Available online at www.sciencedirect.com
Optics Communications 281 (2008) 3806–3819 www.elsevier.com/locate/optcom
Spectroscopic and lasing characterisation of a dicarbazovinylene-MEH-benzene dye A.K. Bansal a, W. Holzer a, A. Penzkofer a,*, E.B. Kley b a
Institut II – Experimentelle und Angewandte Physik, Universita¨t Regensburg, Universita¨tstrasse 31, D-93053 Regensburg, Germany b Institut fu¨r Angewandte Physik, Friedrich-Schiller Universita¨t Jena, Max-Wien Platz 1, D-07743 Jena, Germany Received 18 January 2008; received in revised form 18 March 2008; accepted 20 March 2008
Abstract The dye 1,4-bis(9-ethyl-3-carbazovinylene)-2-methoxy-5-(20 -ethyl-hexyloxy)-benzene (abbreviated 2CzV-MEH-B) dissolved in tetrahydrofuran (THF) and as neat film is characterised by optical absorption and emission spectroscopy. The absorption and stimulated emission cross-section spectra, the fluorescence quantum distributions, fluorescence quantum yields, degrees of fluorescence polarisation, and fluorescence lifetimes are determined. A lasing characterisation is carried out by pumping with single second harmonic pulses of a mode-locked ruby laser (wavelength 347.15 nm, pulse duration 35 ps). The excited-state absorption at the pump laser wavelength is determined by saturable absorption measurements. Laser oscillation of the dye in THF in a rectangular cell is achieved by transverse pumping using the uncoated cell windows for light feedback. From the emission behaviour around threshold the excited-state absorption cross-section spectrum in the laser active spectral region is extracted. The wave-guided travelling-wave lasing behaviour of the dye as neat film is studied by analysis of the amplification of the transverse pumped spontaneous emission. Surface emitting distributed-feedback lasing was achieved with a neat film on corrugated second-order periodic gratings. Ó 2008 Elsevier B.V. All rights reserved. PACS: 42.70.Hj; 42.55.Mv; 33.20.Kf Keywords: Dicarbazovinylene-MEH-benzene dye; ADS084BE light emitting oligomer; Neat film; Absorption spectroscopy; Fluorescence spectroscopy; Amplified spontaneous emission; Low-Q laser oscillation; Travelling-wave lasing; Distributed-feedback lasing; Saturable absorption; Excited-state absorption; Fluorescence self-quenching
1. Introduction The synthesis of laser active organic materials, their physical characterisation, and application in wave-guided thin-film lasers is an active field of research (for reviews see [1,2]). Laser action was achieved for various poly-phenylene-vinylenes (PPV, for review see [2]), poly-phenylene-ethenylenes (PPE) [3], ladder-type poly-paraphenylenes (PPP) [2], polyfluorenes (PF) [4], poly-phenylacethylenes (PPA) [5], poly-arylene-vinylenes (PAV) [6], poly-thiophenes (PT) [7], triphenylamine (TPA) based con*
Corresponding author. Tel.: +49 941 943 2107; fax: +49 941 943 2754. E-mail address:
[email protected] (A. Penzkofer). 0030-4018/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.optcom.2008.03.032
jugated polymers [8], and triphenylamine dimer (TPD) based conjugated and non-conjugated polymers [9]. Thin films made of low-molar-mass organic laser dyes generally loose their lasing ability because of concentration dependent fluorescence quenching [10]. Several spiro-dyes have high solid-state fluorescence efficiency and show optically pumped neat thin-film amplified spontaneous emission [11,12]. For the spiro-linked material 2,20 ,7,70 -tetrakis(4-fluorophenyl)spiro-9,90 -bifluorene optically pumped distributed-feedback laser action in the near ultraviolet wavelength range between 377.7 nm and 394.8 nm was achieved [12]. The widely used electroluminescent TPD molecules turned out to work as efficient blue-wavelength neat thin-film lasers [13,14]. Neat film laser action was also achieved with a thianthrene-substituted distyrylbenzene
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
dye (thianthrene-DSB) [15]. Single-mode tunable laser emission was achieved with an electroluminescent oligothiophene (quinquethiophene-S,S-dioxide T5oA) [16]. Electroluminescent organic molecular dyes applied in light emitting diodes have high solid-state fluorescence efficiency [17–20]. Whether these electroluminescent dyes in neat solid films are laser active has to be tested in each particular case. Here the light emitting oligomer ADS084BE bought from American Dye Source, Inc., Quebec, is characterised spectroscopically and investigated for neat thinfilm and organic-solution laser action. The structural formula of the molecule is shown in Fig. 1. Its chemical name is 1,4-bis(9-ethyl-3-carbazovinylene)-2-methoxy-5(20 -ethyl-hexyloxy)-benzene. It is abbreviated by 2CzVMEH-B to remind of the two carbazovinylene cap-groups and the central MEH-benzene (or phenylene) part. The 2CzV-MEH-B oligomer was selected for investigation because its central part is the repeat unit of MEH-PPV (poly(2-methoxy-5-(20 -ethyl-hexyloxy)-1,4-phenylene-vinylene)) which is one of the most widely used polymers for laser action [5,21], and the two 9-ethyl-9H-carbazole molecules contain the 9H-carbazole core which is the basic repeat unit in different kinds of poly-carbazoles which are used as host materials for dye doped travelling-wave lasers [22] and distributed-feedback lasers [23,24] as well as holetransport layer of organic light emitting diodes (OLEDs) [25–27]. 4,40 -N,N0 -Dicarbazole-biphenyl (CBP) is used as host material in phosphorescent OLEDs like PtOEP [28,29] and Ir(ppy)3 [28,30]. A carbazole-paraphenylenevinylene (CC-PVV) copolymer was synthesized and used as active material of a light emitting diode [31]. The 2CzV-MEH-B oligomer is studied in tetraydrofuran (THF) and as a thin film on a glass substrate. Absorption cross-section spectra, stimulated emission cross-section spectra, fluorescence quantum distributions, fluorescence quantum yields, degrees of fluorescence polarisation, and fluorescence lifetimes are determined. The saturable absorption of the dye at 347.15 nm (second harmonic of picosecond ruby laser) is measured and analysed. The amplification of spontaneous emission (wave-guided travelling-wave lasing) in neat films and the dye solution low-Q lasing in a cell (cell windows act as low reflectivity resonator mirrors) are studied by transverse sample pumping. The
O N C2H5 N
C2H5
H3C O
Fig. 1. Structural formula of studied dye 2CzV-MEH-B (product name: ADS084BE). Sum formula C47H50N2O2, molar mass Mm = 674.91 g mol1.
