Fluorescence Enhancement of Europium Ions in a Scattering Matrix
Mitsunori Saito and Takahiro Koketsu
Department of Electronics and Informatics, Ryukoku University, Seta, Otsu 520-2194, Japan
Keywords: Fluorescence, Europium, Polyethylene Glycol, Scattering, Droplet Laser.
Abstract: Microlasers are usually composed of organic dyes that emit fluorescence with a high efficiency. Those dyes,
however, lose their fluorescence function in a short time because of optically- or thermally-induced bleaching.
This degradation is particularly serious with microdevices, since a high-powered beam is focused into a small
volume of the device. The problem of the device degradation can be solved, if organic dyes are replaced by
fluorescent lanthanide ions (europium, erbium, neodymium, etc.) that have a superior durability against
optical and thermal hazards. The lanthanide ions, however, have a smaller absorption cross-section than
organic dyes, and hence, a pump light for exciting the ions is absorbed insufficiently inside a microdevice. A
long optical path is therefore required to enhance the excitation efficiency. Polyethylene glycol is a useful
solvent for dispersing europium ions, since it turns to a translucent matrix by solidification. In this translucent
matrix, pump light (396 nm wavelength) is scattered heavily, which leads to extension of the optical path and
enhancement of the absorbance. Consequently, fluorescence of the europium ions (613 nm) becomes twofold
stronger in the solid phase than the liquid phase.
1 INTRODUCTION
Microfluidic devices are attracting interests in
various technical fields including chemical analysis,
biological sensing, and material synthesis (Hawkins
and Schmidt, 2010). Organic dyes are usually used
for yielding a fluorescence function in the liquid.
Microdroplet laser is a unique fluidic device in
which fluorescence circulates numerous times indu-
cing a stimulated emission. A droplet of dye solution
self-forms a perfect sphere with a smooth surface,
which facilitates the light circulation. In addition,
deformable spheres realize the emission wavelength
tuning. Since these unique features are unattainable
with solids, droplet lasers have been studied
extensively in the last three decades (Tzeng et al,
1984). Disadvantages of the droplet lasers include
handling difficulty (fluidity) and instability
(volatility), which are essential features of liquids.
These problems are avoidable by enclosing a droplet
in a silicone rubber (Saito et al, 2008); i.e., whereas
an enclosed droplet can be handled like a solid, its
fluidity is preserved in a flexible silicone rubber.
In spite of the improvement above, microdroplet
lasers still have another problem; i.e., organic dye
molecules degrade rapidly because of photochemical
or thermochemical reactions during excitation and
emission processes. This degradation generally takes
place in organic dye lasers (Gersborg-Hansen et al,
2007). Conventional dye lasers, therefore, circulate a
dye solution to replace bleached molecules by fresh
ones. As regards dye-doped polymer lasers, disk
rotation (Kytina et al, 2004) or molecular diffusion
(Yoshioka et al, 2012) are achieved to extend the
device lifetime. Unfortunately these methods are
unadoptable for microdroplet lasers, although the
dye bleaching is even more serious in a small droplet
that is exposed to a strong laser beam (Barnes et al,
1993).
A possible solution for the degradation problem
is replacement of organic dyes by lanthanide
elements, e.g., europium (Eu), erbium (Er),
neodymium (Nd), thulium (Tm), samarium (Sm), etc.
Photoluminescence of these ions are well known
(Shionoya and Yen, 1999) and widely used in lasers,
amplifiers, and illuminators. Many researchers
fabricated microspherical lasers by doping
lanthanide ions in crystals and glasses, e.g., Sm-
doped fluorite (Garret et al, 1961), Nd-doped glass
(Miura et al, 1996), and Er-doped glass (Klitzing et
al, 2000). A droplet laser was also fabricated by
using a Eu chelate (Lin et al, 1992).
The lanthanide ions, however, have a disadvanta-
ge that an absorbance is too low to absorb pump
light efficiently in a small droplet. In this respect,
organic dyes are superior to the lanthanide ions. It
Saito M. and Koketsu T.
Fluorescence Enhancement of Europium Ions in a Scattering Matrix.
DOI: 10.5220/0006089600150021
In Proceedings of the 5th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2017), pages 15-21
ISBN: 978-989-758-223-3
Copyright
c
2017 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
15
has been reported, for example, that the pump-light
absorbance of rhodamine is 10,000-fold stronger
than that of Er (Tagaya et al, 1997). The optical path
has to be extended to absorb pump light efficiently.
