to the dipole moments, which play crucial role in the optical and nonlinear optical properties. The empirical parameter C was chosen at reasonable value of about 0.12 eV, what gives the best agreement in contourshape of calculated optical absorption spectra compared to experimental data. It is assumed that in such type of compounds parameter C is almost constant for all absorption lines whereas the broadening does not influence on their spectral positions [4]. The molecular nonlinear hyperpolarizabilities bijk have been calculated using the model of socalled multi-photon excitation [16]
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
bijk ¼
405
X h j 1 ðlgn0 M in0 n lkng þ lkgn0 M in0 n ljng Þ 2 8h n0 ;n;n0 6¼n6¼g 1 1 þ ðxn0 g xÞðxng þ xÞ ðxn0 g þ xÞðxng xÞ
þðlign0 M jn0 n lkng þ lign0 M kn0 n ljng Þ 1 1 þ ðxn0 g þ 2xÞðxng þ xÞ ðxn0 g 2xÞðxng xÞ þ ðljgn0 M kn0 n ling þ lkgn0 M jn0 n ling Þ 1 1 þ ðxn0 g xÞðxng þ 2xÞ ðxn0 g þ xÞðxng 2xÞ ð3Þ
where indices gn(gn 0 ) correspond to transitions from the ground state g to excited state n(n 0 ); xn 0 g and xng are the corresponding transition frequenðx;y;zÞ ðx;y;zÞ ðx;y;zÞ ðx;y;zÞ cies; M n0 n ¼ ðM gn0 M ðx;y;zÞ Þ, M n0 n and M n0 n gn are the excited state dipole momentums. Since the molecules in the polymer matrix are randomly oriented the calculated second-order susceptibilities will be presented by its average (effective) magqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi nitude bav ¼
ðb2xxx þ b2yyy þ b2zzz Þ=3.
4. Experimental results and discussion Measured absorption spectra of 6-fluoro-1,3-dimethyl-[PQ], 6-bromo-1,3-dimethyl-[PQ], 1,3-dimethyl-7-trifluoromethyl-[PQ], 6-fluoro-3-methyl1-phenyl-[PQ] and 7-trifluoromethyl-3-methyl-1phenyl-[PQ] are presented in the Figs. 2(a), 3(a), 4(a), 5(a) and 6(a), respectively. The calculated absorption spectra (C = 0.12 eV) are shown in the Figs. 2(b)–6(b). One can see that absorption spectra of 6-fluoro-1,3-dimethyl-[PQ] (Fig. 2(b)), 6-bromo-1,3-dimethyl-[PQ] (Fig. 3(b)) and 1,3-dimethyl-7-trifluoromethyl-[PQ] (Fig. 4(b)) calculated within the AM1-(PM3-) methods are rather similar. They can be characterized generally by five or six relatively strong absorption bands. In particular one obtains for 6-fluoro-1,3-dimethyl-[PQ]: 381(398.5), 360.5(359), 287(294), (256), 245(243), 224(224) nm; for 6-bromo-1,3-dimethyl-[PQ]: 374.5(394), 355(355), 284(282), (254), 246(243), 223(223) nm and for 1,3-dimethyl-7-trifluoromethyl-[PQ]: 379.5(402.5), 355.5(358), 242(256), 239(242), 221.5(223). The corresponding experimental spectra (see Figs. 2(a), 3(a) and 4(a)) have
Fig. 2. Measured absorption spectra of 6-fluoro-1,3-dimethyl[PQ] (a) and calculated spectra within the semi-empirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 6-fluoro-1,3-dimethyl-[PQ] obtained within the geometrical optimization procedure.
in all the cases an absorption threshold at about 399 nm which appears to be in fairly good agreement with quantum chemical simulations, namely for the spectra calculated by PM3-method. It is also amazing that spectral positions of the most strongest broad absorption band at about 240– 250 nm in the measured spectra well coincide with the calculated spectra. Both quantum-chemical methods give here sufficiently good agreement. A substitution of the methyl groups by the phenyl group (see Figs. 5 and 6) causes the substantial changes of the calculated spectra. In particular, quantum chemistry simulations reveal several additional relatively strong absorption bands within the spectral range 240–370 nm. This feature is common for all the [PQ]-derivatives containing phenyl groups (see e.g. [4,5]). An appropriate interpretation of their origin obviously refers to additional molecular doubled bonding p-conjugated segments
406
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
Fig. 3. Measured absorption spectra of 6-bromo-1,3-dimethyl[PQ] (a) and calculated spectra within the semi-empirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 6-bromo-1,3-dimethyl-[PQ] obtained within the geometrical optimization procedure.
C@C of the substituted phenyl groups. The corresponding absorption bands may be attributed to p ! p* electronic transition from the bonding to anti-bonding molecular configuration. Total spectra of this organic compound contain eight or nine relatively intense bands as calculated within AM1-(PM3-) procedure, in particular, one obtains for 6-fluoro-3-methyl-1-phenyl-[PQ]: 385.5(410.5), 362(368.5), 315(316), 284(288), 254(260), 245(246.5), 238.5(229) 229.5(217) and 210.5 nm and for 7-trifluoromethyl-3-methyl-1-phenyl-[PQ]: 387.5(417), 357.5(366), 310.5(314.5), 285.5(288.5), 252.5(259), 243.5(247.5), 236(227.5), 227(217.5), 211.5(207). The measured absorption threshold for both these phenyl containing [PQ]-derivatives kg = 399 nm differs slightly from the one calculated by AM1(PM3) methods which give 385.5(410.5) and 387.5(417) nm, respectively. In addition, the experimental spectra in UV region are characterized by only two broad bands. The most strongest
Fig. 4. Measured absorption spectra of 1,3-dimethyl-7-trifluoromethyl-[PQ] (a) and calculated spectra within the semiempirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 1,3-dimethyl-7-trifluoromethyl[PQ] obtained within the geometrical optimization procedure.