3807
distributed-feedback lasing of the dye spin-coated on corrugated gratings etched into a quartz glass is investigated. The low-Q lasing onset is analysed to extract the excitedstate absorption cross-section spectrum of the dye in the fluorescence spectral region. 2. Experimental The dye 1,4-bis(9-ethyl-3carbazovinylene)-2-methoxy-5(20 -ethyl-hexyloxy)-benzene (abbreviated 2CzV-MEH-B) was bought from American Dye Source Inc., Quebec, Canada (product name: ADS084BE) where it is classified as light emitting oligomer. The dye was used without further purification. Its structural formula is shown in Fig. 1. The dye was studied dissolved in tetrahydrofuran (THF) and as neat film. Thin films on substrates for travellingwave lasing and distributed feedback lasing were prepared by spin-coating from THF solution. The substrates were fused silica for optical constants measurements, optical glass microscope carrier plates for travelling-wave lasing experiments, and corrugated fused silica gratings for distributed-feedback laser experiments. The applied dye concentration for spin-coating was 10 mg per ml for the travelling-wave lasing films (spinning speed 1600 rpm), and 80 mg per ml for the distributed-feedback laser films (spinning speed 2400 rpm). The optical constants (refractive index spectrum, n(k), and absorption coefficient spectrum, a(k)) of a 2CzVMEH-B neat film and its film thickness, df, were determined by reflectance, R(k), and transmittance, T(k), measurements as described previously [32,33]. The absorption cross-section spectrum, ra,s(k), of 2CzVMEH-B in THF was measured with a commercial spectrophotometer (Cary 50 from Varian). The absorption cross-section spectrum, ra,f(k), of neat film 2CzV-MEH-B was determined by using the spectral shape of the absorption coefficient spectrum and assuming equal absorption cross-section integrals for the film and the liquid solution in the displayed wavelength range [6]. The relation R R ra;f ðkÞ ¼ af ðkÞ ra;s ð~mÞd~m= af ð~mÞd~m is used where ~m ¼ k1 is the wavenumber. The dye number density in the thin-film is given by Nf = af/ra,f. The fluorescence analysis of 2CzV-MEH-B in THF and of 2CzV-MEH-B thin films was carried out with a selfassembled fluorimeter [34]. The data analysis is described in [35]. The dye coumarin 314T in ethanol was used as reference standard (fluorescence quantum yield /F,R = 0.87 according to Kodak data sheet). The degree of fluorescence polarisation, P F ðkÞ ¼ ½S F;k ðkÞ S F;? ðkÞ=½S F;k ðkÞþ S F;? ðkÞ, was determined by vertical polarised excitation and detection of the fluorescence signal polarised parallel ðS F;k Þ and perpendicular (SF,\) to the excitation light. Temporal fluorescence traces of 2CzV-MEH-B in THF were measured by excitation the solution with second harmonic pulses of an active and passive mode-locked ruby laser [36] (pulse duration 35 ps, wavelength 347.15 nm) and detection of the fluorescence signal with a fast micro-
3808
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
channel-plate photomultiplier (Hamamatsu type R1564U01) connected to a fast real-time digital oscilloscope (LeCroy type 9362). Temporal fluorescence traces of a 2CzV-MEH-B thinfilm were measured by excitation with second harmonic pulses of a titanium sapphire femtosecond laser (pulse duration 130 fs, wavelength 400 nm, laser system Hurricane from Spectra-Physics) and fluorescence signal detection with a picosecond streak-camera (Hamamatsu type C1587 with M1952 plug-in). The saturable absorption behaviour [37] of 2CzV-MEHB in THF at 347.15 nm was studied by excitation intensity dependent energy transmission measurement of 35 ps laser pulses at 347.15 nm through a cell of 1 mm thickness. The experimental arrangement is shown in Fig. 2a. The excitation intensity is determined by two-photon transmission measurement through a KI crystal [38] (photo-detectors PD1 and PD2 in Fig. 2a). The energy transmission through the dye cell is measured with photo-detectors PD1 and PD3. The excited-state absorption cross-section of the sample at the excitation wavelength is extracted from the dependence of the energy transmission on the input laser peak intensity (see below). Low-Q lasing studies on 2CzV-MEH-B in THF were carried out by transverse pumping the dye solution in a quartz glass cell of 1 cm width and 1 mm thickness. The
dye concentration was C = 4.5 104 mol dm3 (transmission T0 = 0.08 at excitation wavelength kP = 347.15 nm). The dye was excited with pulses of a frequency-doubled mode-locked ruby laser (duration DtP = 35 ps, wavelength kP = 347.15 nm, single pulse energy up to 1 mJ). The beam profile of the excitation pulses was shaped to a line-focus of 14.5 mm width (FWHM) and 0.245 mm height (FWHM) using crossed defocusing and focusing cylindrical lenses. The excitation occurred along the 1 cm cell side, perpendicular to the 1 mm cell windows. The cell windows acted as resonator mirrors. The emitted signal from the cell windows was collimated and directed to a spectrometer – diode-array detection system (observation of spectral amplification and spectral narrowing). A scheme of the experimental arrangement is shown in Fig. 2b. Wave-guided travelling-wave lasing studies on 2CzVMEH-B thin films were carried out with the same experimental setup. Only the cell was replaced by a thin-film which was spin-coated on an optical glass plate (microscope carrier plate cut after coating in the film region for optimum edge emission). The plate surface was tilted 7° off the perpendicular direction towards the excitation beam direction in order to better collect the light propagating along the film substrate interface [39]. A scheme of the experimental arrangement is shown in Fig. 2c.
Fig. 2. Experimental setups for (a) saturable absorption measurement, (b) low-Q laser action using dye cell, (c) wave-guided travelling-wave laser action using neat film on glass substrate, and (d) distributed feedback laser action using neat film on corrugated grating. M.L.Laser: mode-locked ruby laser system, SHG: KD*P crystal for second harmonic generation, F: second harmonic pass filter, BS: beam splitters, L1–L6: lenses, TPA: KI crystal for peak intensity detection by two-photon transmission measurement [38], C1, C2: cylindrical lenses, S: sample, PD1–PD3: photodiodes, A1–A3: apertures, SP: spectrometer, and DA: silicon diode-array detection system.
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
3809
For distributed-feedback lasing studies a neat thin 2CzV-MEH-B film was spin-coated on a fused silica plate with three corrugated grating areas (grating length 1.5 mm, grating width 0.5 mm, grating spacing K = 320 nm, 300 nm, and 280 nm, groove width c = K/2, groove depth t = 50 nm, for grating structure see [40]). The pump laser excitation was restricted to an area inside the grating space (pumped area A = 0.9 mm 0.13 mm) to avoid amplified spontaneous emission outside the grating region and to avoid imperfections at the grating – flat-surface borders. The gratings were tilted 45° to the excitation direction. The surface emission was collected and analysed. A scheme of the experimental arrangement is shown in Fig. 2d. The gratings were operated in second grating order (p = 2, Bragg wavelength in medium km,Br = K) and in first diffraction order (M = 1, surface emission) [40]. The grating fabrication was described previously [40].
Fig. 4. The film thickness turned out to be df = 122 nm. The absorption maximum occurs at ka,max = 414 nm. At this wavelength the light penetration depth is d p ¼ a1 f ¼ 57 nm. The refractive index spectrum shows the expected dispersion shape. In the displayed wavelength range the film refractive index is larger than the fused silica or optical glass refractive index. Therefore optical wave-guiding takes place in films above a critical film thickness [41]. For transverse electric modes (TE modes, electrical field vector in plane of the film) the minimum film thickness is [41] " 1=2 # k n2s n2c d min;TE ¼ arctan ; ð1aÞ 1=2 n2f n2s 2pðn2f n2s Þ
3. Results and discussion
k is the considered wavelength in vacuum, nf, ns, and nc are the refractive indices of the film, the substrate, and the surrounding air (nc = 1) at k, respectively. The absorption cross-section spectra, ra(k), of 2CzVMEH-B in THF and of 2CzV-MEH-B neat film are displayed in Fig. 5 (the neat film curve is calculated assuming the same S0–S1 absorption cross-section integral for the solution and the film, see above). The absorption cross-section spectra of the film and the solution are similar, only the solution absorption maximum is approximately 5 nm blue-shifted compared to the film absorption maximum. The 2CzV-MEH-B molecule number density in the film is estimated to be Nf = af/ra,f 8.51 1020 cm3 (calculated
3.1. Optical and spectroscopic characterisation The transmittance spectrum and the reflectance spectrum of a 2CzV-MEH-B thin-film on a fused silica substrate are shown in Fig. 3. The film was prepared by spin-coating of a 2CzV-MEH-B THF solution of dye concentration 15 mg/ml with a speed of 1500 rpm. The absorption coefficient spectrum and the refractive index spectrum together with the film thickness are extracted by a Fresnel equation approach [32,33]. The obtained spectra are shown in
Fig. 3. (a) Transmittance, T(k), and (b) reflectance, R(k), spectra of a 2CzV-MEH-B film on fused silica substrate (solid lines). Film thickness df = 122 nm (determined by Fresnel equation analysis [32,33]). Film was prepared by spin-coating from THF solution. Dash-dotted lines belong to blank fused silica.
and for transverse magnetic films (TM modes, magnetic field vector in plane of the film) it is " 1=2 # k n4f n2s n2c d min;TM ¼ arctan 4 2 : ð1bÞ 1=2 nc nf n2s 2pðn2f n2s Þ
Fig. 4. (a) Absorption coefficient spectrum, af(k), and (b) refractive index spectra, n(k), of 2CzV-MEH-B film and fused silica substrate. Results of distributed-feedback laser analysis on the TE mode refractive index, nf,TE, and the TM mode refractive index, nf,TM, are included.
