Optical absorption measurements and quantum-chemical simulations of optical properties of novel fluoro derivatives of pyrazoloquinoline

Article abstract of DOI:10.1016/j.chemphys.2010.01.001

The results of experimental research and quantum-chemical simulations of the absorption spectra of 1-(4-fluorophenyl)-3,4-diphenyl, 3-(4-fluorophenyl)-1,4-diphenyl, and 4-(4-fluorophenyl)-1,3-diphenyl-pyrazolo[3,4-b] quinoline are presented. Although the

Full text of DOI:10.1016/j.chemphys.2010.01.001

Chemical Physics 370 (2010) 194–200
Contents lists available at ScienceDirect
Chemical Physics
journal homepage: www.elsevier.com/locate/chemphys
Optical absorption measurements and quantum-chemical simulations of optical
properties of novel fluoro derivatives of pyrazoloquinoline
b
c
d
e
e
e
f
M.G. Brik a, , W. Kuznik , E. Gondek , I.V. Kityk , T. Uchacz , P. Szlachcic , B. Jarosz , K.J. Plucinski
*
a Institute of Physics, University of Tartu, Riia 142, Tartu 51014, Estonia
b Chemical Department, Silesian Technological University, Ul. M. Strzody 9, Gliwice, Poland
c
_
Institute of Physics, Cracow University of Technology, Ul. Podchorazych 1, 30-084, Poland
d Electrical Engineering Department, Czestochowa University of Technology, Al. Armii Krajowej 17/19, Cze˛stochowa, Poland
e Department of Chemistry, Hugon Kollataj Agricultural University, Al. Mickiewicza 24/28, 30-059 Cracow, Poland
f Military University of Technology, Electronics Department, 2 Kaliski Str., Warsaw, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
The results of experimental research and quantum-chemical simulations of the absorption spectra of 1-
(4-fluorophenyl)-3,4-diphenyl, 3-(4-fluorophenyl)-1,4-diphenyl, and 4-(4-fluorophenyl)-1,3-diphenyl-
pyrazolo[3,4-b] quinoline are presented. Although the fluorine atom is located on different phenyl rings
in these molecules, the absorption spectra do not differ significantly. Semi-empirical AM1, PM3 and RM1
methods, as well as ab initio ADF code-based calculations were used to optimize geometry, calculate the
infrared and visible spectra of the afore mentioned compounds and analyze the molecular orbitals
schemes. The results of calculations are in good agreement with the experimental data. It was also dem-
onstrated that the positions of the fluorescence maxima depend significantly on the solvent (contrary to
the absorption spectra), in which the molecules are embedded, which allows for manipulating with fluo-
rescence properties of the synthesized molecules by changing the solvent.
Received 10 August 2009
In final form 6 January 2010
Available online 11 January 2010
Keywords:
Electronic properties
Visible spectra
Ultraviolet spectra
DFT calculations
Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction
phenyl-1H-pyrazolo[3,4-b]quinoline,
3-(4-fluorophenyl)-1,4-di-
phenyl-1H-pyrazolo[3,4-b]quinoline and 4-(4-fluorophenyl)-1,3-
diphenyl-1H-pyrazolo[3,4-b]quinoline, for the sake of brevity
hereafter referred to as 3a (PQF1), 3b (PQF3) and 3c (PQF4),
respectively.
The AM1, PM3 and RM1 semi-empirical computational meth-
ods and density functional theory (DFT) calculations (as imple-
mented in the Amsterdam Density Functional code (ADF) [10])
were used to simulate the UV and IR spectra; the results of these
calculations are compared with experimental absorption spectra
of these molecules in acetonitrile and dichloromethane possessing
different polarizabilities. These methods are well known to be reli-
able for organic molecules properties’ simulations.