band exists at about 272 nm, i.e. it is clearly red shifted for both compounds with respect to the spectral positions of the series of most intense absorption bands obtained by the quantum chemical simulations. The reasons for the discrepancies observed between the measured and calculated spectra of [PQ]-derivatives have been discussed already in our recent publications [4,5]. Initially [4] they were attributed to the solvatochromic effect [17]. However, the experiments with strongly and weakly polar solvents [5] have indicated indeed rather small its influence on the absorption spectra. All the solutions clearly have shown only a weak hypsochromic (blue) shift of kg as the solvent polarity increased. The strongest one was observed on about 4 nm only. Broadened spectra as well as observed band position shifts are, therefore, rather a result of two other eventual mechanisms which are not taken into account in quantum-chemical calcula-
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
407
Fig. 5. Measured absorption spectra of 6-fluoro-3-methyl-1phenyl-[PQ] (a) and calculated spectra within the semi-empirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 6-fluoro-3-methyl-1-phenyl-[PQ] obtained within the geometrical optimization procedure.
Fig. 6. Measured absorption spectra of 7-trifluoromethyl-3methyl-1-phenyl-[PQ] (a) and calculated spectra within the semi-empirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 7-trifluoromethyl-3methyl-1-phenyl-[PQ] obtained within the geometrical optimization procedure.
tions. One of those mechanisms is its broadening due to the electron-vibronic coupling. Indeed the lowest energy absorption band reveals features very similar to vibronic replica (see e.g. Fig. 2(a) and Fig. 3(a)). The measurements on a frozen solution or by cooling a disperse solid solution (e.g. polymer matrix) would made these features more obvious. The second reason is much related to the derivatives containing methyl, phenyl or CF3 groups. One should be emphasized that our quantum-chemical calculations have been performed for isolated molecules in vacuum. The molecular optimization procedure gives here a planar molecular configuration for all [PQ]-derivatives. That is unlikely to be quite correct for a solution phase assuming just that the methyl, phenyl or CF3 groups are presumably to rotate dynamically affecting the symmetry selection rules for different electronic transitions.
Fig. 7 shows the experimental efficiency of second-order optical susceptibility d vs photoinduced power density Ip for each chromophore. Table 2 presents the saturated second order optical susceptibility deff (as obtained from the data in the Fig. 7) as well as the dipole moments j ~ p j, effective (averaged) molecular optical nonlinear constant bav and excited state dipole moments qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðxÞ2
ðyÞ2
ðzÞ2
M g ¼ M g1 þ M g1 þ M g1 corresponding to the first active oscillator (as calculated within semi-empirical AM1 i PM3 methods). The experiment shows that maximally achieved second-order optical susceptibility (up to 1.53 pm/V) appears for the chromophore containing two methyl groups and fluor. In the phenyl-containing derivatives the second-order susceptibilities are substantially lower. It may be a consequence of competition between the charge transfer within two backside groups. However, the lowest magnitude of the
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
408
Fig. 7. Efficiency of second-order optical susceptibility deff vs photoinduced power density. Comp.1: 6-fluoro-1,3-dimethyl[PQ]; Comp.2: 7-trifluoromethyl-3-methyl-1-phenyl-[PQ]; Comp.3: 6-bromo-1,3-dimethyl-[PQ]; Comp.4: 6-fluoro-3-methyl-1-phenyl-[PQ]; Comp.5: 1,3-dimethyl-7-trifluoromethyl[PQ].
nonlinear optical susceptibility has been observed for 6-bromo-1,3-dimethyl-[PQ]. The presence of Br leads here to a suppressing of the charge transfer. One must also mention that studied composites lost the output second-order effective susceptibility not more than 24% which is comparable with other organic materials, e.g. such as cyan derivatives of stilbene [18] for which the second-order susceptibility reaches the magnitudes up to 1 pm/V. Several other important notes would be useful in the interpretation of nonlinear optical properties of polymer matrices doped by organic [PQ]-derivatives. Since we deal with the optical poling effect an important role plays the static polarization created by two coherent waves. Accordingly,
a substantial meaning in this case has the magnitude of the excited state dipole moments (not only in the occupied ground state). From the Table 2 one can conclude that quantum chemical PM3 method gives again better agreement with the experimental data compared to AM1 method. Several discrepancies between the values likely reflect specific features of the local Lorentz field related to macroscopic hyperpolarizabilities and macroscopic susceptibilities. To understand them better a simple model for optical grating formation would be rather useful. Generally, the optical polarizability can be presented by the linear and nonlinear terms L NL P~i ¼ P~i þ P~i ; L ðxÞ P~i ¼ aij Ej ;
P~i
NL
ðxÞ
ð4Þ ðxÞ
ðxÞ
ðxÞ
ðxÞ
¼ bijk Ej Ek þ cijkl Ej Ek El :
The microscopic susceptibilities aij, bijk, cijkl are related with the macroscopic susceptibility by the equation ðxÞ
ðxÞ ðxÞ
ðxÞ
ðxÞ ðxÞ ðxÞ
ðxÞ
vij ¼ Li Lj aij ; vijk ¼ Li Lj Lk bijk ; vijkl ðxÞ ðxÞ ðxÞ ðxÞ
¼ Li Lj Lk Lk cijkl ;
ð5Þ
LðxÞ z
where ¼ ½ðn2z 2Þ=ðn2z þ 3Þ=q are the compoðxÞ nents (z ” i,j,k) of the local Lorentz field and vijk is the macroscopic second-order nonlinear susceptibility responsible for the photoinduced SHG. The later one is very crucial in the case of guest–host polymers what may be a main reason of discrepancies between the local hyperpolarizabilities and macroscopic optical susceptibilities. Because the quantum-chemical calculations of Lorentz factor are
Table 2 Saturated second-order optical susceptibility deff (experiment) and calculated dipole moment j ~ p j, effective (averaged) molecular optical nonlinear constant bav and excited state dipole moments Mg Compound
Comp.2 Comp.3 Comp.4 Comp.5
deff
j~ pj
j~ pj
bav
bav
Mg
Mg
[pm/V] (exp.) 1.53 1.21 1.15 1.27 1.51
[D] (AM1) 0.596 3.555 0.842 1.084 3.184
[D] (PM3) 0.461 2.927 0.230 0.439 2.738
[au] (AM1) 722.8 1605 451.4 133.2 504.1
[au] (PM3) 682.4 134.1 131.1 182.2 392.3
[D] (AM1) 1.630 4.374 2.070 1.807 4.346
[D] (PM3) 3.796 5.694 3.516 4.215 5.543
Comp.1: 6-fluoro-1,3-dimethyl-[PQ]; Comp.2: 7-trifluoromethyl-3-methyl-1-phenyl-[PQ]; Comp.3: 6-bromo-1,3-dimethyl-[PQ]; Comp.4: 6-fluoro-3-methyl-1-phenyl-[PQ]; Comp.5: 1,3-dimethyl-7-trifluoromethyl-[PQ].
