Full text of DOI:10.1007/s003400100619

Appl. Phys. B 73, 105–109 (2001) / Digital Object Identifier (DOI) 10.1007/s003400100619
Applied Physics B
Lasers
and Optics
A microchip-laser-pumped DFB-polymer-dye laser
T. Voss, D. Scheel, W. Schade∗
Institut für Physik und Physikalische Technologien, TU Clausthal, Germany
Received: 1 March 2001/Revised version: 23 May 2001/Published online: 18 July 2001 –  Springer-Verlag 2001
Abstract. A miniaturized, high repetition rate, picosecond
all solid state photo-induced distributed feedback (DFB)
polymer-dye laser is described by applying a passively
Q-switched and frequency-doubled Cr4+:Nd3+:YAG-
microchip laser (pulse width ∆τ = 540 ps, repetition rate
ν = 3 kHz, pump energy Epump = 0.15 µJ) as a pump source.
A poly-methylmethacrylate film doped with rhodamine B
dye serves as active medium. The DFB-laser pulses are tem-
porally and spectrally characterized, and the stability of the
thin polymer/dye film at high repetition rates is analyzed.
The shortest DFB-laser pulses obtained have a duration of
11 ps. After the emission of 350000 pulses the intensity of
the DFB-laser output has decreased by a factor of two and
the pulse duration has increased by a factor of 1.2. For single
DFB-laser pulses of 20-ps duration the spectral bandwidth is
measured to be ∆λ = 0.03 nm, which is only 0.005 nm above
the calculated Fourier limit assuming a Gaussian profile for
the temporal shape of the pulses. Coarse wavelength tuning of
the DFB laser between 590 and 619 nm is done by turning the
prism. Additionally, a fine tuning of the DFB-polymer-laser
wavelength is achieved by changing the temperature of the
polymer laser: (1) the use of a laser-active polymer such as
a ladder-type poly(p-phenylene)(LPPP) to prepare a thin film
with a permanent DFB structure [7–9] and (2) the use of
a thin film of a laser-active polymer or a polymer doped with
a laser dye to generate a photo-induced distributed feedback
structure for the duration of the pump pulse [10–14]. For
the latter, the fabrication of the film is much easier, since no
permanent structure has to be printed into the material. Fur-
thermore, the DFB laser emits short pulses in the picosecond
regime and its wavelength can easily be tuned within the gain
profile of the polymer or the dye.
In this paper we report on the temporal and the spectral
characterization of a DFB-polymer laser based on rhodamine-
B-doped poly-methylmethacrylate (PMMA) as active laser
medium. We show that a simple prism setup and a frequency-
doubled Cr4+:Nd3+:YAG-microchip laser [15, 16] can be
used to generate single DFB-laser pulses with a temporal
width of typically 20 ps at repetition rates of up to 3 kHz. The
pulse durations calculated by a numerical simulation based
on a rate-equation model developed by Bor and Müller [17]
are in good agreement with the experimental results. The
dependence of the DFB wavelength on the temperature
(λDFB ∼ n(T)) can easily be used for a fine tuning or a stabi-
lization of the output wavelength. Coarse wavelength tuning
between 590 and 619 nm and the selection of a center wave-
length is done by changing the angle of incidence of the pump
beam. Furthermore, our results indicate that efficient pump-
ing of the DFB structure by the frequency-doubled microchip
laser at λ = 532 nm and Epump = 0.15 µJ gives significantly
larger lifetimes of the polymer/dye film compared to recent
results applying ultraviolet pulses [14]. In this investigation
a 50% decrease of the initial DFB-pulse intensity is obtained
after 350000 pulses.
dλ
polymer/dye layer ( = −0.05 nm/◦C) in the range from 20
dT
to 40 ◦C.
PACS: 42.70.Jk; 42.55.Mv; 42.60.By
Distributed feedback (DFB) dye lasers are well known as
simple laser sources providing nearly bandwidth-limited pi-
cosecond pulses for spectroscopic applications [1–6]. Typ-
ically, dyes that are dissolved in organic solvents, e.g.
methanol, are used as laser-active media. When pumped by
excimer or Nd:YAG lasers, output powers of several nJ/pulse
and repetition rates between 10–100 Hz are obtained.
Recently, the application of laser-active polymers and
polymers doped with conventional laser dyes as active media
for photo-induced distributed feedback lasers has been inves-
tigated. There are two different approaches towards a DFB-
1 Experimental setup
The experimental setup is shown in Fig. 1. A frequency-
doubled microchip laser is used as a pump source for
a DFB-polymer laser. The microchip laser consists of
a Cr4+:Nd3+:YAG crystal (ALPHALAS GmbH) which com-
∗Corresponding author.