Recently considerable attention was paid to organic light emit-
ting materials that could be used in various optoelectronic devices,
such as the light emitting diodes (LED), sensors and electrolumi-
nescent displays (ELD). Pyrazoloquinoline and its derivatives rep-
resent a class of chromophore, which is highly fluorescent in a
wide spectral range, from blue to green–yellow, depending upon
the substituents [1–6]. These compounds are also promising mate-
rials for optoelectronic applications [7,8]. Thus, the problem of
understanding of the nature of electronic transitions in these sys-
tems, theoretical modeling of their optical properties and search
for new materials with optimized characteristics is crucially
important, in particularly for the understanding of chemical phys-
ics of observed phenomena. Recently several papers dealing with
solvatochromic effect and the influence of phenyl ring rotation
on the UV–Vis absorption spectra of pyrazoloquinoline and some
of their derivatives have been published [6,9].
2. Synthesis and experimental
1H-pyrazolo[3,4-b]quinolines were prepared either by Fried-
lander condensation (3a and b) or three-component reaction
(3c). In the first case o-aminobenzophenone 1 was heated with
fluorine substituted 2,5-diphenyl-2,4-dihydropyrazol-3-ones 2a
or 2b. Three-component reaction involved heating of aniline 4, p-
fluorobenzaldehyde and 2,5-diphenyl-2,4-dihydropyrazol-3-one.
The last method is one of the best procedures described for the
1H-pyrazolo[3,4-b]quinoline synthesis (Fig. 1).
In this paper, we present the results of experimental investi-
gations and theoretical simulations of absorption spectra of
novel pyrazoloquinoline derivatives: 1-(4-fluorophenyl)-3,4-di-
* Corresponding author.
E-mail address: brik@fi.tartu.ee (M.G. Brik).
0301-0104/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.chemphys.2010.01.001

M.G. Brik et al. / Chemical Physics 370 (2010) 194–200
195
Fig. 1. Schemes of preparation of the studied complexes. Top: ethylene glycol, 180–190 °C, 19 h. Bottom: ethylene glycol, 180–190 °C, 2 h. See the text for all notations.
2.1. 1-(4-Fluorophenyl)-3,4-diphenyl-1H-pirazolo[3,4-b]quinoline
(3a)
1H NMR (300 MHz, CDCl3) d[ppm]: 7.03–7.35 (m, 12H), 7.38
(ddd, J = 8.6, 6.7, 1.2 Hz, 1H), 7.76 (ddd, J = 8.5, 6.7, 1.5 Hz, 1H),
7.90 (dd, J = 8.7, 0.8 Hz, 1H), 8.21 (dd, J = 8.7, 0.5 Hz, 1H), 8.53–
8.60 (m, 2H); 13C NMR (75 MHz, CDCl3) d[ppm]: 115.5, 116.4 (d,
2-Aminobenzophenone 1 (395 mg, 2 mmol) and 1-(4-fluoro-
phenyl)-3-phenyl-2-pyrazoline-5-one 2a (508 mg, 2 mmol) were
stirred with 1.5 g of ethylene glycol for 19 h in 190 °C. Then the
reaction mixture was cooled and poured into water. Product was
extracted with chloroform (common procedure). The extract was
flashed through Al2O3 and the solvent was evaporated. Crude prod-
uct was dissolved in toluene and flashed through Al2O3. The sol-
vent was evaporated and the product was crystallized few times
from toluene: petroleum ether to give 317 mg (38%) of desired
product as yellow crystalline powder, m.p. 206–207 °C.
3
2JC–F = 22.7 Hz), 123.2 (d, JC–F = 8.0 Hz), 124.2, 124.9, 127.8,
128.2, 128.3, 128.6, 129.1, 129.7, 131.0, 131.2, 133.1, 135.1, 136.8
(d, 4JC–F = 2.6 Hz), 145.5, 147.5, 149.3, 150.8, 161.5 (d,
1JC–F = 244.4 Hz).
2.2. 1,4-Diphenyl-3-(4-fluorophenyl)-1H-pirazolo[3,4-b]quinoline
(3b)
2-Aminobenzophenone 1 (1.0 g, 5 mmol) and 1-phenyl-3-(4-
fluorophenyl)-2-pyrazoline-5-one 2b (1.3 g, 5 mmol) were stirred
with 3 g of ethylene glycol for 24 h in 190 °C. Then the reaction
mixture was cooled slightly, ethanol was added to hot solution
and the whole mixture was boiled with ethanol. Then the mixture
was cooled, and crude product was filtered off. Crude compound
was crystallized few times from toluene to give 831 mg (40%) of
desired product as yellow needles, m.p. 216–217 °C.