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
too complicate a new theoretical formalism should be developed which is a subject for a separate work. The results presented above appear in some contradiction to the generally adopted conception that the optical poling is associated only with the third order hyperpolarizibilities. The performed measurements and quantum chemical simulation indeed indicate on a substantial role of the second-order hyperpolarizibility as well.
409
ter agreement with the experiment compared to the AM1 method. The results of performed measurements and quantum chemical simulations contradict to the generally adopted conception that the optical poling is associated only with the third order hyperpolarizibilities. A substantial role of the second-order hyperpolarizibility is presented here as well. Acknowledgement
5. Conclusion The optical poling experiment shows, that maximally achieved second-order nonlinear optical susceptibility (up to 1.53 pm/V) appears for the chromophore containing two methyl groups and fluor as substituents. In the phenyl-containing derivatives the nonlinear second-order susceptibility is substantially lower. It may be a consequence of competition between the charge transfer within two backside groups. The lowest magnitude of the nonlinear optical susceptibility is revealed for Brcontaining [PQ]-derivative. It is assumed that Br leads here to a suppressing of the charge transfers mechanism what is the reason for a relatively low nonlinear optical efficiency. We present also the results of experimental studies and quantum chemical simulations of optical absorption and optical poling performed on a new synthesized halogen derivatives of 1H-pyrazolo[3,4-b]quinoline. The quantum chemical analysis reveals similarity in the absorption spectra of metyl-containing halogen derivatives which are characterized by four or five strong absorption bands in the spectral range 200–500 nm. A substitution of the methyl groups by the phenyl group causes the substantial changes of the absorption spectra mainly in the spectral range 240–370 nm. We attribute these differences to additional molecular double bonding segments C@C of the substituted phenyl groups, thereby the electronic p ! p* transitions appear to be involved into the absorption process. The discrepancies between the calculated and the measured spectra are attributed to electron-vibronic coupling as well as to specific rotational dynamics of phenyl rings. For all compounds the semi-empirical PM3 method gives bet-
This work was supported by KBN Grant No. 3-T11-B-074-26. References [1] Z. He, G.H.W. Milburn, K.J. Baldwin, D.A. Smith, A. Danel, P. Tomasik, J. Lumin. 86 (2000) 1. [2] Z. He, G.H.W. Milburn, A. Danel, A. Puchala, P. Tomasik, D. Rasala, J. Mater. Chem. 7 (1997) 2323. [3] A. Danel, Z. He, G.H.W. Milburn, P. Tomasik, J. Mater. Chem. 9 (1999) 339. [4] E. Koscien, J. Sanetra, E. Gondek, A. Danel, A. Wisla., A.V. Kityk, Opt. Commun. 227 (2003) 115. [5] E. Gondek, E. Koscien, J. Sanetra, A. Danel, A. Wisla, A.V. Kityk, Spectrochimica Acta, Part A, (to appear). [6] M.K. Balakirev, I.V. Kityk, V.A. Smirnov, L.I. Vostrikova, J. Ebothe, Phys. Rev. A 67 (2003) 023806. [7] I.V. Kityk, M. Makowska-Janusik, E. Gondek, L. Krzeminska, A. Danel, K.J. Plucinski, S. Benet, B. Sachraoui, J. Phys.: Cond. Matter 16 (2004) 231. [8] B.P. Antonyuk, Opt. Commun. 174 (2000) 427. [9] B.P. Apostoluk, D. Chapron, G. Godret, B. Sahraoui, J.M. Nunzi, C. Fiorini-Debuichat, P. Raimond, Opt. Lett. 27 (2002) 2028. [10] F. Charra, F. Kaizar, J.M. Nunzi, P. Raimond, E. Idiart, Opt. Lett. 18 (1993) 941. [11] J. Becher, B.F. Hansen, N. Malhotra, J. Heterocycl. Chem. 28 (1991) 1837. [12] L. Hennig, T. Mu¨ller, M. Grosche, J. Prakt. Chem. 332 (1990) 693. [13] A. Danel, K. Chaczatrian, P. Tomasik, ARKIVOC 1 (2000) 51. [14] A. Danel, Synthesis 1H-pyrazolo[3,4-b]quinolines as a luminofores for LEDÕs, Ph.D Thesis, 1996. [15] I.V. Kityk, B. Marciniak, A. Mefleh, J. Phys. D: Appl. Phys. 34 (2001) 1. [16] R. Boyd, Principles of Nonlinear Optics, Academic Press, New York, 1983. [17] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, VCH Verlag, Germany, 1988. [18] I. Yan, L. Liu, L. Ji, M. Ye, L. Lu, W. Wang, J. Phys. D: Appl. Phys. 37 (2004) 1597.