Dispersion of scattering particles is effective to
extend the optical path (Watanabe et al, 2005).
Turbid media are also attracting interests for
development of random lasers (Wiersma et al, 1995).
Polyethylene glycol (PEG) is a useful matrix for
dispersing fluorescent materials, since it dissolves
various dyes and ions in the liquid phase and
becomes translucent in the solid phase (Saito and
Nishimura, 2012). Random lasing has been realized
by the use of scattering in the solid PEG that
contains rhodamine 6G (Saito and Nishimura, 2016).
In this study, we prepared a PEG solution of
europium chloride (EuCl
3
) and examined the
enhancement of the pump-light absorbance and
fluorescence in this scattering solid.
2 SAMPLE PREPARATION
Figure 1(a) shows transmission spectra of aqueous
solutions of rhodamine 6G (typical organic dye) in a
sample cell of 10 mm thickness. A strong absorption
band emerged when the concentration exceeded 10
5
mol/l. As the black line in Fig. 1(b) shows, however,
Figure 1: Transmission spectra of the aqueous solutions
containing (a) rhodamine 6G (10
-6
10
-4
mol/l) or (b) EuCl
3
(10
-2
mol/l). The solutions were put in a sample cell of 10
or 50 mm thickness. Pure water in the same sample cell
was used as a blank for the transmittance evaluation.
an aqueous solution of EuCl
3
in the same sample cell
(10 mm) exhibited only a trace of absorption (394
nm) in spite of 1000-fold increase in the
concentration (10
2
mol/l). The gray line shows the
absorption band that became clear by extension of
the sample thickness (50 mm). However, both the
depth and width of the absorption band were still
smaller in a Eu solution than a rhodamine solution.
These facts indicate that europium needs a long
optical path to absorb pump light. In addition, the
pump light wavelength has to be tuned accurately to
meet the absorption wavelength.
To extend the optical path, we used PEG as a
solvent. As Fig. 2(a) shows, the melting point of
PEG changes notably depending on the molecular
weight. PEG with a molecular weight of 300 (PEG
300), for example, takes a transparent liquid phase at
room temperature, whereas PEG 1000 takes a
translucent solid phase. Figure 2(b) shows
transmission spectra of PEG 300 and PEG 1000 in a
sample cell of 10 mm thickness. PEG 300 is
transparent over the visible spectral range, whereas
PEG 1000 scatters visible light too strongly to allow
transmission. In the sample preparation process,
PEG 1000 was heated to ~50 °C for dissolving
EuCl
3
, and then cooled to room temperature. No
precipitation took place in the solidification process,
and Eu ions were dispersed uniformly in the solid.
Figure 2: (a) Melting points of PEGs with various
molecular weights. (b) Transmission spectra of PEGs at
room temperature. The photograph shows the PEGs in the
liquid and solid phases.
PHOTOPTICS 2017 - 5th International Conference on Photonics, Optics and Laser Technology
16
3 ABSORBANCE
Spectral measurement was conducted for the PEG
300 that contained EuCl
3
of 10
2
mol/l. The solution
was put into a glass cell of 1050 mm thickness. The
PEG containing no salts [the black line in Fig. 2(b)]
was used as a blank for evaluation of the
transmittance. Figure 3(a) shows the transmission
spectra that were measured for the samples of 10 or
50 mm thickness. The absorption band at 394 nm is
slightly weaker in the PEG than water [Fig. 1(b)]. In
the ultraviolet region (<380 nm), however, the
transmittance decreases more notably in the PEG
than water. We used the absorption band at 394 nm
in the current experiment, since no suitable
ultraviolet laser was available in our laboratory.
Figure 3: (a) Transmission spectra of EuCl3 in PEG 300
(liquid). The concentration was 102 mol/l and the sample
thickness was 10 or 50 mm. (b, c) The optical densities of
EuCl3 in PEG 300 or water. The numerals beside the
marks denote wavelengths. The concentration was 102
mol/l.
Figure 3(b) shows the absorbance or the optical
densities (logT) that were evaluated from the
transmittances (T
) at several wavelengths around the
absorption band. The optical density increases in
proportion to the sample thickness, although some
deviation is visible at 30 mm probably due to
insufficient optical alignment in the measurement
process. The highest absorbance is attainable at 394
nm, and it is halved at 396 nm. Figure 3(c) shows
the result for the aqueous solution [Fig. 1(b)].