1H NMR (300 MHz, CDCl3) d[ppm]: 7.76 (tt, J = 8.8, 2.1 Hz, 2H),
7.07–7.14 (m, 2H), 7.20–7.42 (m, 7H), 7.58 (tt, J = 9.0, 1.9 Hz, 2H),
7.77 (ddd, J = 8.5, 6.8, 1.4 Hz, 1H), 7.89 (dd, J = 8.7, 0.8 Hz, 1H),
8.24 (dd, J = 8.7, 0.5 Hz, 1H), 8.59 (dt, J = 7.6, 1.0 Hz, 2H); 13C
2
NMR (75 MHz, CDCl3) d[ppm]: 114.5 (d, JC–F = 21.7 Hz), 114.8,
120.9, 123.5, 124.2, 125.5, 127.0, 128.0, 128.5, 128.7 (d,
3
4JC–F = 3.1 Hz), 129.0, 129.1, 130.3, 130.5, 130.8 (d, JC–F = 8.4 Hz),
1
134.5, 139.8, 144.5, 145.7, 148.6, 150.2, 162.4 (d, JC–F = 247.6 Hz).
2.3. 1,3-Diphenyl-4-(4-fluorophenyl)-1H-pirazolo[3,4-b]quinoline
(3c)
Aniline 4 (0.93 g, 10 mmol), 4-fluorobenzaldehyde 5 (1.24 g,
10 mmol), 1,3-diphenyl-2-pyrazoline-5-one 2c (2.36 g, 10 mmol)
and 4.5 g of ethylene glycol were placed together in round bottom
Fig. 2. Experimental UV–Vis absorption spectra of PQF1, PQF3 and PQF4 in
acetonitrile solvent (mass concentration approximately 0.01%).

196
M.G. Brik et al. / Chemical Physics 370 (2010) 194–200
Fig. 3. Optimized structures of: 1-(4-fluorophenyl)-3,4-diphenylpyrazoloquinoline (PQF1) (top left); 3-(4-fluorophenyl)-1,4-diphenylpyrazoloquinoline (PQF3) (top right); 4-
(4-fluorophenyl)-1,3-diphenylpyrazoloquinoline (PQF4) (bottom). Positions of the fluorine and nitrogen atoms are indicated.
Fig. 4. PQF1 experimental (top panel) and calculated UV–Vis absorption spectra of
Fig. 5. PQF3 experimental (top panel) and calculated spectra of MM + optimized
structure (bottom panel). ACN – acetonitrile solution, CH2Cl2 – dichloromethane
solution, approximately 0.01%.
MM + optimized molecular structure (bottom panel) in different solvents: ACN –
acetonitrile solution, CH2Cl2 – dichloromethane solution, approximately 0.01%.
flask and refluxed for 2 h. Then the mixture was cooled, ethanol
was added, and yellow precipitate was filtered of and washed with
ethanol. Crude product was dissolved in chloroform and flashed
through Al2O3 and the solvent was evaporated. Yellow solid was

M.G. Brik et al. / Chemical Physics 370 (2010) 194–200
197
Table 1
Spectral positions of absorption maxima (in nm) in the experimental spectra of PQF1,
PGF3, and PQF4 compounds. The calculated values are in the RM1, PM3, AM1 rows,
and the experimental values – in the CH2Cl2 and CAN rows.