3810
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
included in Fig. 5. They are calculated from the S0–S1 absorption spectra and the fluorescence quantum distributions (see top part of Fig. 6) using the Strickler-Berg formula [43,44] " #1 R Z 8pc0 n3F em EF ðkÞdk ra ðkÞ dk ; srad;SB ¼ ð2Þ R nA E ðkÞk3 dk abs k em F and the Einstein relation [45,46] rem ðkÞ ¼
Fig. 5. Absorption cross-section, ra(k), and stimulated emission crosssection, rem(k), spectra of 2CzV-MEH-B thin films (solid curves) and of 2CzV-MEH-B dissolved in tetrahydrofuran (THF) (dashed curves, concentration 7.1 105 mol dm3). Upper border, ku, of S0–S1 transition for stimulated emission cross-section calculations is included. Additionally is shown the absorption cross-section spectrum and the stimulated emission cross-section spectrum of Horner-type MEH-PPV in toluene (from [21]) as well as the absorption cross-section spectrum of carbazole (from [42]).
at k = 416 nm). The mass density, q, of 2CzV-MEH-B neat thin films is determined to be q = (Nf/NA)Mm 0.954 g cm3, where NA is the Avogadro constant and Mm = 674.91 g mol1 is the molar-mass. The absorption cross-section spectra of carbazole (from [42]) and of the repeat unit of MEH-PPV in toluene (Horner-type synthesis, from [21]) are included in Fig. 5. The longest-wavelength absorption peak of 2CzV-MEH-B is determined by the absorption of MEH-B. The long-wavelength absorption peak of MEH-PPV is approximately 90 nm red-shifted because of p electron delocalisation over some repeat units. The MEH-PPV chromophore was determined in [21] to extend on the average over about 5.7 repeat units. With the extension the average transition dipole moment per repeat unit is lowered as seen by the lower absorption cross-section of MEH-PPV probably due to conformational changes over the chromophore size. For the CC-PPV copolymer in chloroform [31] the S0–S1 absorption peak occurs at 408 nm at approximately the same position as the S0–S1 absorption peak of 2CzV-MEH-B in THF (410 nm). This indicates that the carbazole copolymer part limits the electron cloud extension to one phenylene-vinylene molecule extension (chromophore size equal to repeat unit size). The 2CzV-MEH-B absorption cross-section spectrum is approximately the sum of one MEH-B sub-unit and two carbazole sub-units of the oligomer. The stimulated emission cross-section spectra, rem(k), of the solution and the neat film of 2CzV-MEH-B are
k4 EF ðkÞ R : 2 8pnF c0 srad;SB em EF ðk0 Þdk0
ð3Þ
In Eqs. (2) and (3) the integrals extend over the regions of S1 ? S0 emission (em) and S0 ? S1 absorption (abs, border ku is indicated in Fig. 5). nF and nA are the average refractive indices in the S0–S1 fluorescence and absorption region, respectively. c0 is the light velocity in vacuum. srad,SB is the theoretical radiative lifetime determined by use of Eq. (2). The shapes of the stimulated emission cross-section spectra of 2CzV-MEH-B in solution and of 2CzV-MEH-B neat film resolve a vibronic structure. The neat film spectrum is about 25 nm red-shifted compared to the solution spectrum, and the spectral half-width of the neat film stimulated emission cross-section spectrum is considerably broader than that of the solution [D~mem (film) 5060 cm1, D~mem (solution) 3680 cm1 (FWHM)]. Additionally the stimulated emission cross-section spectrum of MEH-PPV in toluene is included. The fluorescence quantum distributions, EF (k), of 2CzV-MEH-B in THF and of 2CzV-MEH-B neat film are shown in Fig. 6a. The fluorescence quantum yield, R /F ¼ EF ðkÞdk, of the solution is /F = 0.86 ± 0.02 and
Fig. 6. (a) Fluorescence quantum distributions, EF(k), and (b) degrees of fluorescence polarisation, PF(k), of 2CzV-MEH-B neat film (solid curves, film thickness 122 nm) and of 2CzV-MEH-B in THF (dashed curves, concentration 6.7 105 mol dm3).
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
3811
of the film is /F = 0.062 ± 0.003. The fluorescence efficiency in the solution is high, but in the film it is reduced (self-quenching). The fluorescence peak of the solution occurs at 459 nm, while the fluorescence peak of the film occurs at 480 nm. The degrees of fluorescence polarisation, PF, of 2CzVMEH-B in THF and of 2CzV-MEH-B neat film are shown in Fig. 6b. In the solution it is PF 0.12, and in the neat film it is PF 0.16. For isotropic media with parallel orientation of the absorption and emission transition dipole moments, PF approaches 0.5 if there occurs no reorientation within the fluorescence lifetime (high viscosity and low concentration); and PF approaches 0 if complete reorientation occurs within the fluorescence lifetime (either low viscosity for molecular reorientation, or high concentration for reorientation by site-to-site excitation transfer) [47]. In the solution the obtained degree of fluorescence polarisation is determined by molecular reorientation, while in the neat film it is determined by site-to-site excitation transfer. The degree of fluorescence polarisation, PF, is related to the reorientation time, sor, of the transition dipole moments by [47,48] sor ¼
1=P F;0 1=3 P F sF ; 1 P F =P F;0
ð4Þ
where PF,0 = 0.5, and sF is the fluorescence lifetime. For 2CzV-MEH-B in THF it is PF 0.12 and sF = 1.5 ns (see below) giving sor 375 ps. For 2CzV-MEH-B thinfilm the parameters are PF 0.16 and sF,av 60 ps (see below) giving sor 26 ps. This short reorientation time is caused by fast excitation transfer in the random oriented molecules of the neat film (dipolar Fo¨rster-type energy transfer [49,50] and Dexter-type exchange transfer [50,51]). The temporal fluorescence signals of a 4 105 molar solution of 2CzV-MEH-B in THF (solid line in part (a)) and of a 2CzV-MEH-B neat film (thickness 122 nm, solid line in part (b)) are shown in Fig. 7. The dotted curves show the detection system response functions (attenuated pump pulses directed to micro-channel-plate photomultiplier or streak-camera). A single-exponential fluorescence decay is observed for the dye in solution with a fluorescence lifetime of sF = 1.5 ns. The fluorescence decay of the neat film fits well to a two-exponential fluorescence decay according to SF(t) = SF,0[x1exp(t/sF,1) + x2exp(t/sF,2)] with x1 = 0.72, sF,1 = 47.3 ps, x2 = 1x1 = 0.28 and sF,2 = 238 ps. An average fluorescence lifetime, defined by sF,av = x1sF,1 + x2sF,2 is found to be sF,av = 100 ps. The fluorescence self-quenching seems to be distance and orientation dependent causing the non-single-exponential decay. The self-quenching is thought to be caused by reductive electron transfer (HOMO level of excited molecule is filled by electron from neighbour molecule), oxidative electron transfer (electron in LUMO level of excited molecule moves to LUMO level of a neighbouring unexcited molecule), and charge recombination [50,52,53]. The radiative lifetime determined from the fluorescence lifetime, sF,av, and the fluorescence quantum yield, /F, is
Fig. 7. Temporal fluorescence traces of (a) 2CzV-MEH-B in THF (concentration 3.9 105 mol dm3) measured with micro-channel-plate photomultiplier, and (b) of 2CzV-MEH-B neat film measured with streakcamera (film thickness 122 nm). Dotted curves show response functions.
srad ¼
sF;av : /F
ð5Þ
The experimental results are srad (solution) 1.74 ns and srad (neat film) 1.6 ns. The radiative lifetime in the neat film is slightly shorter than in the solution because of the higher refractive index of the neat film (see Eq. (2): srad / nA n3 F ). The calculated monomeric radiative lifetimes, srad,SB, determined by use of the Strickler-Berg formula (Eq. (2)) give srad,SB (solution) 1.74 ns and srad,SB (neat film) 1.5 ns. In liquid solution the experimental radiative lifetime, srad, and the calculated monomeric Strickler-Berg radiative lifetime, srad,SB (ra is the absorption cross-section per molecule) give the same value within our experimental accuracy, i.e. the emitting chromophore size and the molecule size are the same [45]. This result is expected for diluted solutions (no dimers or higher aggregates present). The determined neat film radiative lifetime, srad, is found to be slighter longer than the calculated monomeric radiative lifetime, srad,SB. The accuracy of srad and srad,SB determination is rather low for the neat film because of the two-exponential fluorescence behaviour. Again the emitting chromophore size and the molecule size are about the same (wave-function of emitting state is restricted to monomeric unit). Some optical and spectroscopic parameters of 2CzVMEH-B in THF and of 2CzV-MEH-B neat films are collected in Table 1.