(Fax: +49-5323/723-600, E-mail: wolfgang.schade@tu-clausthal.de)

106
Fig. 2b). The repetition rate of the passively Q-switched mi-
crochip laser is determined by the concentration of the Cr4+
ions, which act as a saturable absorber in the YAG crystal. An
acousto-optical modulator is used as a pulse picker to tune the
repetition rate between 10 Hz and 1.6 kHz.
The pump beam is imaged on the polymer/dye film
of the prism-DFB laser by a system of a cylindrical lens
( f = 90 mm) and a spherical lens ( f = 150 mm). The layer
is prepared by first solving 20 mg rhodamine B dye and
then 3.7 g PMMA pellets in a mixture of 10 ml chloroform
and 0.8 ml DMSO (dimethylsulfoxide). This solution is after-
wards spin-coated directly onto the side AC of a 90◦-SF11
prism and baked for one hour at a temperature of T = 80 ◦C
(see Fig. 3). The thickness of the layer is adjusted by chang-
ing the concentration of the PMMA in the solution, the tem-
perature of the solution which determines the viscosity and by
changing the rotational frequency of the spin-coater. The best
performance of the DFB laser is obtained with films having
a thickness between 1 and 3 µm. The prism material SF11 is
chosen because of its high index of refraction n = 1.795 to en-
sure that the generated DFB-laser wavelengths are within the
gain profile of the rhodamine B laser dye (refer to (1)).
The pump beam enters the prism through the side AB.
One part of the beam is refracted on the polymer/dye film di-
rectly, while the other part is totally reflected on the side BC
of the prism. The two parts of the beam form an interference
pattern in the polymer/dye film. The length of this pattern
is approximately L = 2.0 mm and its height is b = 0.1 mm.
This interference pattern acts as distributed feedback struc-
ture of the laser and only exists for the duration of the pump
pulse (see Fig. 3). Since the wavelength selection is achieved
by first-order Bragg scattering, the wavelength of the DFB-
polymer laser in vacuum is given by the relation λ = 2npΛ,
where Λ is the separation of the interference fringes [2]. In
the case of the prism geometry which is used in this inves-
tigation, the wavelength of the DFB-polymer laser is given
by [18]:
Fig. 1. Experimental setup for the microchip-laser-pumped DFB-polymer
laser
bines the laser medium and the passive Q-switch in one crys-
tal and is pumped by a continuous-wave (cw) diode laser (Os-
ram SPL CG81, P = 1 W, λ = 808 nm) via a gradient-index
lens. The pulsed output of the microchip laser (λ = 1064 nm)
is frequency-doubled in a KTP crystal. A bandpass filter
(Schott BG 39) is used for the separation of the 532-nm
beam from the fundamental. The microchip laser typically
generates pulses with a duration of 540 ps at 532 nm and
a repetition rate of 3 kHz (see Fig. 2a). The pulse to pulse
fluctuations in the output intensity are less than 2% (see
√
ꢀ
ꢁ
npλp
np 2λp
α
3 α2
2 n2
λ =
≈
1−
+
+. . . .
(1)
n sin γ
n
n
Here, np = 1.49 is the index of refraction of the polymer/dye
film, n = 1.79 the index of refraction of the prism, λp =
a
Time (ps)
b
Pulses Measured
Fig. 2. a Temporal shape and b pulse to pulse stability of the frequency-
doubled microchip laser
Fig. 3. Setup of the prism-DFB-polymer-dye laser

107
532 nm the wavelength of the pump beam, α the angle of in-
cidence of the pump beam on the prism surface and γ the
angle of incidence on the polymer/dye film. The expansion
(1) holds for |α| ꢀ 1. For this geometry only one standing
wave can be generated in the DFB resonator. The corres-
ponding wavelength can be calculated by (1). It further shows
that for a fixed pump wavelength λp the wavelength of the
DFB-polymer laser can be tuned by changing the angle of
incidence α and/or the indices of refraction n and np.
The pulses of the DFB-polymer laser are temporally ana-
lyzed using a streak camera (Optronis Optoscope, time reso-
lution < 2 ps, repetition rate < 2.5 kHz). The spectral analy-
sis is done with an optical multichannel analyzer (OMA4,
EG&G Princeton Applied Research, spectral resolution of
≈ 0.1 nm) and a 0.5-m monochromator with a spectral reso-
lution of ≈ 0.01 nm.
types of lasers, the non-linearity of τc in (4) is responsible
for the typical self-cavity-dumping of the DFB laser, which
makes the laser emit short pulses in the picosecond regime.