Although the absorbance at 394 nm is slightly higher
in water than PEG 300, these solutions will exhibit
similar absorption efficiency for the pump light. As
regards the PEG 1000 (solid), the transmission
spectrum was not measurable, since no light passed
through the sample [the gray line in Fig. 2(b)].
4 FLUORESCENCE
The results in Fig. 3 indicate that efficient excitation
is achievable at around 394 nm wavelength. We
used a laser diode of 396 nm in the current
experiment. The laser power was 100 mW. Figure 4
shows an optical setup for fluorescence measure-
ments. Samples were put into a glass cell of 10 or 50
mm thickness. The laser beam was focused on the
input side of the glass cell. Fluorescence was picked
up in the direction perpendicular to the laser beam
axis, since no strong light emerged from the output
side of the glass cell when the sample was
translucent (PEG 1000). As the photograph in Fig. 4
shows, the transparent sample (PEG 300) also
emitted fluorescence in the perpendicular direction.
The collection lens and the probe fiber (core
diameter: 400 μm) were moved to change the pickup
position, i.e., the distance z from the input side. The
fluorescence spectrum was measured by using a
multichannel spectrometer (B & W Tek, BTC112E).
Figure 4: Optical setup for the fluorescence measurement.
Fluorescence was picked up from the sample side or the
exit end by using a probe fiber of the multichannel
spectrometer. The photograph shows fluorescence that is
emitted from the laser beam path inside the PEG solution
of EuCl3.
Fluorescence Enhancement of Europium Ions in a Scattering Matrix
17
Figure 5 shows the fluorescence spectra of the
solid sample (PEG 1000, Eu concentration: 10
-2
mol/l). Measurement position was 0, 1, 3, or 10 mm
apart from the input side. Fluorescence peaks are
visible at 592, 613, and 698 nm. These peaks
correspond to the electronic transitions of
5
D
0
7
F
1
,
5
D
0
7
F
2
, and
5
D
0
7
F
4
, respectively. The peak at
396 nm is attributed to the scattered pump light. The
pump light attenuates gradually as it propagates in
the sample because of scattering by the matrix and
the absorption by Eu ions. As Fig. 5(d) shows, the
pump light penetrates only ~10 mm depth from the
input end. According to the pump light attenuation,
the fluorescence peaks become lower as the position
becomes distant from the input side.
Figure 5: Fluorescence spectra of the solid sample (PEG
1000) that were measured from the sample side. The peak
at 396 nm was caused by scattering of the pump laser.
Measurement position was (a) 0, (b) 1, (c) 3, or (d) 10 mm
from the input side.
Spectral measurements for the liquid sample
(PEG 300, Eu concentration: 10
2
mol/l) were
conducted in the same manner. Figure 6 shows the
spectra that were measured at different positions.
The peak at 396 nm is weak and its height changed
little with the measurement position. This fact
indicates that the pump light propagates with a
negligible absorption loss in the transparent solution.
Accordingly, fluorescence peaks at 592, 613, and
698 nm exhibit no position dependence. Although
the fluorescence in the transparent solution is
apparently stronger than that in the translucent solid,
the actual fluorescence intensity is thought to be
stronger in the solid, since fluorescence also suffers
scattering before it is picked up by the probe fiber.
Figure 6: Fluorescence spectra of the liquid sample (PEG
300) that were measured from the sample side. The peak
at 396 nm was caused by scattering of the pump laser.
Measurement position was (a) 0, (b) 1, (c) 3, or (d) 10 mm
from the input side.
PHOTOPTICS 2017 - 5th International Conference on Photonics, Optics and Laser Technology
18
Figure 7(a) shows the position dependence of the
pump light intensity, i.e., the peak height at 396 nm,
in the solid and liquid phases of PEG. The
attenuation of the pump light is negligible in the
liquid phase. By contrast, the pump light attenuates
rapidly in the solid phase. The attenuation length is
about 1 mm; i.e., the pump light intensity becomes
1/e of the original value within 1 mm. This fact
indicates that the pump light is confined efficiently
in a microdroplet if it is composed of the solid PEG.
Figure 7(b) shows the position dependence of the
fluorescence intensity in the solid phase. All peaks
become lower with the attenuation of the pump
power. The attenuation length, however, extends to
~7 mm, which is longer than that of the pump light.
This difference is explained by the scattering of the
fluorescence rays. That is, fluorescence rays are
emitted in all directions, and some of them emerge
form a distant position due to the scattering.