3812
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
Table 1 Optical and spectroscopic parameters of 2CzV-MEH-B (light emitting oligomer ADS084BE) Parameter
2CzV-MEH-B in THF
2CzV-MEH-B neat film
Comments
ka,max (nm) kem,max (nm) kF,max (nm) nA nF /F PF sF,av (ns) srad (ns) srad,SB (ns) sor (ps)
410 492 459 1.4178 [64] 1.4105 [64] 0.86 0.12 1.5 1.74 1.74 375
414 516 480 1.7 1.8 0.062 0.16 0.1 1.61 1.5 26
Fig. 5 Fig. 5 Fig. 6 For film: Fig. 4b For film: Fig. 4b Fig. 6 Fig. 6 Fig. 7 Eq. (5) Eq. (2) Eq. (4)
3.2. Saturable absorption behaviour The experimental energy transmission, TE, of second harmonic picosecond ruby laser pulses (duration DtP = 35 ps, wavelength kP = 347.15 nm) through a 1 mm cell filled with 2CzV-MEH-B in THF as a function of the input peak pulse intensity, I0P, is shown by the circles in Fig. 8. The transmission rises from the small-signal transmission of T0 = 0.073 at low excitation intensity to TE 0.26 at excitation intensity I0P = 2 1010 W cm2. The solid curves in Fig. 8 are numerical simulations to the transmission measurements. The applied energy level system for the saturable absorption simulations is inserted in Fig. 8. The pump laser excites the 2CzV-MEH-B dye
Fig. 8. Saturable absorption behaviour of 2CzV-MEH-B in THF at wavelength kP = 347.15 nm (excitation with second harmonic laser pulses of ruby laser, pulse duration 35 ps FWHM). Circles are measured. Curves are calculated using excited-state absorption cross-sections rex = 3.5 1017 cm2 (1), 4 1017 cm2 (2), and 4.5 1017 cm2 (3). The inset shows the energy level scheme applied to the excited-state absorption simulations.
molecules from the S0 ground-state 1 to a Franck-Condon level 20 in the S1 band. From there the molecules relax to a thermalized level 2 with the Franck-Condon relaxation time constant sFC (sFC = 0.5 ps is used in the simulations [54]). From the S1 band excited-state absorption occurs to a higher lying singlet band Sn (level 3), and ground-state recovery occurs with the fluorescence lifetime sF. The higher excited molecules relax quickly back to the S1 band with a time constant sex (sex = 60 fs is used in the simulations [55]). The differential equation system for the intensity dependent pump pulse transmission is given in [56] (Eqs. (6–14) there) and are not repeated here. The two-photon absorption contribution of the solvent is included in ð2Þ the calculations by addition of the term aP I 2P to the right ð2Þ 10 side of Eq. (11) in [56] (aP ¼ 3:2 10 cm W1 for THF [57]). For the solid curves in Fig. 8 the excited-state absorption cross-section is varied. The best fit to the experimental data is obtained for rex,P = 4 1017 cm2. The groundstate absorption cross-section is ra,P = 9.4 1017 cm2. 3.3. Lasing performance The transverse pumped lasing of 2CzV-MEH-B/THF in a 1 cm 1 mm dye-laser cell, the wave-guided travellingwave lasing (amplification of spontaneous emission) of neat thin films of 2CzV-MEH-B on microscope carrier plates, and the distributed-feedback laser action of neat thin films of 2CzV-MEH-B on corrugated Bragg gratings etched into a fused silica plate are studied. 3.3.1. Transverse pumped lasing of 2CzV-MEH-B in THF A 4.46 104 molar solution of 2CzV-MEH-B in THF (dye number density N0 = 2.69 1017 cm3) kept in a dyelaser cell of ‘s = 1 cm length and d = 1 mm length was transverse pumped with line-focused picosecond second harmonic pulses of the ruby laser system (pump beam cross-section 14.5 mm 0.245 mm). The dye cell itself formed the optical laser resonator (low-Q resonator): the end surfaces of the cell acted as mirrors (reflectance R = (n 1)2/(n + 1)2 = 0.03527, n = 1.4624 is refractive index of cell fused silica glass at 500 nm). In Fig. 9a an output spectrum of the low-Q laser, SLQL(k), is shown. It belongs to an excitation pulse energy density of w0P = 7.66 104 J cm2. The emission maximum is at kLQL,max = 501 nm. It is slightly red-shifted compared to the wavelength position of peak stimulated emission cross-section which occurs at 492 nm (see Fig. 5). The laser line-width is DkLQL 11.5 nm (FWHM) at the applied excitation pulse energy. The dependences of the laser output energy, WLQL, of the laser wavelength peak position, kLQL,max, and of the spectral half-width (FWHM), DkLQL, on the peak input pump laser energy density, w0P, are displayed in Fig. 10a, b, and c, respectively. In Fig. 10a, above a certain threshold pump pulse energy density of laser action, w0P,th 0.6 mJ cm2, the emission rises steeply beyond the spontaneous emission and amplified spontaneous emission level due to low-Q laser
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
Fig. 9. Spectral shapes of (a) low-Q laser emission, SLQL(k)/SLQL,max, of 2CzV-MEH-B in THF (concentration 4.46 104 mol dm3, excitation energy density w0P = 7.66 104 J cm2), and of (b) travelling-wave laser emission, STWL(k)/STWL,max, of 2CzV-MEH-B neat film (thickness 90 nm, w0P = 4.0 104 J cm2). For comparison the normalized fluorescence quantum distributions, EF(k)/EF,max, are also displayed.
oscillation action. In Fig. 10b it is seen, that the wavelength position of light emission below laser oscillator threshold is at about 490 nm, and above laser threshold is at about 500 nm. The peak position of the stimulated emission cross-section is at about 492 nm. The red-shift of peak laser emission is caused by the S1-state excited-state absorption cross-section spectral distribution (peak of effective stimulated emission cross-section, rem,eff (k) = rem(k) rex(k), occurs at 500 nm). In Fig. 10c it is seen that the spectral half-width, DkLQL, of the light emission shrinks down from about 55 nm before laser action (w0P < w0P,th) to about 12 nm above laser threshold. Low-Q laser emission spectra, SLQL(k), normalized to the excitation pump pulse peak energy density, w0P, are shown in Fig. 11a for various pump pulse peak energy densities. Up to w0P 0.4 mJ cm2 the normalized spectra are unchanged and are determined by the spontaneous emission, Ssp(k) (normal fluorescence). At higher pump pulse energy densities the normalized spectra rise steeply in the wavelength range of maximum stimulated emission cross-section (see Fig. 5). In Fig. 11b the amplification, S LQL ðkÞ=N ex;0 ðw0P Þ S sp ðkÞ=N ex;0 ðw0P;sp Þ h i w S LQL ðkÞ 1 exp w0P;sp P;sat h i ; ¼ w0P S sp ðkÞ 1 exp wP;sat
AðkÞ ¼
ð6Þ
3813
Fig. 10. Low-Q laser oscillator performance of 2CzV-MEH-B dissolved in THF filled in a dye-laser cell (1 cm length, 1 mm thick) transverse pumped with line-focused (beam cross-section 14.5 mm 0.245 mm) second harmonic pulses of a mode-locked ruby laser (pulse duration DtP 35 ps, wavelength kP = 347.15 nm). Dye concentration C0 = 4.46 104 mol dm3. (a) Collected emission signal, WLQL, versus input pump pulse energy density, w0P. Full acceptance angle of fluorescence collection is Dh = 0.83°. Pump laser threshold energy density w0P,th 0.6 mJ cm2. (b) Peak wavelength position of low-Q laser oscillator emission, kLQL,max, versus input pump pulse energy density, w0P. (c) Spectral line-width (FWHM) of low-Q laser oscillator, DkLQL, versus input pump pulse peak energy density, w0P.