PQF1
PQF3
PQF4
RM1
PM3
AM1
210
204
207
238
237
233
263
260
257
320
310
306
345
350
335
415
438
405
CH2Cl2 0.01 g/l
ACN 0.015 g/l
260
260
316
314
328
328
398
396
212
RM1
PM3
AM1
205
202
198
235
234
231
259
262
265
311
308
300
358
360
350
417
440
408
CH2Cl2
ACN
260
260
316
314
328
328
398
396
216
RM1
PM3
AM1
204
201
198
236
235
232
265
260
259
312
303
297
419
442
410
CH2Cl2
ACN
262
260
316
316
330
328
398
396
220
Fig. 7. Sensitivity of the constants k to the PQF1 calculated spectra.
of the fluorescence maxima are very similar; there is only a slight
difference in the emission intensities determined by their oscillator
strengths.
3. Calculation procedure
The calculations of UV–Vis absorption spectra were performed
with HyperChem 8.04 package. The molecular structure was opti-
mized with the MM + force field model, which is known as a
widely used reliable method for prediction of molecular structure,
especially for organic molecules and polymers [11,12]. After that,
the calculations of the IR spectra were performed using the ADF
package. Fig. 3 shows optimized configurations for the considered
molecules. For the sake of brevity, we do not give here the calcu-
lated atomic positions (50 atoms for each of this three molecules
would mean a rather lengthy table); they can be available from
the authors upon request. The structures are different mainly with
respect to the mutual positions of the F and N atoms and their
attachments to the phenyl rings. The shortest F–C distance is
1.401 Å for the PQF1 molecule, 1.400 Å for the PQF3 molecule,
and 1.398 Å for the PQF4. Distances from the fluorine atom to
the three nitrogen atoms are (all in Å): 5.54, 6.332, 6.575 for
PQF1; 6.371, 7.662, 9.022 for PQF3, and 7.801, 8.477, 8.495, for
PQF4.
Fig. 6. PQF4 experimental (top panel) and calculated UV–Vis spectra of MM + opti-
mized structure (bottom panel). ACN – acetonitrile solution, CH2Cl2 – dichloro-
methane solution, approximately 0.01%.
Parametrized semi-empirical methods AM1, RM1 and PM3
were used to calculate the optical spectra taking into account only
the singly excited configuration interactions. The orbital criterion
of 12 occupied and 12 un-occupied orbitals was chosen for all
the three methods. The relatively new Recife Model 1 (RM1) is a
re-parametrized Austin Model 1 (AM1) version. Unlike AM1, RM1
uses optimized parameters and has been proven to give better re-
sults than both AM1 and PM3 in many cases [13].
crystallized from toluene to give 673 mg (16%) of desired product
as yellow needles, m.p. 240–242 °C.
1H NMR (300 MHz, CDCl3) d[ppm]: 6.93 (tt, J = 8.7, 2.1 Hz, 2H),
7.12–7.26 (m, 7H), 7.32 (tt, J = 7.4, 1.1 Hz, 1H), 7.40 (ddd, J = 8.6,
6.7, 1.2 Hz, 1H), 7.58 (tt, J = 8.0, 1.9 Hz, 2H), 7.77 (ddd, J = 8.5, 6.8,
1.5 Hz, 1H), 7.87 (dd, J = 8.7, 0.8 Hz, 1H), 8.25 (d, J = 8.3 Hz, 1H),
8.58–8.62 (m, 2H); 13C NMR (75 MHz, CDCl3) d[ppm]: 115.0 (d,
2JC–F = 21.8 Hz), 115.2, 120,9, 123.5, 124.3, 125.5, 126.6, 127.6,
4. Results and discussion
3
127.8, 129.0, 129.1, 129.2, 130.4, 130.5, 132.1 (d, JC–F = 8.4 Hz),
1
132.5, 139.9, 143.3, 148.6, 150.2, 162.9 (d, JC–F = 248.6 Hz).
4.1. Optical spectra simulation
The UV–Vis absorption and fluorescence spectra were recorded
in acetonitrile and dichloromethane (HPLC grade) solutions (con-
centrations approximately 0.001% w/w) in a standard 1 cm path
length quartz cell using HP 8452A UV–Vis Diode Array Spectropho-
tometer in the 190–820 nm range with spectral resolution 1 nm.