The equations (2) to (5) are solved numerically for pa-
rameters describing the PMMA/rhodamine B film [19, 20]
and are referred to in Table 1. The results of the simulation
for two different values of the pump energy are shown in
Figs. 4a and 5a. With Epump = 0.09 µJ the DFB laser is oper-
ated just slightly above the laser threshold. Then only a single
pulse with a duration of ∆τ = 21 ps is emitted. When in-
creasing the pump energy to Epump = 0.15 µJ the simulation
2 Temporal characterization of the DFB-polymer laser
The temporal behavior of the generated DFB-laser pulses is
analyzed experimentally by recording streak images, and the
measured results are compared to numerical simulations that
apply a system of coupled differential equations given by Bor
and Müller [17]:
a
Time (ps)
dN
dt
dQ
dt
σec
np
N
= Ipσp (N0 − N)−
(σe −σa) c
NQ −
,
(2)
(3)
τc
Q
ΩN
=
NQ −
+
,
np
τc
τf
npL3
8π2c
hcQ
τc =
Pd =
[N (σe −σa) V]2 ,
Lab.
(4)
(5)
2λτc
In these equations N(t) describes the density of dye molecules
in the first excited singlet state, Q(t) the density of DFB
photons, τc the average lifetime of the photons in the DFB
resonator and Pd the output power of the DFB laser. N0 is
the density of dye molecules in the PMMA film, Ip the spa-
tially averaged pump photon flux per unit area, σp the absorp-
tion cross section for the pumping wavelength for the ground
state, σa the absorption cross section for the DFB laser wave-
length for the first excited singlet state to the second, σe the
stimulated emission cross section for the laser wavelength
and τf the fluorescence lifetime of the first excited state. c
is the speed of light in vacuum, L the length of the pumped
volume, a the penetration depth of the pump light into the
polymer/dye layer, b the height of the excited volume, V the
visibility of the interference fringes and np the index of re-
fraction of the polymer/dye layer. Ω determines the fraction
of the spontaneous emission which propagates into the an-
gular and the spectral ranges of the DFB-laser beam and h
is Planck’s constant. Whereas (2) and (3) hold for different
b
Time (ps)
c
Pulses Measured
Fig. 4. a Numerical simulation and b experimental data for Epump
=
0.09 µJ. Duration of the single pulse obtained with a Gaussian fit:
a ∆τ = 22 ps and b ∆τ = 21 ps. c Pulse to pulse stability of the DFB-laser
output for single-pulse operation
Table 1. Values used for the numerical simulation [19, 20]
N0
σp
σe
σa
τf
ꢀ
η
c
V
7.6×1024 m−3
3.2×10−20 m2
1.5×10−20 m2
< 0.7×10−20 m2
3.9×10−9 s
1.18 kgm−3
1.49
3×108 ms−1
1

108
tral bandwidth of a 20-ps single DFB-laser pulse is meas-
ured to be ∆λ = 0.03 nm (resolution of the spectrometer
≈ 0.01 nm). This is only 0.005 nm above the theoretically
2
0.41λ
c∆τ
calculated Fourier limit ∆λ =
= 0.025 nm assuming
a Gaussian-shaped single pulse of 20-ps duration.
A fine tuning of the DFB-polymer-laser wavelength by
temperature control of the polymer/dye film was also investi-
gated [27]. Our results show that by changing the temperature
of the prism and of the polymer/dye film from 20 to 41 ◦C
the DFB wavelength can be tuned from 616.2 nm to 615.2 nm
(see Fig. 6). A linear fit to the measured data gives a depen-
dence of the wavelength on the temperature of
a
Time (ps)
dλ
dT
nm
= −0.05 ◦
.
(6)
C
Therefore, a coarse tuning of the wavelength can be done
by changing the angle γ, and a fine tuning of about 1 nm is
achieved by controlling the temperature of the prism and the
polymer/dye film.
The dependence of the DFB-laser wavelength on the tem-
perature is calculated by expanding the indices of refraction
in (1) in powers of (T − T0). For fixed values of γ and λp the
wavelength is approximately given by
n0 −n1 (T − T0)
b
Time (ps)
λ (T) = C
.
(7)
n2 +n3 (T − T0)+n4 (T − T0)2
Fig. 5. a Numerical simulation and b experimental data for Epump
=
0.15 µJ. Duration of the first pulse obtained with
a ∆τ = 9 ps and b ∆τ = 13 ps
a Gaussian fit:
As a starting point for the simulation the measured wave-
length λ = 616 nm for a temperature T0 = 24 ◦C is chosen.