Optical poling of several halogen derivatives of pyrazoloquinoline E. Kos´cien´ a, J. Sanetra b, E. Gondek b, B. Jarosz c, I.V. Kityk J. Ebothe e, A.V. Kityk a,*
d,*,1
,
a
e
Institute for Computer Science, Technical University of Czestochowa, Armii Krajowej 17, 42-200 Czestochowa, Poland b Institute of Physics, Technical University of Krakow, Podhorazych 1, 30-084 Krakow, Poland c Department of Chemistry, Hugon Kołłotaj Agricultural University, Al. Mickiewicza 24/28, 30-059 Krakow, Poland d Institute of Biology and Biophysics, Technical University of Czestochowa, PL-42200 Czestochowa, Poland Universite de Reims, UFR Sciences, Laboratoire de Microscopie et dÕEtude de Nanostructures, E.A.n 3799, Boite Postale 138, 21 Rue Clement Ader, F-51685 Reims Cedex 02, France Received 5 June 2004; received in revised form 31 August 2004; accepted 2 September 2004
Abstract Paper deals with optical absorption and second-order optical poling effect in a new synthesized halogen derivatives of 1H-pyrazolo[3,4-b]quinoline. The experimental study and quantum chemical simulations are presented. In optical poling experiment (fundamental wavelength k = 1.76 lm) we have found that the maximal output of second-order susceptibility (up to 1.53 pm/V) is observed for the chromophore possessing two methyl groups and fluor. The quantum chemical analysis reveals similarity in the absorption spectra of methyl-containing halogen derivatives which are characterized by four or five strong absorption bands in the spectral range 200–500 nm. A substitution of the methyl groups by the phenyl group causes the substantial changes of the absorption spectra mainly in the spectral range 240–370 nm. The comparison of measured and calculated absorption spectra manifests rather good agreement, namely in the part regarding the spectral positions of the first oscillator (absorption threshold). At the same time, the measured spectra reveal the considerable broadening almost of all absorption bands and even complete damping some of them in the case of phenyl derivatives. Semi-empirical PM3 method demonstrate substantially better agreement with the experimental values compared to the AM1 method. The lowest magnitude of the nonlinear optical susceptibility is revealed for Br-containing [PQ]-derivative. It is assumed that Br leads here to a suppressing of the charge transfers mechanism what is the reason for a relatively low nonlinear optical efficiency. 2004 Elsevier B.V. All rights reserved.
*
Corresponding authors. Tel.: +48601504268; fax: +48223612228. E-mail addresses: [email protected] (I.V. Kityk), [email protected] (A.V. Kityk). 1 Present address: Solid State Department, Institute of Physics WSP Czestochowa, Al.Armii Krajowej 1315, Czestochowa 42201, Poland. 0030-4018/$ - see front matter 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.optcom.2004.09.012
402
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
PACS: 33.20.Lg; 33.20.Kf; 33.70.w; 42.70.Ky Keywords: Organic materials; Visible and ultraviolet spectra; Optical poling; Optical second harmonic generation
1. Introduction 1H-pyrazolo[3,4-b]quinoline and its derivatives are known as a class of highly fluorescent materials in the blue spectral range [1] as well as promising materials for electroluminescent applications [2,3]. The fundamental optical properties of these materials significantly depend on the pattern of substitution as well as on the type of substituents. In particular, the results of experimental studies and quantum chemical simulation of the absorption spectra of methyl- and/or phenyl-containing derivatives 1H-pyrazolo[3,4-b]quinoline have been presented in several our recent publications (see e.g. [4,5]). It is shown that a substitution of the methyl groups by at least one phenyl group causes the drastic changes of the absorption spectra mainly in the spectral range 240–370 nm. These differences are attributed to additional molecular double bounded segments C@C of the substituted phenyl groups i.e. to p ! p* conjugated transitions. A comparison of measured and the calculated absorption spectra manifests quite satisfactory agreement for all the compounds in the part regarding the spectral position of the first oscillator (absorption threshold). At the same time, the measured spectra demonstrate the considerable broadening almost of all the absorption bands and even complete damping some of them in the case of phenyl derivatives. From the other hand, the experiments performed with highly and weakly polar organic solvents [5] show that the solvatochromic effect on the absorption spectra is indeed small. For this reason, the discrepancies between the calculated and the measured spectra have been attributed mainly to electron-vibronic coupling as well as to rotational dynamics of aromatic groups. In this paper, we present the results of experimental studies and semi-empirical quantum chemical simulations (PM3 and AM1 methods) performed on a new synthesized halogen derivatives of 1H-pyrazolo[3,4-b]quinoline (referred
hereafter as [PQ]): 6-fluoro-1,3-dimethyl-1H-pyrazolo[3,4-b]quinoline (6-fluoro-1,3-dimethyl-[PQ]), 6-bromo-1,3-dimethyl-1H-pyrazolo [3,4-b]quinoline (6-bromo-1,3-dimethyl-[PQ]), 1,3-dimethyl-7trifluoromethyl-1H-pyrazolo[3,4-b]quinoline (1,3dimethyl-7-trifluoromethyl-[PQ]), 6-fluoro-3-methyl1-phenyl-1H-pyrazolo[3,4-b]quinoline (6-fluoro-3methyl-1-phenyl-[PQ]), 7-trifluoromethyl-3-methyl1-phenyl-1H-pyrazolo[3,4-b]quinoline (7-trifluoromethyl-3-methyl-1-phenyl-[PQ]). Optical absorption and optical poling have been chosen as appropriate methods to study the basic opto-electronic properties of these materials. The absorption spectra in UV and visible regions give an important information about interacting p-electrons which obviously are also responsible for the relevant luminescent and electroluminescent properties of the organic materials. Optical poling is a relatively new and quite effective technique providing an important information about second-order optical properties of disordered materials like organic solutions or organic molecules incorporated into a polymer matrix (see e.g. [6,7]). Two coherent optical beams possessing different frequencies (x and 2x) lead to the appearance of reversible long-lived static polarization inside the medium [6]. The magnitude of medium polarization is determined [6,8,9] by the formation of a spatially periodic electrostatic field. Such effect is actually called as optical poling or v(2) optical grating. Due to the loss of the center of symmetry there appear the properties of a non-centrosymmetric uniaxial crystal with a photoinduced optical axes determined by light polarization of the pumping beam. As a result, the optically induced second harmonic generation (SHG) and parametric amplification become possible in such perturbed medium. All optical poling of organic materials is well studied procedure [10]. However particular interest present the chromophore possessing simultaneously good luminescent properties and which are incorporated in the host polymer matrices. This
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
is very crucial for creation of the multifunctional materials possessing simultaneously v(2)-grating as well as the luminescent properties. Contrary to the organic liquid solvents which are used usually for the optical poling in this case a substantial role begin to play intermolecular interaction between the chromophore and polymer. So, comparing the evaluations of the hyperpolarizability performed for the individual molecule and incorporated inside the polymer could give an answer concerning the role of the matrix (in our case PMMA) in the properties observed. The goal of the present work is to study how the halogen substituents influence the photoinduced SHG output in 1H-pyrazolo[3,4-b]quinoline derivatives. The main goal here is to search correlation between the experimental data and quantum chemical simulation, in particular regarding the correlation between the measured macroscopic second-order optical susceptibility and calculated static molecular dipoles as well as calculated second-order nonlinear optical molecular susceptibilities. The experimental results are therefore supplied by quantum-chemical semi-empirical modeling for molecules in vacuum. Dealing with this approach, the quantum chemical PM3- and AM1-method have been found as the most appropriate for simulation of the experimental data.