Experimental fluorescence spectra of the studied samples are
shown in Fig. 2. The shape of the spectra and the spectral positions
Figs. 4–6 show the comparison between the experimental and
calculated UV–Vis absorption spectra of pyrazoloquinoline fluoro
derivatives. There is practically no significant difference between
the UV–Vis absorption spectra of PQF1, PQF3 and PQF4, in spite
of the fact that fluorine atom is attached to different phenyl rings
in their molecular structures (Fig. 3).

198
M.G. Brik et al. / Chemical Physics 370 (2010) 194–200
In each spectrum there is a wide spectral peak in the visible
its location accurately has to be considered as a serious drawback
of the PM3 method. AM1 prediction of this band was slightly better
than that of RM1, but both were closer to the experimental data
(see Figs. 4–6, Table 1).
spectral range with maximum at approximately 400 nm and sev-
eral peaks of different intensities in the ultraviolet range, namely
two small peaks at 330 and 315 and two higher peaks at 260 and
220 nm. No distinct solvatochromic effect can be observed as the
experimental absorption spectra of pyrazoloquinoline derivatives
in acetonitrile and dichloromethane are practically identical, which
is in accordance with earlier observations [14]. The only difference
is that the highest energy peak in dichloromethane spectra
(210 nm) is not detected, but this is probably a result of relatively
low solvent transparency below 230 nm and nothing should be de-
duced from any results in the far-UV region of these spectra.
Generally, all the applied semi-empirical methods produced
quite satisfactory results: the number of peaks simulated and their
spectral positions prevailingly coincide with experimental spectra.
However, some significant differences between the AM1, RM1 and
PM3-derived spectra were found. The first one is the red shift of
the visible range absorption band by approximately 40 nm in all
the PM3-calculated spectra (Figs. 4–6, Table 1).
Fig. 7 shows how choice of fitting constant influences the
resultant spectra of RM1 prediction of PQF1 absorption. The
graphs are built for low values of the fitting constant k: 0.1
and 0.18. This constant is related to the electron-vibration broad-
ening and for these compounds this constant is comparable with
other similar materials. One can see that even such a small dif-
ference in the constant’s value changes the spectrum signifi-
cantly. When the 0.18 value is used, the spectral peaks are
wider, which resembles more the experimental spectra. On the
other hand, as the spectral peaks are wide, they may merge to
form a wide spectral band instead of two peaks with distinct
maxima (Fig. 6, 300–360 nm).
The calculated DOS diagrams (shown below in Fig. 8) are all
very reach in the number of the molecular orbitals and include
292 orbitals in total. Analysis of the calculated excitation spectra
allows to assign the excitation peaks and narrow down the number
of the orbitals involved into the optical transitions: these are the
Usually this band is of crucial importance for potential opto-
electronic applications of these materials. The failure to predict
Fig. 8. Calculated DOS. From top to bottom: PQF1, PQF3, PQF4. The HOMO level in each case is shown by a vertical arrow. Region of the optical transitions is shown by a brace.

M.G. Brik et al. / Chemical Physics 370 (2010) 194–200
199
Fig. 9. Molecular orbitals scheme in the PQF1 molecule. The occupied orbitals are shown in black color, and un-occupied – in grey. For the sake of brevity similar diagrams for
PQF3 and PQF4 are not shown. Vertical scale is the energy in eV.
nearest neighbors (to the lower and higher energies) of the highest
occupied molecular orbital (HOMO).
4.2. IR spectra simulation
Fig. 9 shows the overall scheme of the molecular orbitals in
PQF1 molecule. It is seen from this diagram, that the lowest un-
occupied orbitals consists mainly of the C and H contributions. In
the highest un-occupied orbitals the H contribution is replaced
by that one of N.
Fluorescence spectra were measured in cyclohexane (CHX, non-
polar solvent) and acetonitrile (ACN, polar solvent) media. PQF1
spectrum (Fig. 10) shows strong solvatochromic shift from 455 to
480 nm, PQF3 – 460–485 nm, and PQF4 – 455–480 nm.