of the spontaneous emission is displayed. Thereby Nex,0 is the initial emission-state population number density. It is given by w0P N ex;0 ¼ N 0 1 exp ; ð7Þ wP;sat taking the level population saturation at high excitation energy density into account. wP,sat is the saturation energy density. It is given by [58] wP;sat ¼ hmP =ra;P :
ð8Þ
The second part of Eq. (6) is obtained by insertion of Eq. (7) in the first part of Eq. (6). For Ssp(k)/w0P,sp the curve with w0P = 0.079 mJ cm2 from Fig. 11a is used. Above laser threshold pump pulse energy density, w0P > w0P,th 0.6 mJ cm2 the amplification around the peak laser wavelength rises steeply. Over the whole displayed wavelength range it is A(k) > 1 indicating that rem(k) > rex(k). For vertical polarised excitation and vertical polarised emission detection, the spontaneous emission amplification, A(k), of the short-length low-Q laser oscillator is approximately given by
3814
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
Fig. 11. Spectral lasing performance of low-Q laser oscillator. (a) Normalized spectral low-Q laser output, SLQL(k)/w0P, of picosecond-laser transversely pumped 2CzV-MEH-B/THF in 10 mm 1 mm dye-laser cell. Dye concentration C0 = 4.46 104 mol dm3. Cell wall thickness ‘w = 1.25 mm. (b) Spectral fluorescence light amplification A(k) (Eq. (6)) for several pump laser energy densities, w0P.
the pump laser propagation. ‘s is the amplification length of the dye solution (inner dye cell length). ‘w is the wall thickness of the dye cell. ns and nw are the refractive indices of the dye solution and of the dye cell walls, respectively. c0 is the speed of light in vacuum. d is the inner cell width. w0P is the peak pump laser energy density at the dye cell entrance. aP is the absorption coefficient of the dye solution at the pump laser wavelength kP = c0/ mP. R is the reflectance of the dye cell. sor is the transition dipole reorientation time. rem(k) and rex(k) are the orientation averaged stimulated emission cross-section and the orientation averaged excited-state absorption cross-section of the dye at k, respectively. ra(k) is the ground-state absorption cross-section in the fluorescence spectral region. Its contribution takes into account the reduced fluorescence re-absorption due to ground-state population depletion (the spontaneous emission is more strongly absorbed than the amplified spontaneous emission by this contribution). The orientation dependence of the stimulated emission cross-section is given by rem(k)fem(t) [59]. At time t = 0 the transition dipole moment ~ lS1–S0 is parallel to the vertical polarisation of the excitation and the stimulated emission cross-section is 3rem(k). For times t sor the stimulated emission cross-section is rem(k). The same arguments apply for the ground-state fluorescence re-absorption reduction term ra(k)fem(t). The orien-
AðkÞ ¼ AASE ðkÞ þ ALQL ðkÞ h i R1 R1 2 Rd 2 drer 0 dzeap z 0 dtet=sF exp ½ðrem ðkÞ þ ra ðkÞÞfem ðtÞ rex ðkÞfex ðtÞN ex;0 er ap zt=sF ‘2s 1 ð1 RÞ Rd R1 R1 dr expðr2 Þ 0 dz expðap zÞ 0 dt expðt=sF Þ 1 n h iot=trt R1 R R 2 r2 d ap z 1 t=sF r2 ap zt=sF dre dze dte R exp ½r ðkÞf ðtÞ r ðkÞf ðtÞN e 2‘ em em ex ex ex;0 s 1 0 0 þ ð1 RÞ Rd R1 R1 dr expðr2 Þ 0 dz expðap zÞ 0 dt expðt=sF Þ 1
with the resonator round-trip time ns ‘s þ 2nw ‘w ; ð10Þ trt ¼ 2 c0 and the orientation factors for vertical polarised excitation and vertical polarised emission detection w0P fem ðtÞ ¼ 2 expðt=sor Þ exp þ 1; ð11aÞ wP;sat and fex ðtÞ 2 expðt=sor Þ expðw0P =wP;sat Þ þ 1 for ~ lS1–Sn k~ lS0–S1 ¼ : 1 expðt=sor Þ expðw0P =wP;sat Þ for ~ lS1–Sn ? ~ lS0–S1 ð11bÞ r is the coordinate perpendicular to the direction of the pump laser propagation and perpendicular to the line-focus direction of the pump laser. z is the coordinate along
ð9Þ
tation dependence of the excited-state absorption crosssection is given by rex(k)fex(t). If the transition dipole moments of S1–S0 emission and excited-state absorption are parallel then without considering the ground-state depopulation it is: fex(t) = 2exp(t/sor) + 1. If the transition dipole moments of S1–S0 emission and excited-state absorption are perpendicular to one another then it is without considering the ground-state depopulation: fex(t) = 1 exp(t/sor). Eq. (9) neglects emission level depopulation due to the laser action, and is therefore only valid for moderate amplification factors (here emission state depopulation is given by fluorescence lifetime sF). The first sum term of Eq. (9) takes care of the amplified spontaneous emission, AASE. The effective gain length for this amplified spontaneous emission is approximated by ‘s/2. The second sum term of Eq. (9) describes the amplification of the fed-back light (laser oscillator), ALQL.
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
In Fig. 12a the dotted-line connected circles show the light amplification at 492 nm versus the input pump pulse energy density. The solid curves are calculated A(k) curves (Eq. (9)) using the known dye-laser parameters and varying the excited-state absorption cross-section, rex(492 nm). The excited-state absorption transition dipole moment is assumed to be parallel to the ground-state absorption transition dipole moment (Eq. (11b)). The dashed curves are calculated AASE(k) curves (first sum of Eq. (9)) for the same parameters. The steep rise in amplification occurs at the laser oscillator threshold where the light amplification compensates the output losses (small reflectance R). The best fit of the calculation to the experimental data points is obtained for rex (492 nm) = 1.2 1016 cm2. In Fig. 12b the same experimental data are shown as in Fig. 12a. Only the A(k) and AASE(k) curves are calculated for the situation of the excited-state absorption transition dipole moment perpendicular to the ground-state absorption transition dipole moment (Eq. (11b)). The experimental data cannot be fitted by the theoretical curves. For times t < sor the S1–Sn excited-state transition does not couple to the induced emission since the excited-state absorption cross-section for perpendicular oriented transi-
tion is zero, leading to a higher calculated amplification than experimentally observed. This indicates that for the considered transitions the transition dipole moments are not perpendicular to one another. In Fig. 13 the extracted rem,eff(k) = rem(k) rex(k) and rex(k) spectra together with the rem(k) spectrum (from Fig. 5) are shown for the situation of parallel orientation of the excited-state absorption and the ground-state absorption transition dipole moments. At laser oscillator threshold the light amplification V = exp(rem,eff,LNex,0,th‘s) compensates the reflection losses L = R1. The effective stimulated emission is given by rem,eff,L = rem,L rex,L, where rem,L is the stimulated emission cross-section at the peak laser wavelength, and rex,L is the excited-state absorption cross-section at the peak laser wavelength. It occurs no laser action if rex,L P rem,L. The laser oscillator threshold is defined by expðrem;eff;L N ex;0;th ‘s Þ w0P;th ¼ exp rem;eff;L N 0 1 exp ‘s ¼ R1 ; wP;sat
ð12Þ
where w0P,th is the pump laser threshold energy density. Rewriting Eq. (12) to the effective stimulated emission cross-section gives rem;eff;L ¼
Fig. 12. Spectral light amplification, A(k = 492 nm), versus input pump pulse peak energy density, w0P. The experimental results are shown by the dotted-line connected circles. The dashed curves are calculated amplifications due to the amplified spontaneous emission, AASE (first term of Eq. (9)). The solid curves are calculated total amplifications, A, including amplified spontaneous emission (AASE) and low-Q laser oscillation (ALQL). The top part (a) considers parallel orientation of ground-state and excite-state transition dipole moments. The bottom part (b) considers perpendicular orientation of ground-state and excited-state transition dipole moments. Experimental parameters are used in the calculations (rem (492 nm) = 2.53 1016 cm2). The excited-state absorption crosssection is varied using (1) rex = 0, (2) rex = 1 1016 cm2, (3) rex = 1.5 1016 cm2, (4) rex = 2 1016 cm2, and (5) rex = 4 1016 cm2.