2. Synthesis and experimental Depending on lateral substituents, several chemical compounds mentioned below were used for the synthesis of a particular [PQ]-derivative. All reagents were used as received from Aldrich or Fluka without further purification. Column chromatography was performed on Merck silica gel 60 (230–400 mesh) using toluene/ethyl acetate (3:1) as eluent. All the [PQ]-derivatives 3 were prepared according to procedures described in [11–14] by condensation of substituted aromatic amines 2 with 5-chloro-1,3-dimethyl- and 5-chloro-3-methyl-1-phenyl-1H-pyrazole-4-carbaldehydes 1 (see Scheme 1 and Table 1). As the result, the following compounds were synthesized for further optical studies: 6-fluoro-1,3-dimethyl-[PQ], 6-bromo-1, 3-dimethyl-[PQ], 1,3-dimethyl-7-trifluoromethyl-
403
Scheme 1. Synthesis of the 1H-pyrazolo[3,4-b]quinoline ([PQ]-) derivatives. Conditions: at the temperature of 140–190 C within 30–60 min. The radicals R1, R2 and R3 are mentioned in the Table 1.
Table 1 Supplement to the Scheme 1: chemical radicals R1, R2 and R3 Compound
R1
R2
R3
6-fluoro-1,3-dimethyl-[PQ] 7-trifluoromethyl-3-methyl-1-phenyl-[PQ] 6-bromo-1,3-dimethyl-[PQ] 6-fluoro-3-methyl-1-phenyl-[PQ] 1,3-dimethyl-7-trifluoromethyl-[PQ]
Me Ph Me Ph Me
Me Me Me Me Me
6-F 7-CF3 6-Br 6-F 7-CF3
Here, [PQ] „ 1H-pyrazolo[3,4-b]quinoline.
[PQ], 6-fluoro-3-methyl-1-phenyl-[PQ] and 7-trifluoromethyl-3-methyl-1-phenyl-[PQ]. The purity of these compounds was checked by TLC (Merck). The optical absorption spectra were recorded in tetrahydrofuran solution (mass concentration of about 0.1%) using Shimadzu UV–VIS 2101 scanning spectrophotometer in range 200–500 nm. The measurements were performed using standard 1 cm path length quartz cuvette for absorption spectrometry. The measurements of the optical second-order susceptibility have been performed in standard optical poling experiments (see setup shown in the Fig. 1). As a fundamental laser source we have used Gd:YAG-laser generating at k = 1.76 lm with pulse power P = 21 MW and pulse duration about 30 ps. The laser beam is consequently split into two coherent beams by the beamsplitter BS. The second beam is converted by b-BBO crystal to a doubled-frequency signal (k = 0.88 lm). Using two-channel scheme we achieved the interaction of two coherent waves. We have found that coherence length is equal to about 28–36 lm. The observed output signal is detected by photomultiplier PM. Evaluation of the effective second-order
404
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
Fig. 1. The setup in optical poling experiment. BS is the beamsplitter; P1, P2 and P3 are the polarizers; S is the sample; F is the filter; PM is the photomultiplier; BC is the electronic boxcar.
susceptibility was done using a simple procedure described in the [15]. The investigated chromophores were incorporated into the PMMA polymer matrices with weight concentration of about 5.5% and their space homogeneity was equal about 2.1%.