Using (7) the constant C is determined with these initial
values to be C = 741.1 nm. The constants n0 and n1 are used
for the approximation of the index of refraction of PMMA.
The constants n2 −n4 characterize the prism material SF11.
These constants are taken from [21, 22] and are listed in
Table 2. The results of this calculation (line) together with the
experimental data (dots) are shown in Fig. 6.
predicts two DFB-laser pulses, with the first pulse having
a duration of ∆τ = 9 ps. The second pulse is emitted 260 ps
after the first and has a duration of 17 ps.
The corresponding experimental data are shown in
Figs. 4b and 5b. A good agreement between experiment and
simulation in the predicted pulse durations, their spacing to
each other and their relative intensities is obtained. Typical
single-pulse durations are ∆τ = 20 ps. When increasing the
pump energy from Epump = 0.09 µJ to Epump = 0.15 µJ (the
highest pump energy that is obtained with the present experi-
mental setup), a less intense second pulse with a duration of
21 ps appears 240 ps after the first pulse, while the duration
of the first pulse decreases from ∆τ = 21 ps to ∆τ = 13 ps.
When changing the repetition rate between 20 and 800 Hz no
significant change of the temporal output characteristics of
the DFB-polymer-dye laser is measured.
In Fig. 4c the pulse to pulse fluctuation in the output in-
tensity of the DFB-polymer-dye laser is shown for 50 single-
shot measurements. While the intensity of the microchip laser
shows a fluctuation of 2% (which is the standard deviation;
refer to Fig. 2b) the pulse to pulse stability of the DFB-laser
output is 6% (see Fig. 4c).
Temperature (°C)
Fig. 6. Fine tuning of the DFB wavelength by changing the temperature of
the prism and polymer/dye film from 20 to 41 ◦C The dots correspond to
the experimental data. The line is the result of the calculation using (7)
3 Spectral characterization of the DFB-polymer laser
Table 2. Values for the refractive indices of PMMA and SF11 [21, 22]
The wavelength of a photo-induced DFB-polymer-dye laser
can easily be tuned by changing the angle of incidence
γ of the pump beam on the polymer/dye film (refer to
(1)) [12–14]. In the present experiment a tuning range of the
wavelength from 618.8 to 590.0 nm is obtained. The spec-
ꢄ
ꢅ
ꢂ
ꢃ
ꢂ
ꢃ
1
K
1
K
1
K
n0
n1
n2
n3
n4
2
1.492
1.1×10−4
1.795
1.12×10−5 1.18×10−8

109
4 Stability of the polymer/dye film
The wavelength of the almost bandwidth-limited prism-DFB-
laser pulses can be tuned over a 30-nm range by rotating
the prism, and a fine tuning over a range of about 1 nm and
a stabilization of the DFB wavelength is achieved by control-
ling the temperature of the prism and the polymer/dye film.
The temporal characteristics of the DFB-polymer-laser out-
put are quantitatively verified by computer simulations using
a simple rate-equation system that is successfully applied
to a DFB-polymer-dye laser. Compared to previous experi-
ments the ‘lifetime’ of the PMMA/rhodamine B film film
is increased to 350000 pulses when using the frequency-
doubled output of a passively Q-switched microchip laser
as pump source. The photophysical properties of the layer
concerning the characterization of the photo-induced DFB
structure and the photodegradation of the laser dye are cur-
rently under investigation with a near-field scanning optical
microscope.
Practical applications of DFB-polymer-dye lasers are limited
by the lifetime of the polymer/dye film. Recently reported
results indicate [14, 20, 23–26] that photodegradationand dif-
fusion of the dye molecules probably caused by the local
heating of the layer by the pump pulse contribute to a de-
crease of the output power. Most of these experiments are
performed at repetition rates between 5 and 20 Hz. They show
a ‘lifetime’ of several tens of thousand pulses depending on
the wavelength and the power of the pump laser, where the
‘lifetime’ is defined as the number of pulses after which the
output power has decreased to 50% of its initial value.
In the present investigation a typical ‘lifetime’ of 350000
pulses for the PMMA/rhodamine B film is obtained with
a microchip laser as pump source for the DFB-polymer-dye
laser (see Fig. 7). Furthermore, our results show that a change
of the repetition rate between 20 and 800 Hz has no sig-
nificant influence on the stability of the polymer/dye film.