3815
lnðRÞ i ; w N 0 ‘s 1 exp w0P;th P;sat h
ð13Þ
Fig. 13. Cross-sections of 2CzV-MEH-B in THF in lasing spectral region extracted by fit of Eq. (9) to the amplification A(k, w0P = 0.609 mJ cm2) of Fig. 11b. Solid curve: effective stimulated emission cross-section spectrum, rem,eff(k). Dashed curve: stimulated emission cross-section spectrum, taken from Fig. 5. Dash-dotted curve: excited-state absorption cross-section spectrum, rex (k) = rem(k) rem,eff(k).
3816
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
Insertion of experimental values into Eq. (13) (R = 0.03527, kP = 347.15 nm, w0P,th = 6 104 J cm2, wP,sat = 6.09 103 J cm2, N0 = 2.69 1017 cm3, ra,P = 9.4 1017 cm2, ‘s = 1 cm) gives rem,eff,L = 1.33 1016 cm2. This value is in good agreement with the result of Fig. 13. The laser threshold is the lower the higher the dye cell reflectance R, the higher the dye number density N0, the longer the excited dye cell length, the higher the dye absorption cross-section, ra,P, at the pump laser wavelength, the higher the stimulated emission cross-section rem,L, and the lower the excited-state absorption cross-section rex,L. 3.3.2. Wave-guided travelling-wave lasing of 2CzV-MEH-B neat thin films The wave-guided travelling-wave laser action (waveguided amplification of spontaneous emission) was studied using thin films of 2CzV-MEH-B spin-coated from a THF solution (concentration 80 mg/ml, speed 1600 rpm) onto microscope carrier plates (optical glass similar to Schott type BK7) and cut in the film region for optimum edge emission. The films were stored under ambient conditions in the dark before usage. No special drying or aging procedure was applied. In Fig. 9b an edge-emitted travelling-wave laser spectral shape is shown (solid line) for a pump pulse energy density of w0P = 4.0 104 J cm2. For comparison the shape of the fluorescence quantum distribution is included in the figure. The dependences of the collected laser output energy, WTWL, of the laser wavelength peak position, kTWL,max, and of the spectral half-width (FWHM), DkTWL, on the peak input pump laser energy density, w0P, are displayed in Figs. 14a–c, respectively. In Fig. 14a, above a threshold pump pulse energy density, w0P,th 4 105 J cm2, the emission begins to rise beyond the spontaneous emission level due to stimulated emission (amplification of spontaneous emission). At high pump pulse energy density the output signal saturates to a maximum value, WTWL,max 7.5 nJ. The experimental behaviour is fitted by w0P w0P;th W TWL ¼ W TWL; max 1 exp ; ð14Þ w0P;g;sat w0P;th (solid curve) where w0P,g,sat 6 103 J cm2 is the pump pulse energy density of gain saturation. The saturation is thought to be caused by exciton–exciton annihilation processes at high densities of excited molecules [60,61]. In Fig. 14b, the wavelength position of peak light emission, kTWL,max, blue-shifts from about 530 nm at low excitation energy density to 519 nm at high excitation energy density (shift towards the position of peak stimulated emission cross-section, reduction of ground-state re-absorption). In Fig. 14c it is seen that the spectral half-width, DkTWL, of the light emission shrinks around the laser threshold from a spontaneous emission line-width of DkF 85 nm to DkTWL 10 nm, and then remains nearly constant.
Fig. 14. Wave-guided travelling-wave laser performance of a 2CzV-MEHB neat film on an optical glass substrate. Film thickness, df = 90 nm; pumped film area, 5 mm 0.245 mm. (a) Collected emission signal, WTWL, versus input pump pulse energy density, w0P. Full acceptance angle of fluorescence collection is Dh = 0.3 rad. Curve is calculated by use of Eq. (14) with a pump laser threshold energy density w0P,th = 35 lJ cm2, a pump pulse energy density of gain saturation w0P,g,sat = 6 mJ cm2, and a maximum output energy WTWL,max = 7.5 nJ. (b) Peak wavelength of travelling-wave laser, kTWL,max, versus input pump pulse energy density, w0P. (c) Spectral line-width (FWHM) of travelling-wave laser, DkTWL, versus input pump pulse energy density, w0P.
The initial laser slope efficiency, gsl,ini, at threshold concerning the collected TWL light is obtained from Eq. (14) by [21] gsl;ini ¼
1 @W TWL 1 W TWL; max jw0P ¼w0P;th ¼ ; Aexp @w0P Aexp w0P;g;sat w0P;th
ð15Þ
where Aexp is the exposed film area. A value of gsl,ini 0.002 is estimated. It should be noted that the true initial laser slope efficiency is larger since travelling-wave laser emission occurs in forward and backward direction along the pump line focus and the emission angle at the film edge is larger than the acceptance angle of the collecting lens (Dh = 0.3 rad). In Fig 15 the inverse ratio of the pump laser threshold energy density, w0P,th,min/w0P,th, versus the excited film length, ‘ap, is presented in order to determine approximately the effective length of light amplification, ‘TWL. The constant dashed line approximates the behaviour for pumped lengths longer than the effective gain length, and the dash-dotted-line approximates the situation for pump lengths shorter than the effective gain length. From the crossing point of the lines one obtains ‘TWL 1.15 mm.
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
3817
Fig. 15. Normalized inverse laser threshold pump pulse energy density, w0P,th,min/w0P,th, versus exposed film length, ‘ap. The kink of the line gives the effective gain length, ‘TWL. Film thickness df = 90 nm.
If the gain length is limited by ground-state tail absorption, a(kTWL,max), at the peak laser wavelength, kTWL,max 1 530 nm, then it is a(530 nm) ‘1 and the TWL 8:7 cm absorption cross-section is ra(530 nm) = a(530 nm)/Nf 1 1020 cm2. 3.3.3. Distributed-feedback laser action of 2CzV-MEH-B on corrugated gratings Distributed-feedback laser studies have been carried out for 2CzV-MEH-B neat thin films on corrugated gratings etched into a fused silica plate. Grating spacing of K = 320 nm, 300 nm, and 280 nm were used. The grating size was 1.5 mm 0.5 mm (groove length 0.5 mm). The exposed grating area was 0.9 mm 0.13 mm (see above). The surface-emitted radiation was collected and detected. The laser wavelength, kDFB, is proportional to the grating spacing, and it increases with film thickness. TE modes (electrical field vector in film plane) and TM modes (electrical field vector perpendicular to film plane) are excited. A detailed description is given in [40]. Some lasing results are shown in Fig. 16 and some physical and spectroscopic parameters of the DFB lasers are collected in Table 2. In the figure the DFB laser spectra (solid curves) are compared with the travelling-wave laser spectrum and with the shape of the fluorescence quantum distribution. The spin-coated film on the substrate was 320 nm thick (solution 80 mg 2CzV-MEH-B per ml THF, spinning speed 2400 rpm). The DFB laser spectra shown were measured with no polarizer in the detection path. In Fig. 16a (grating spacing K = 320 nm) lasing occurred at kDFB = 513.5 nm (TM1 mode) with a spectral width of DkDFB 2 nm. The applied pump pulse energy was WP 190 nJ, and the collected DFB laser energy was
Fig. 16. Spectra of surface emitting thin-film 2CzV-MEH-B distributed feed-back lasers (solid curves, pumped area 0.9 mm 0.13 mm, full acceptance angle Dh = 36°). For comparison normalized edge-emitted wave-guided thin-film travelling-wave laser emission spectrum (from Fig. 9b) and normalized fluorescence quantum distribution (from Fig. 6) are included. Several parameters are listed in Table 2. (a) DFB laser A: grating spacing K = 320 nm. Film thickness df = 320 nm. Applied pump pulse energy density w0P = 1.6 104 J cm2. (b) DFB laser B: K = 300 nm. df = 320 nm. w0P = 2.2 104 J cm2. (c) DFB laser C: K = 280 nm. df = 320 nm. w0P = 1.9 104 J cm2. Table 2 Physical and spectroscopic parameters of 2CzV-MEH-B distributed feedback lasers Parameter
DFB laser A
DFB laser B
DFB laser C
K (nm) t (nm) c (nm) df(nm) deff (nm) M kDFB (nm) N
320 50 160 320 345 1 513.5 1.605 TM 1 50.35 2.08 69.47 44.58 0.9 0.13
300 50 150 320 345 1 513.5 1.712 TE 1 56.96 2.05 36.15 45.41 0.9 0.13
280 50 140 320 345 1 506.2 1.808 TE 1 58.05 2.13 31.23 43.27 0.9 0.13
j hj (°) nf dmin (nm) hcrit (°) ‘k;exc (mm) ‘\,exc (mm)
1 524.9 1.750 TM 0 68.76 1.88 89.27 50.95 0.9 0.13
Abbreviations: K: groove spacing. t: groove depth. c: groove width. df: film thickness. deff: effective film thickness in grating region. M: diffraction order. kDFB: distributed-feedback laser wavelength. N: effective refractive index. j: guided mode number. hj: propagation angle. nf: refractive index of film. dmin: minimal film thickness for wave-guiding. hcrit: critical angle for total internal reflection (hcrit = arcsin(ns/nf) with ns refractive index of substrate). ‘k;exc : length of exposed grating area. ‘\,exc: width of exposed grating area.