3. Calculation procedure The calculation of optical absorption spectra, static dipole moments and molecular second-order nonlinear constants have been performed using the molecular dynamics and quantum chemical package Hyperchem 7.5. The geometrical optimization has been carried out by (MM+) force field method. Together with semi-empirical AM1 and PM3 methods this model is known to be as the most general and frequently used for molecular mechanics calculations developed particularly for organic molecules. The calculations of optical absorption spectra and nonlinear susceptibility were consequently performed within several semi-empirical quantum chemical models available in this software. However, the results will be presented for AM1 and parameterized PM3 methods, only. These methods indeed give the best agreement with the experiment. The optical spectra were calculated considering only the singly excited configuration interactions (CI); the excitation energies have been limited in both methods by the orbital criterion, i.e. to 12 occupied and 12 unoccupied orbitals. The intensity of spectral optical absorption I(x) is determined by the following expression (for details see e.g. [4,5]):
IðxÞ x
n X X
~r j Wki ij2 j hWji j ihr
k¼1 i¼x;y;z
h2 ðx xjk Þ2 þ ðC=2Þ2
;
ð1Þ
where Wj and Wk are wave functions of jth ground ~r is state and kth excited state energy levels and ihr a transition dipole momentum operator of molecule, hxjk = (EjEk) is the energy difference between the ground and excited states, x is the frequency of incident electromagnetic wave. Taking into account that the integration is performed over the space volume the intensity of spectral absorption may be expressed as: IðxÞ x
i
2
n X X
j ðlÞjk j
k¼1 i¼x;y;z
ðx xjk Þ þ ðC=2hÞ
2
2
;
ð2Þ
i ~r j Wki i j are the correspondwhere ðlÞjk ¼j hWji j r ing transition dipole moments, parameter C is related inversely with relaxation times. The resonance frequencies xjk and transition dipole momentums ðlÞijk have been obtained directly within PM3 and AM1 procedures, as described above. The specific features of these two methods consists in the fitting of semi-empirical parameters to the dipole moments, which play crucial role in the optical and nonlinear optical properties. The empirical parameter C was chosen at reasonable value of about 0.12 eV, what gives the best agreement in contourshape of calculated optical absorption spectra compared to experimental data. It is assumed that in such type of compounds parameter C is almost constant for all absorption lines whereas the broadening does not influence on their spectral positions [4]. The molecular nonlinear hyperpolarizabilities bijk have been calculated using the model of socalled multi-photon excitation [16]
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
bijk ¼
405
X h j 1 ðlgn0 M in0 n lkng þ lkgn0 M in0 n ljng Þ 2 8h n0 ;n;n0 6¼n6¼g 1 1 þ ðxn0 g xÞðxng þ xÞ ðxn0 g þ xÞðxng xÞ
þðlign0 M jn0 n lkng þ lign0 M kn0 n ljng Þ 1 1 þ ðxn0 g þ 2xÞðxng þ xÞ ðxn0 g 2xÞðxng xÞ þ ðljgn0 M kn0 n ling þ lkgn0 M jn0 n ling Þ 1 1 þ ðxn0 g xÞðxng þ 2xÞ ðxn0 g þ xÞðxng 2xÞ ð3Þ
where indices gn(gn 0 ) correspond to transitions from the ground state g to excited state n(n 0 ); xn 0 g and xng are the corresponding transition frequenðx;y;zÞ ðx;y;zÞ ðx;y;zÞ ðx;y;zÞ cies; M n0 n ¼ ðM gn0 M ðx;y;zÞ Þ, M n0 n and M n0 n gn are the excited state dipole momentums. Since the molecules in the polymer matrix are randomly oriented the calculated second-order susceptibilities will be presented by its average (effective) magqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi nitude bav ¼
ðb2xxx þ b2yyy þ b2zzz Þ=3.
4. Experimental results and discussion Measured absorption spectra of 6-fluoro-1,3-dimethyl-[PQ], 6-bromo-1,3-dimethyl-[PQ], 1,3-dimethyl-7-trifluoromethyl-[PQ], 6-fluoro-3-methyl1-phenyl-[PQ] and 7-trifluoromethyl-3-methyl-1phenyl-[PQ] are presented in the Figs. 2(a), 3(a), 4(a), 5(a) and 6(a), respectively. The calculated absorption spectra (C = 0.12 eV) are shown in the Figs. 2(b)–6(b). One can see that absorption spectra of 6-fluoro-1,3-dimethyl-[PQ] (Fig. 2(b)), 6-bromo-1,3-dimethyl-[PQ] (Fig. 3(b)) and 1,3-dimethyl-7-trifluoromethyl-[PQ] (Fig. 4(b)) calculated within the AM1-(PM3-) methods are rather similar. They can be characterized generally by five or six relatively strong absorption bands. In particular one obtains for 6-fluoro-1,3-dimethyl-[PQ]: 381(398.5), 360.5(359), 287(294), (256), 245(243), 224(224) nm; for 6-bromo-1,3-dimethyl-[PQ]: 374.5(394), 355(355), 284(282), (254), 246(243), 223(223) nm and for 1,3-dimethyl-7-trifluoromethyl-[PQ]: 379.5(402.5), 355.5(358), 242(256), 239(242), 221.5(223). The corresponding experimental spectra (see Figs. 2(a), 3(a) and 4(a)) have
Fig. 2. Measured absorption spectra of 6-fluoro-1,3-dimethyl[PQ] (a) and calculated spectra within the semi-empirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 6-fluoro-1,3-dimethyl-[PQ] obtained within the geometrical optimization procedure.
in all the cases an absorption threshold at about 399 nm which appears to be in fairly good agreement with quantum chemical simulations, namely for the spectra calculated by PM3-method. It is also amazing that spectral positions of the most strongest broad absorption band at about 240– 250 nm in the measured spectra well coincide with the calculated spectra. Both quantum-chemical methods give here sufficiently good agreement. A substitution of the methyl groups by the phenyl group (see Figs. 5 and 6) causes the substantial changes of the calculated spectra. In particular, quantum chemistry simulations reveal several additional relatively strong absorption bands within the spectral range 240–370 nm. This feature is common for all the [PQ]-derivatives containing phenyl groups (see e.g. [4,5]). An appropriate interpretation of their origin obviously refers to additional molecular doubled bonding p-conjugated segments
406
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
Fig. 3. Measured absorption spectra of 6-bromo-1,3-dimethyl[PQ] (a) and calculated spectra within the semi-empirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 6-bromo-1,3-dimethyl-[PQ] obtained within the geometrical optimization procedure.
C@C of the substituted phenyl groups. The corresponding absorption bands may be attributed to p ! p* electronic transition from the bonding to anti-bonding molecular configuration. Total spectra of this organic compound contain eight or nine relatively intense bands as calculated within AM1-(PM3-) procedure, in particular, one obtains for 6-fluoro-3-methyl-1-phenyl-[PQ]: 385.5(410.5), 362(368.5), 315(316), 284(288), 254(260), 245(246.5), 238.5(229) 229.5(217) and 210.5 nm and for 7-trifluoromethyl-3-methyl-1-phenyl-[PQ]: 387.5(417), 357.5(366), 310.5(314.5), 285.5(288.5), 252.5(259), 243.5(247.5), 236(227.5), 227(217.5), 211.5(207). The measured absorption threshold for both these phenyl containing [PQ]-derivatives kg = 399 nm differs slightly from the one calculated by AM1(PM3) methods which give 385.5(410.5) and 387.5(417) nm, respectively. In addition, the experimental spectra in UV region are characterized by only two broad bands. The most strongest
Fig. 4. Measured absorption spectra of 1,3-dimethyl-7-trifluoromethyl-[PQ] (a) and calculated spectra within the semiempirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 1,3-dimethyl-7-trifluoromethyl[PQ] obtained within the geometrical optimization procedure.