Since all synthesized molecules consist of 50 atoms, 144 normal
vibration modes should form the observed experimentally vibra-
tional spectrum. The calculations of the vibrational spectra were
performed after optimizing the geometry of the molecules using
the ADF program with LDA potential and DZ basis set. The results
of the calculations are shown in Fig. 11 (the tables with the calcu-
lated vibrational frequencies and coordinates of the atoms in the
optimized structures are not given here for the sake of brevity,
but are available from the authors upon request). The overall
appearance of the spectra is similar: a low intensity vibration at
about 3100 cmÀ1, and two groups of structured intensive lines at
about 750 and 1500 cmÀ1 (the density of the vibrational states in
these regions is very high). However, it should be emphasized that
redistribution of the IR bands intensities takes place: for example,
intensity of the 1500 cmÀ1 line is decreased in PQF3 and PQF4,
compared with that for PQF1. This can be explained by different
positions of the fluorine atom in each of these molecules.
5. Conclusions
In the present paper, we report on the synthesis and spectral
studies of three organic molecules: 1-(4-fluorophenyl)-3,4-diphe-
nyl, 3-(4-fluorophenyl)-1,4-diphenyl, and 4-(4-fluorophenyl)-1,3-
diphenyl-pyrazolo[3,4-b]quinoline. The experimental studies of
the excitation and emission spectra were supported by theoretical
quantum-chemical calculations (both semi-empirical and ab initio)
of the optimized geometry, excitation spectra, DOS and IR spectra
of the compounds. The calculated results include the positions of
atoms in the optimized structures, normal vibration frequencies
and molecular orbital energies. The fluorescence spectra measure-
ments have shown a profound solvatochromic shift of the lumines-
Fig. 10. Fluorescence spectra of studied compounds. See text for further details.

200
M.G. Brik et al. / Chemical Physics 370 (2010) 194–200
cence maxima, which gives an opportunity of tuning the fluores-
cence properties of the synthesized molecules in different solvents.
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] S. Calus, E. Gondek, A. Danel, B. Jarosz, M. Pokladko, A.V. Kityk, Mater. Lett. 61
(2007) 3292.
[4] A. Danel, Z. He, G.H.W. Milburn, P. Tomasik, J. Mater. Chem. 9 (1999) 339.
[5] Y.T. Tao, E. Balasubramanian, A. Danel, B. Jarosz, P. Tomasik, Appl. Phys. Lett. 77
(2000) 1575.
[6] E. Koscien, E. Gondek, A.V. Kityk, Opt. Commun. 280 (2007) 95.
[7] L. Mu, Z. He, J. Wang, G. Hui, Y. Wang, X. Jing, A. Danel, E. Kulig, IEEE Photon.
Technol. Lett. 20 (2008) 1781.
[8] E. Gondek, A. Danel, I.V. Kityk, J. Mater. Sci.: Mater. Electron. 20 (2009) 461.
[9] S. Całus, E. Gondek, M. Pokladko, E. Kulig, B. Jarosz, A.V. Kityk, Spectrochim.
Acta A 67 (2007) 1007.
[10] G. te Velde, F.M. Bickelhaupt, E.J. Baerends, C. Fonseca Guerra, S.J.A. van
Gisbergen, J.G. Snijders, T. Ziegler, J. Comput. Chem. 22 (2001) 931.
´
[11] M. Czerwinski, J. Bieleninik, J. Napieralski, I.V. Kityk, J. Kasperczyk, R.I.
Mervinskii, Eur. Polym. J. 33 (1997) 1441.
[12] S.W. Tkaczyk, I.V. Kityk, R. Schiffer, J. Phys. D: Appl. Phys. 35 (2002) 563.
[13] G.B. Rocha, R.O. Freire, A.M. Simas, J.J. P Stewart, J. Comput. Chem. 27 (2006)
10.
[14] E. Gondek, E. Koscien, J. Sanetra, A. Danel, A. Wisła, A.V. Kityk, Spectrochim.
Acta A 60 (2004) 3101.
Fig. 11. Calculated vibrational spectra of PQF1 (top), PQF3 (middle) and PQF4
(bottom) vibrations.