3818
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
WDFB 0.6 nJ. In Fig. 16b (K = 300 nm) lasing occurred at kDFB = 513.5 nm with a spectral line-width of DkDFB 2.6 nm (TE1 mode) and at kDFB = 524.9 nm with DkDFB 1.2 nm (TM0 mode). The pump pulse energy was WP 260 nJ, and the collected DFB laser energy was WDFB 3 nJ. In Fig. 16c (K = 280 nm) lasing occurred at kDFB = 506.2 nm with DkDFB 0.65 nm (TE1 mode). The pump pulse energy was WP 220 nJ, and the collected DFB laser energy was WDFB 1.2 nJ. Depending on the grating spacing and the film thickness the DFB laser wavelength could be tuned across the main part of the fluorescence spectrum. The DFB laser action suppressed the travelling-wave laser action. A theoretical description of the DFB lasing allows the determinations of the polarisation mode (TE or TM, can also be determined by polarizer application), the number of the guided modes (j = 0, 1, 2, . . . e.g. TE0, TE1, TM0, . . .), the refractive index of the film (nf), and the angle of propagation of the wavevector of the mode in the film (hj). These determinations are carried out in the following. The DFB laser wavelength in vacuum, kDFB, is given by [62,63] kDFB ¼
2KN 2K ¼ nf sinðhj Þ; p p
ð16Þ
where N = nf sin(hj) is the effective refractive index, nf is the refractive index of the film, and hj is the wavevector angle of incidence in the film. p is the grating order (here used p = 2). In the case of using the grating in second-order (p = 2) with surface emission the diffraction order is M = 1 (angle of diffraction hd = 0°) and the condition for constructive interference is given by [40] KN ¼ Knf sinðhj Þ ¼ kDFB :
ð17Þ
At fixed grating spacing, K, and film refractive index, nf, the occurring laser wavelength, kDFB, is determined by the allowed propagation angles hj which are determined by the resonance condition [41] 4p 4p nf d eff cosðhj Þ ¼ d eff N cotðhj Þ kDFB kDFB ¼ 2pj þ 2/s þ 2/c
ð18Þ
where j is the number of the guided mode (j = 0,1,2,3,. . .), deff = d + t c/K is the effective film thickness (d is film thickness, t is groove depth, c is groove width), /s is the phase change at the interface between the film and the substrate, and /c is the phase change at the interface between the film and the cover (air in our case). For TE modes these phase changes are (i = s or c) [41] (
1=2 ) n2f sin2 ðhj Þ n2i /TE;i ¼ arctan ; ð19aÞ nf cosðhj Þ and for TM modes they are /TM;i ¼
n2f n2i
/TE;i :
To get a solution of Eq. (18) a minimum film thickness, dmin,TE for the TE0 mode (Eq. (1a)) and dmin,TM for the TM0 mode (Eq. (1b)), is required. The application of Eq. (17) to the measured kDFB values of Fig. 16 determine the experimental effective refractive indices, N = nfsin(hj). At fixed N the application of Eq. (18) to the experimental situation (fixed kDFB, deff) determines the angle of propagation, hj, the film refractive index, nf, the polarisation mode (TE or TM), and the number j of the guided mode (for other parameters no solution at the fixed DFB laser emission wavelength, kDFB). The obtained parameters for the DFB laser spectra of Fig. 16 are listed in Table 2. The obtained film refractive indices, nf,TE and nf,TM, from the DFB laser analysis are included in Fig. 4. The refractive indices obtained for the TE polarisation and the TM polarisation are approximately the same indicating that the spin-coated film on the grating is isotropic (no refractive index anisotropy due to special molecular alignment). The refractive indices obtained from the grating analysis agree reasonably well with the film refractive index data obtained from reflectance and transmittance measurement and Fresnel equation analysis. 4. Conclusions In this paper a dicarbazovinylene-MEH-benzene dye (light emitting oligomer ADS084BE from American Dye Source, Inc.) was characterised by optical constants determination, absorption, emission, saturable absorption, and lasing studies. In liquid solution a transverse pumped low-Q resonator dye cell laser was realized. Wave-guided travelling-wave laser edge emission was achieved for a thin-film spin-coated on a glass substrate. Narrow spectral line-width surface emitting laser action with low laser threshold was obtained by transverse pumping of a spincoated film on corrugated gratings. The excited-state absorption cross-section spectrum in the lasing wavelength region was determined by theoretical analysis of low-Q laser oscillator output. The high fluorescence efficiency of the dye in liquid solution (/F 0.86) was reduced by self-quenching in the neat solid-state film (/F 0.062) which is thought to be due to reductive and oxidative electron transfer followed by charge recombination. The performed laser studies show that the application of the blue and green emitting dicarbazovinylene-MEH-benzene dye ADS084BE in organic light emitting devices may be extended to integrated-optics laser devices. Acknowledgement The authors are grateful to Anja Merkel for excellent technical assistance. References
ð19bÞ
[1] Z.V. Vardeny, O. Korovyanko, in: T.A. Skotheim, J.R. Reynolds (Eds.), Conjugated Polymers. Theory, Synthesis, Properties, and
A.K. Bansal et al. / Optics Communications 281 (2008) 3806–3819
[2]
[3] [4] [5] [6]
[7]
[8] [9] [10] [11] [12]
[13] [14] [15] [16]
[17] [18] [19]
[20] [21] [22]
[23] [24] [25] [26] [27] [28] [29]