band exists at about 272 nm, i.e. it is clearly red shifted for both compounds with respect to the spectral positions of the series of most intense absorption bands obtained by the quantum chemical simulations. The reasons for the discrepancies observed between the measured and calculated spectra of [PQ]-derivatives have been discussed already in our recent publications [4,5]. Initially [4] they were attributed to the solvatochromic effect [17]. However, the experiments with strongly and weakly polar solvents [5] have indicated indeed rather small its influence on the absorption spectra. All the solutions clearly have shown only a weak hypsochromic (blue) shift of kg as the solvent polarity increased. The strongest one was observed on about 4 nm only. Broadened spectra as well as observed band position shifts are, therefore, rather a result of two other eventual mechanisms which are not taken into account in quantum-chemical calcula-
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
407
Fig. 5. Measured absorption spectra of 6-fluoro-3-methyl-1phenyl-[PQ] (a) and calculated spectra within the semi-empirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 6-fluoro-3-methyl-1-phenyl-[PQ] obtained within the geometrical optimization procedure.
Fig. 6. Measured absorption spectra of 7-trifluoromethyl-3methyl-1-phenyl-[PQ] (a) and calculated spectra within the semi-empirical quantum chemical PM3 and AM1 models (b). Inset shows the chemical structure of 7-trifluoromethyl-3methyl-1-phenyl-[PQ] obtained within the geometrical optimization procedure.
tions. One of those mechanisms is its broadening due to the electron-vibronic coupling. Indeed the lowest energy absorption band reveals features very similar to vibronic replica (see e.g. Fig. 2(a) and Fig. 3(a)). The measurements on a frozen solution or by cooling a disperse solid solution (e.g. polymer matrix) would made these features more obvious. The second reason is much related to the derivatives containing methyl, phenyl or CF3 groups. One should be emphasized that our quantum-chemical calculations have been performed for isolated molecules in vacuum. The molecular optimization procedure gives here a planar molecular configuration for all [PQ]-derivatives. That is unlikely to be quite correct for a solution phase assuming just that the methyl, phenyl or CF3 groups are presumably to rotate dynamically affecting the symmetry selection rules for different electronic transitions.
Fig. 7 shows the experimental efficiency of second-order optical susceptibility d vs photoinduced power density Ip for each chromophore. Table 2 presents the saturated second order optical susceptibility deff (as obtained from the data in the Fig. 7) as well as the dipole moments j ~ p j, effective (averaged) molecular optical nonlinear constant bav and excited state dipole moments qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðxÞ2
ðyÞ2
ðzÞ2
M g ¼ M g1 þ M g1 þ M g1 corresponding to the first active oscillator (as calculated within semi-empirical AM1 i PM3 methods). The experiment shows that maximally achieved second-order optical susceptibility (up to 1.53 pm/V) appears for the chromophore containing two methyl groups and fluor. In the phenyl-containing derivatives the second-order susceptibilities are substantially lower. It may be a consequence of competition between the charge transfer within two backside groups. However, the lowest magnitude of the
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
408
Fig. 7. Efficiency of second-order optical susceptibility deff vs photoinduced power density. Comp.1: 6-fluoro-1,3-dimethyl[PQ]; Comp.2: 7-trifluoromethyl-3-methyl-1-phenyl-[PQ]; Comp.3: 6-bromo-1,3-dimethyl-[PQ]; Comp.4: 6-fluoro-3-methyl-1-phenyl-[PQ]; Comp.5: 1,3-dimethyl-7-trifluoromethyl[PQ].
nonlinear optical susceptibility has been observed for 6-bromo-1,3-dimethyl-[PQ]. The presence of Br leads here to a suppressing of the charge transfer. One must also mention that studied composites lost the output second-order effective susceptibility not more than 24% which is comparable with other organic materials, e.g. such as cyan derivatives of stilbene [18] for which the second-order susceptibility reaches the magnitudes up to 1 pm/V. Several other important notes would be useful in the interpretation of nonlinear optical properties of polymer matrices doped by organic [PQ]-derivatives. Since we deal with the optical poling effect an important role plays the static polarization created by two coherent waves. Accordingly,
a substantial meaning in this case has the magnitude of the excited state dipole moments (not only in the occupied ground state). From the Table 2 one can conclude that quantum chemical PM3 method gives again better agreement with the experimental data compared to AM1 method. Several discrepancies between the values likely reflect specific features of the local Lorentz field related to macroscopic hyperpolarizabilities and macroscopic susceptibilities. To understand them better a simple model for optical grating formation would be rather useful. Generally, the optical polarizability can be presented by the linear and nonlinear terms L NL P~i ¼ P~i þ P~i ; L ðxÞ P~i ¼ aij Ej ;
P~i
NL
ðxÞ
ð4Þ ðxÞ
ðxÞ
ðxÞ
ðxÞ
¼ bijk Ej Ek þ cijkl Ej Ek El :
The microscopic susceptibilities aij, bijk, cijkl are related with the macroscopic susceptibility by the equation ðxÞ
ðxÞ ðxÞ
ðxÞ
ðxÞ ðxÞ ðxÞ
ðxÞ
vij ¼ Li Lj aij ; vijk ¼ Li Lj Lk bijk ; vijkl ðxÞ ðxÞ ðxÞ ðxÞ
¼ Li Lj Lk Lk cijkl ;
ð5Þ
LðxÞ z
where ¼ ½ðn2z 2Þ=ðn2z þ 3Þ=q are the compoðxÞ nents (z ” i,j,k) of the local Lorentz field and vijk is the macroscopic second-order nonlinear susceptibility responsible for the photoinduced SHG. The later one is very crucial in the case of guest–host polymers what may be a main reason of discrepancies between the local hyperpolarizabilities and macroscopic optical susceptibilities. Because the quantum-chemical calculations of Lorentz factor are
Table 2 Saturated second-order optical susceptibility deff (experiment) and calculated dipole moment j ~ p j, effective (averaged) molecular optical nonlinear constant bav and excited state dipole moments Mg Compound
Comp.2 Comp.3 Comp.4 Comp.5
deff
j~ pj
j~ pj
bav
bav
Mg
Mg
[pm/V] (exp.) 1.53 1.21 1.15 1.27 1.51
[D] (AM1) 0.596 3.555 0.842 1.084 3.184
[D] (PM3) 0.461 2.927 0.230 0.439 2.738
[au] (AM1) 722.8 1605 451.4 133.2 504.1
[au] (PM3) 682.4 134.1 131.1 182.2 392.3
[D] (AM1) 1.630 4.374 2.070 1.807 4.346
[D] (PM3) 3.796 5.694 3.516 4.215 5.543
Comp.1: 6-fluoro-1,3-dimethyl-[PQ]; Comp.2: 7-trifluoromethyl-3-methyl-1-phenyl-[PQ]; Comp.3: 6-bromo-1,3-dimethyl-[PQ]; Comp.4: 6-fluoro-3-methyl-1-phenyl-[PQ]; Comp.5: 1,3-dimethyl-7-trifluoromethyl-[PQ].