Figure 7: Position dependence of the measured peak
height. (a) shows the pump light intensities in the solid or
liquid samples. (b) and (c) shows the fluorescence
intensities of the solid and liquid samples, respectively.
The numerals beside the lines denote peak wavelengths.
By contrast to the solid phase, the liquid phase
yielded no position dependence, as shown in Fig.
7(c). The intensity ratio of the 613-nm peak to the
other two peaks is higher in the liquid phase than the
solid phase. This fact is related to the width of the
613-nm peak; i.e., as Figs. 5 and 6 show, the peak is
narrower in the liquid phase than the solid phase.
The ligand field around Eu ions possibly changes
through the phase transition process (Wong et al,
2002).
Finally, fluorescence emission was measured in
the forward direction to confirm the effectiveness of
the scattering. As Fig. 7(a) shows, the pump light
excites the solid sample efficiently only within its
penetration depth. The samples were therefore put in
a thin glass cell in this experiment. As the dotted
lines in Fig. 4 illustrate, fluorescence was measured
at the exit side of the sample cell. Figures 8(a) and
8(b) show the spectra that were measured for the
solid and liquid samples of 1 mm thickness. The
fluorescence peak height was twice higher in the
solid than the liquid. The effect of the absorption
enhancement was notable in these thin samples. As
Figure 8: Fluorescence spectra that were measured in the
forward direction. The samples were PEG 1000 (solid) or
PEG 300 (liquid) that contained EuCl3. The sample
thickness was (a, b) 1, (c, d) 2, or (e, f) 5 mm.
Fluorescence Enhancement of Europium Ions in a Scattering Matrix
19
Figs. 8(c) and 8(d) show, however, the peak height
difference became less notable for the samples of 2
mm thickness. When the thickness extended to 5
mm, the fluorescence peaks shrank in the solid phase,
as shown in Fig. 8(e). By contrast, the fluorescence
peaks of the liquid grew further with the increase of
the sample thickness, as shown in Fig. 8(f).
5 DISCUSSION
In the transparent matrix, e.g., glasses or liquids,
fluorescence of lanthanide ions increases in
proportion to the sample thickness, since the pump
light reaches the exit end with a small optical
attenuation. Optical amplifiers or lasers, therefore,
use a long fiber to attain a strong fluorescence
intensity. If a high-powered laser beam is required, a
thick crystal (solution) or a long fiber is a better
choice than the scattering PEG. The advantage of the
microlaser is an efficient light emission from a small
volume. From this viewpoint, PEG is a suitable
matrix for creating a fluorescent device whose size
is smaller than 1 mm.
As mentioned earlier, the fluorescence peak at
613 nm exhibited different shapes depending on the
phase (solid or liquid). We also observed different
spectra in the aqueous and PEG solutions. This
phenomenon seems to be related to the surrounding
ligand field. Further investigation on this
phenomenon possibly leads to improvement of the
fluorescence efficiency.
In the current experiment, a laser diode of 396
nm was used as a pump light source, since other
suitable sources were not available in our laboratory.
As Fig. 3(a) shows, the transmittance of the sample
solution decreases in the ultraviolet range below 380
nm. If this transmittance decrease originates from
the absorption by the Eu ions, ultraviolet pump light
will induce more efficient fluorescence emission.
Note that this transmittance decrease was certainly
induced by addition of EuCl
3
since the pure PEG
300 was used as a blank for the transmittance evalu-
ation. It possibly happens, however, that the addition
of EuCl
3
promotes the ultraviolet absorption of PEG
[Fig. 2(b)]. Further experiments are needed to clarify
the origin of this transmittance decrease.
As Fig. 2(a) shows, the phase transition of PEGs
takes place at around room temperature. In addition,
PEGs exhibit a bistable behavior during the phase
transition process; i.e., both the solid and liquid
phases are stable at a certain temperature. This
phenomenon seems useful to create a bistable
microlaser. We are currently thinking of creating a
random droplet laser on the basis of the current
experimental results.
6 CONCLUSIONS
Polyethylene glycol is a suitable matrix for creating
a micro optical device that uses lanthanide ions as a
fluorescence emitter. It exhibits a strong scattering
in the solid phase, and extends an optical path of
pump light, leading to efficient excitation of the
fluorescent ions. This enhanced fluorescence was
demonstrated experimentally by exciting a solution
of EuCl
3
with a laser diode of 396 nm wavelength.