Additionally, we analyzed the stability of the polymer/dye
film at different temperatures. Between 6 and 45 ◦C no sig-
nificant influence on the ‘lifetime’ is measured. However, at
higher temperatures the performance of the laser dye signifi-
cantly decreases and the ‘lifetime’ is lowered, which is in
agreement with previous results [27]. This indicates that at
moderate temperatures only the local heating caused by the
pump beam contributes to the photodegradation of the laser
dye, whereas the global temperature of the polymer/dye film
has no significant influence on the stability.
Acknowledgements. The authors wish to thank Heinz Christian Rost for the
preparation of the polymer/dye films.
References
1. C.V. Shank, J.E. Bjorkholm, H. Kogelnik: Appl. Phys. Lett. 18, 395
(1971)
2. Z. Bor: Opt. Commun. 29, 103 (1979)
3. F.P. Schäfer: Laser Optoelektronik 2, 95 (1984)
4. H. Kogelnik, C.V. Shank: J. Appl. Phys. 43, 2327 (1972)
5. Z. Bor, A. Müller, B. Racz: Opt. Commun. 40, 294 (1982)
6. S. Szatmari, B. Racz: Appl. Phys. B 43, 173 (1987)
7. C. Kallinger, M. Hilmer, A. Haugeneder, M. Perner, W. Spirkl, U. Lem-
mer, J. Feldmann, U. Scherf, K. Müllen, A. Gombert, V. Wittwer: Adv.
Mater. 10, 920 (1998)
8. A. Haugeneder, M. Neges, C. Kallinger, W. Spirkl, U. Lemmer, J. Feld-
mann, M.-C. Amann, U. Scherf: Appl. Phys. B 66, 389 (1998)
9. A. Haugeneder, M. Hilmer, C. Kallinger, M. Perner, W. Spirkl, U. Lem-
mer, J. Feldmann, U. Scherf: J. Appl. Phys. 85, 1124 (1999)
10. M. Maeda, Y. Oki, K. Imamura: IEEE J. Quantum Electron. QE-33,
2146 (1997)
11. Y. Oki, K. Ohno, M. Maeda: Jpn. J. Appl. Phys. 37, 6403 (1998)
12. X. Zhu, S. Lam, D. Lo: Appl. Opt. 39, 3104 (2000)
13. V. Dumacher, L. Rocha, C. Denis, C. Fiorini, J.-M. Nunzi, F. Sobel,
B. Sahraoui, D. Gindre: J. Opt. A: Pure Appl. Opt. 2, 279 (2000)
14. M. Fukuda, K. Mito: Jpn. J. Appl. Phys. 39, 5859 (2000)
15. J.J. Zayhowski: Opt. Lett. 21, 588 (1996)
16. B. Braun, F.X. Kärtner, U. Keller, J.-P. Meyn, G. Huber: Opt. Lett. 21,
405 (1996)
17. Z. Bor, A. Müller: IEEE J. Quantum Electron. QE-22, 1524 (1986)
18. S. Chandra, N. Tekeuchi, S.R. Hartmann: Appl. Phys. Lett. 21, 144
(1972)
Number of Pulses x 10-3
Fig. 7. Stability of the polymer/dye film, ν = 100 Hz, T = 22 ◦C, Epump
=
0.15 µJ
19. A. Tagaya, S. Teramoto, E. Nihei, K. Sasaki, Y. Koike: Appl. Opt. 36,
572 (1997)
20. A. Deshpande, E. Namdas: J. Lumin. 91, 25 (2000)
21. K. Horie, H. Ushiki, F.M. Winnik: Molecular Photonics (Wiley-VCH,
Weinheim 2000) p. 162
5 Summary
We have shown that a miniaturized microchip-laser-pumped
DFB-polymer laser with an active medium that consists of
a PMMA layer doped with rhodamine B dye molecules gen-
erates single laser pulses with a duration of ∆τ = 20 ps at
repetition rates of up to 3 kHz. The layer can easily be fab-
ricated by the spin-coating method, and only a prism and
a system of two lenses are necessary for the DFB-laser setup.
22. Datasheet from Schott Glaskatalog (2000)
23. W. Hu, H. Ye, C. Li, Z. Jiang, F. Zhou: Appl. Opt. 36, 579 (1997)
24. A. Costela, I. Garcia-Moreno, J. Barroso, R. Sastre: Appl. Phys. B 70,
367 (2000)
25. S. Popov: Appl. Opt. 37, 6449 (1998)
26. A. Costela, I. Garcia-Moreno, J.M. Figuera, F. Amat-Guerri, R. Mal-
lavia, M.D. Santa-Maria, R. Sastre: J. Appl. Phys. 80, 3167 (1996)
27. W. Schade, B. Garbe, V. Helbig: Appl. Opt. 29, 3950 (1990)