E. Kos´cien´ et al. / Optics Communications 242 (2004) 401–409
too complicate a new theoretical formalism should be developed which is a subject for a separate work. The results presented above appear in some contradiction to the generally adopted conception that the optical poling is associated only with the third order hyperpolarizibilities. The performed measurements and quantum chemical simulation indeed indicate on a substantial role of the second-order hyperpolarizibility as well.
409
ter agreement with the experiment compared to the AM1 method. The results of performed measurements and quantum chemical simulations contradict to the generally adopted conception that the optical poling is associated only with the third order hyperpolarizibilities. A substantial role of the second-order hyperpolarizibility is presented here as well. Acknowledgement
5. Conclusion The optical poling experiment shows, that maximally achieved second-order nonlinear optical susceptibility (up to 1.53 pm/V) appears for the chromophore containing two methyl groups and fluor as substituents. In the phenyl-containing derivatives the nonlinear second-order susceptibility is substantially lower. It may be a consequence of competition between the charge transfer within two backside groups. The lowest magnitude of the nonlinear optical susceptibility is revealed for Brcontaining [PQ]-derivative. It is assumed that Br leads here to a suppressing of the charge transfers mechanism what is the reason for a relatively low nonlinear optical efficiency. We present also the results of experimental studies and quantum chemical simulations of optical absorption and optical poling performed on a new synthesized halogen derivatives of 1H-pyrazolo[3,4-b]quinoline. The quantum chemical analysis reveals similarity in the absorption spectra of metyl-containing halogen derivatives which are characterized by four or five strong absorption bands in the spectral range 200–500 nm. A substitution of the methyl groups by the phenyl group causes the substantial changes of the absorption spectra mainly in the spectral range 240–370 nm. We attribute these differences to additional molecular double bonding segments C@C of the substituted phenyl groups, thereby the electronic p ! p* transitions appear to be involved into the absorption process. The discrepancies between the calculated and the measured spectra are attributed to electron-vibronic coupling as well as to specific rotational dynamics of phenyl rings. For all compounds the semi-empirical PM3 method gives bet-
This work was supported by KBN Grant No. 3-T11-B-074-26. References [1] Z. He, G.H.W. Milburn, K.J. Baldwin, D.A. Smith, A. Danel, P. Tomasik, J. Lumin. 86 (2000) 1. [2] Z. He, G.H.W. Milburn, A. Danel, A. Puchala, P. Tomasik, D. Rasala, J. Mater. Chem. 7 (1997) 2323. [3] A. Danel, Z. He, G.H.W. Milburn, P. Tomasik, J. Mater. Chem. 9 (1999) 339. [4] E. Koscien, J. Sanetra, E. Gondek, A. Danel, A. Wisla., A.V. Kityk, Opt. Commun. 227 (2003) 115. [5] E. Gondek, E. Koscien, J. Sanetra, A. Danel, A. Wisla, A.V. Kityk, Spectrochimica Acta, Part A, (to appear). [6] M.K. Balakirev, I.V. Kityk, V.A. Smirnov, L.I. Vostrikova, J. Ebothe, Phys. Rev. A 67 (2003) 023806. [7] I.V. Kityk, M. Makowska-Janusik, E. Gondek, L. Krzeminska, A. Danel, K.J. Plucinski, S. Benet, B. Sachraoui, J. Phys.: Cond. Matter 16 (2004) 231. [8] B.P. Antonyuk, Opt. Commun. 174 (2000) 427. [9] B.P. Apostoluk, D. Chapron, G. Godret, B. Sahraoui, J.M. Nunzi, C. Fiorini-Debuichat, P. Raimond, Opt. Lett. 27 (2002) 2028. [10] F. Charra, F. Kaizar, J.M. Nunzi, P. Raimond, E. Idiart, Opt. Lett. 18 (1993) 941. [11] J. Becher, B.F. Hansen, N. Malhotra, J. Heterocycl. Chem. 28 (1991) 1837. [12] L. Hennig, T. Mu¨ller, M. Grosche, J. Prakt. Chem. 332 (1990) 693. [13] A. Danel, K. Chaczatrian, P. Tomasik, ARKIVOC 1 (2000) 51. [14] A. Danel, Synthesis 1H-pyrazolo[3,4-b]quinolines as a luminofores for LEDÕs, Ph.D Thesis, 1996. [15] I.V. Kityk, B. Marciniak, A. Mefleh, J. Phys. D: Appl. Phys. 34 (2001) 1. [16] R. Boyd, Principles of Nonlinear Optics, Academic Press, New York, 1983. [17] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, VCH Verlag, Germany, 1988. [18] I. Yan, L. Liu, L. Ji, M. Ye, L. Lu, W. Wang, J. Phys. D: Appl. Phys. 37 (2004) 1597.