The fluorescence peak at 613 nm became twofold
higher in this scattering matrix than the original
liquid.
ACKNOWLEDGEMENTS
This research was supported by Japan Society for
the Promotion of Science (15K04642).
REFERENCES
Hawkins, A. R., Schmidt, H., eds., 2010. Handbook of
Microfluidics. CRC Press, Boca Raton, Florida.
Tzeng, H.-M., Wall, K. F., Long, M. B., Chang, R. K.,
1984. Laser emission from individual droplets at
wavelengths corresponding to morphology-dependent
resonances. Opt. Lett. 9(11). p. 499–501.
Saito, M., Shimatani, H., Naruhashi, H., 2008. Tunable
whispering gallery mode emission from a
microdroplet in elastomer. Opt. Express, 16(16). p.
11915–11919.
Gersborg-Hansen, M., Balslev, S., Mortensen, N. A.,
Kristensen, A., 2007. Bleaching and diffusion
dynamics in optofluidic dye lasers. Appl. Phys. Lett.
90(14). p. 143501-1–3.
Kytina, I. G., Kytin, V. G., Lips, K., 2004. High power
polymer dye laser with improved stability. Appl. Phys.
Lett. 84(24). p. 4902–4904.
Yoshioka, H., Yang, Y., Watanabe, H., Oki, Y., 2012.
Fundamental characteristics of degradation-
recoverable solid-state DFB polymer laser. Opt.
Express, 20(4). p. 4690–4696.
Barnes, M. D., Ng, K. C., Whitten, W. B., Ramsey, J. M.,
1993. Detection of single rhodamine 6G molecules in
levitated microdroplets. Anal. Chem. 65(17). p. 2360–
2365.
Shionoya, S., Yen, W. M., eds., 1999. Phosphor
Handbook. CRC Press, Boca Raton, Florida.
Garret, C. G. B., Kaiser, W., Bond, W. L., 1961.
Stimulated emission into optical whispering modes of
PHOTOPTICS 2017 - 5th International Conference on Photonics, Optics and Laser Technology
20
spheres. Phys. Rev. 124(6). p. 1807–1809.
Miura, K., Tanaka, K., Hirao, K., 1996. Laser oscillation
of a Nd
3+
-doped fluoride glass microsphere. J. Mater.
Sci. Lett. 15. p. 1854–1857.
Klitzing, W., Jahier, E., Long, R., Lissillour, F., Lefèvre-
Seguin, V., Hare, J., Raimond, J.-M., Haroche, S.,
2000. Very low threshold green lasing in microspheres
by up-conversion of IR photons. J. Opt. B: Quantum
Semiclass. Opt. 2. p. 204–206.
Lin, H.-B., Eversole, J. D., Merritt, C. D., Campillo, A. J.,
1992. Cavity-modified spontaneous-emission rates in
liquid microdroplets. Phys. Rev. A, 45(9). p. 6756–
6760.
Tagaya, A., Kobayashi, T., Nakatsuka, S., Nihei, E.,
Sasaki, K., Koike, Y., 1997. High gain and high power
organic dye-doped polymer optical fiber amplifiers:
absorption and emission cross sections and gain
characteristics. Jpn. J. Appl. Phys. 36(5A). p. 2705–
2708.
Watanabe, H., Oki, Y., Maeda, M., Omatsu, T., 2005.
Waveguide dye laser including a SiO
2
nanoparticle-
dispersed random scattering active layer. Appl. Phys.
Lett. 86(15). p. 151123-1–3.
Wiersma, D. S., Albada, M. P., Lagendijk, A., Lawandy,
N. M., Balachandran, R. M., 1995. Random Laser?
Nature, 373. p. 203–204.
Saito, M., Nishimura, Y., 2012. Bistable optical
transmission properties of polyethylene-glycol. Proc.
SPIE, 8474. p. 847411-1–12.
Saito, M., Nishimura, Y., 2016. Bistable random laser that
uses a phase transition of polyethylene glycol. Appl.
Phys. Lett. 108(13). p. 131107-1–4.
Wong, K. S., Sun, T., Liu, X.-L., Pei, J., Huang, W., 2002.
Optical properties and time-resolved photoluminescen-
ce of conjugated polymers with europium complex
side chain and an emitter. Thin Solid Films, 417. p.
85–89.
Fluorescence Enhancement of Europium Ions in a Scattering Matrix
21