Characterization, third ed., Handbook of Conjugated Polymers, CRC Press, Boca Raton, USA, 1962 (Chapter 22). U. Lemmer, A. Haugeneder, C. Kallinger, J. Feldmann, in: G. Hadziioannou, P.F. van Hutten (Eds.), Semiconducting Polymers. Chemistry, Physics and Engineering, Wiley-VCH, Weinheim, Germany, 2000, p. 309 (Chapter 10). W. Holzer, A. Penzkofer, S.-H. Gong, A.P. Davey, W.J. Blau, Opt. Quant. Electron. 29 (1997) 713. G. Heliotis, D.D.C. Bradley, G.A. Turnbull, I.D.W. Samuel, Appl. Phys. Lett. 81 (2002) 415. M.D. McGehee, A.J. Heeger, Adv. Mater. 12 (2000) 1655. W. Holzer, A. Penzkofer, T. Schmitt, A. Hartmann, C. Bader, H. Tillmann, D. Raabe, R. Stockmann, H.-H. Ho¨rhold, Opt. Quant. Electron. 33 (2001) 121. T. Granlund, M. Theander, M. Berggren, M. Andersson, A. Ruseckas, V. Sundstro¨m, G. Bjo¨rk, M. Granstro¨m, O. Ingana¨s, Chem. Phys. Lett. 288 (1998) 879. W. Holzer, A. Penzkofer, H. Tillmann, E. Klemm, H.-H. Ho¨rhold, Synth. Met. 124 (2001) 455. W. Holzer, A. Penzkofer, H. Tillmann, D. Raabe, H.-H. Ho¨rhold, Opt. Mater. 19 (2002) 283. F.P. Scha¨fer (Ed.), Dye Lasers, 2nd revised ed., Topics in Applied Physics, vol. 1, Springer-Verlag, Berlin, 1997. J. Salbeck, M. Scho¨rner, T. Fuhrmann, Thin Solid Films 417 (2002) 20. D. Schneider, T. Rabe, T. Riedl, T. Dobbertin, M. Kro¨ger, E. Becker, H.-H. Johannes, W. Kowalsky, T. Weinmann, J. Wang, P. Hinze, A. Gerhard, P. Sto¨ssel, H. Vestweber, Adv. Mater. 17 (2005) 31. W. Holzer, A. Penzkofer, H.-H. Ho¨rhold, Synth. Met. 113 (2000) 281. R. Philip, W. Holzer, A. Penzkofer, H. Tillmann, H.-H. Ho¨rhold, Synth. Met. 132 (2003) 297. W. Holzer, A. Penzkofer, A. Lux, H.-H. Ho¨rhold, E.B. Kley, Synth. Met. 145 (2004) 119. M. Zavelani-Rossi, G. Lanzani, S. de Silvestri, M. Anni, G. Gigli, R. Cingolati, G. Barbarella, L. Favaretto, Appl. Phys. Lett. 79 (2001) 4082. P.M. Brosenberger, D.S. Weiss, Organic Photoreceptors for Xerography, Marcel Dekker, New York, 1998. M. Pope, C.E. Swenberg, Electronic Processes in Organic Crystals, Oxford University Press, New York, 1999. R. Farchioni, G. Grosso (Eds.), Organic Electronic Materials Conjugated Polymers and Low Molecular Weight Organic Solvents, Springer-Verlag, Berlin, 2001. J. Shinar (Ed.), Light Emitting Diodes. A Survey, Springer-Verlag, New York, 2004. W. Holzer, A. Penzkofer, H. Tillmann, H.-H. Ho¨rhold, Synth. Met. 140 (2004) 155. L.G. Samsonova, N.S. Eremina, V.A. Svetlichnyi, T.N. Kopylova, G.V. Maier, N.S. Kopataeva, E.E. Sirotkina, J. Appl. Spectrosc. 73 (2006) 194. V. Dumarcher, L. Rocha, C. Denis, C. Fiorini, J.-M. Nunzi, F. Sobel, B. Sahraoui, D. Gindre, J. Opt. A: Pure Appl. Opt. 2 (2000) 279. N. Tsutsumi, A. Fujihara, D. Hayashi, Appl. Opt. 45 (2006) 5748. H.-O. Ha, W.-J. Cho, C.-S. Ha, Mol. Cryst. Liq. Cryst. 349 (2000) 443. J. Li, D. Liu, Y. Li, C.-S. Lee, H.-L. Kwong, S. Lee, Chem. Mater. 17 (2005) 1208. J. Ding, J. Gao, Y. Cheng, Z. Xie, L. Wang, D. Ma, X. Jing, F. Wang, Adv. Funct. Mater. 16 (2006) 575. M.A. Baldo, M.E. Thompson, S.R. Forrest, Pure Appl. Chem. 71 (1999) 2095. A.K. Bansal, W. Holzer, A. Penzkofer, T. Tsuboi, Chem. Phys. 330 (2006) 118.
3819
[30] W. Holzer, A. Penzkofer, T. Tsuboi, Chem. Phys. 308 (2005) 93. [31] J.S. Lee, E.J. Oh, J.Y. Park, M.W. Shin, I.H. Cho, Mater. Sci. Eng. B85 (2001) 186. [32] A. Penzkofer, E. Drotleff, W. Holzer, Opt. Commun. 158 (1998) 221. [33] W. Holzer, M. Pichlmaier, E. Drotleff, A. Penzkofer, D.D.C. Bradley, W.J. Blau, Opt. Commun. 163 (1999) 24. [34] A. Penzkofer, W. Leupacher, J. Lumin. 37 (1987) 61. [35] W. Holzer, M. Pichlmaier, A. Penzkofer, D.D.C. Bradley, W.J. Blau, Chem. Phys. 246 (1999) 445. [36] P. Weidner, A. Penzkofer, Opt. Quantum Electron. 25 (1993) 1. [37] A. Penzkofer, Appl. Phys. B 40 (1986) 85. [38] W. Blau, A. Penzkofer, Opt. Commun. 36 (1981) 419. [39] A. Penzkofer, W. Holzer, H. Tillmann, H.-H. Ho¨rhold, Opt. Commun. 229 (2004) 279. [40] W. Holzer, A. Penzkofer, T. Pertsch, N. Danz, A. Bra¨uer, E.B. Kley, H. Tillmann, C. Bader, H.-H. Ho¨rhold, Appl. Phys. B 74 (2002) 333. [41] H. Kogelnik, in: T. Tamir (Ed.), Integrated Optics, Topics in Applied Physics, vol. 7, Springer-Verlag, Berlin, 1979, p. 13. [42] B. Kabouchi, C.K. Assongo, C. Cazeau-Dubroca, Spectrochim. Acta A 57 (2001) 1111. [43] S.J. Strickler, R.A. Berg, J. Chem. Phys. 37 (1962) 814. [44] J.B. Birks, D.J. Dyson, Proc.R. Soc. London, Ser. A 275 (1963) 135. [45] O.G. Peterson, J.P. Webb, W.C. McColgin, J.H. Eberly, J. Appl. Phys. 142 (1971) 1917. [46] A.V. Deshpande, A. Beidoun, A. Penzkofer, G. Wagenblast, Chem. Phys. 142 (1990) 123. [47] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Plenum, New York, 1983. [48] F. Ammer, A. Penzkofer, P. Weidner, Chem. Phys. 192 (1995) 325. [49] Th. Fo¨rster, Fluoreszenz organischer Verbindungen, Vandenhoeck und Ruprecht, Go¨ttingen, Germany, 1951. [50] B. Valeur, Molecular Fluorescence. Principles and Applications, Wiley-VCH, Weinheim, 2002. [51] D.L. Dexter, J. Chem. Phys. 21 (1953) 836. [52] V. Balzani, Electron Transfer in Chemistry, Wiley-VCH, Weinheim, 2001. [53] J. Shirdel, A. Penzkofer, R. Procha´zka, Z. Shen, J. Daub, Chem. Phys. 336 (2007) 1. [54] A. Penzkofer, W. Falkenstein, W. Kaiser, Chem. Phys. Lett. 44 (1976) 82. [55] F. Graf, A. Penzkofer, Opt. Quantum Electron. 17 (1985) 53. [56] W. Holzer, H. Gratz, T. Schmitt, A. Penzkofer, A. Costela, I. Garcı´aMoreno, R. Sastre, F.J. Duarte, Chem. Phys. 256 (2000) 125. [57] S.-H. Gong, A. Penzkofer, Opt. Quant. Electron. 31 (1999) 269. [58] A. Penzkofer, Appl. Phys. B 46 (1988) 43. [59] A. Penzkofer, W. Falkenstein, Chem. Phys. Lett. 44 (1976) 547. [60] R.G. Kepler, V.S. Valencia, S.J. Jacobs, J.J. McNamara, Synth. Met. 78 (1996) 227. [61] W. Holzer, A. Penzkofer, R. Stockmann, H. Meysel, H. Liebegott, H.-H. Ho¨rhold, Synth. Met. 125 (2002) 343. [62] J. Carroll, J. Whiteway, D. Plumb, Distributed Feedback Semiconductor Lasers. SPIE Press Monograph, vol. PM52, The Institute of Electrical Engineers, London, 1998. [63] G. Morthier, P. Vankwikelberge, Handbook of Distributed Feedback Laser Diodes, Artech House, Boston, 1997. [64] K.H. Hellwege, A.M. Hellwege (Eds.), Landolt-Bo¨rnstein. Zahlenwerte und Funktionen aus Physik, Chemie, Astronomie, Geophysik und Technik. 6. Auflage. II. Band. Eigenschaften der Materie in ihren Aggregatzusta¨nden. 8. Teil: Optische Konstanten, Springer-Verlag, Berlin, 1962.