A HIGHLY EFFICIENT THREE-DIMENSIONAL (3D) LIQUID-LIQUID
WAVEGUIDE LASER BY TWO FLOW STREAMS
Y. Yang1, C. D. Ohl2, H. S. Yoon3 and A. Q. Liu1†
1
School of Electrical & Electronic Engineering
School of Physical & Mathematical Sciences
3
School of Biological Sciences
Nanyang Technological University, SINGAPORE 639798
(†E-mail: [email protected]; Tel: +65-6790-4336; Fax: +65 6793-3318)
2
ABSTRACT
This paper reports a tunable three-dimensional (3D)
liquid-liquid waveguide laser constructed by two flow
streams in a microchannel. It comprises a 3D dye dissolved
liquid-liquid waveguide as gain medium and gold (Au)
mirrors to provide feedback. The 3D liquid-liquid
waveguide is highly efficient to overcome the optical loss as
compared with liquid-liquid three-dimensional (2D)
waveguide. The lasing wavelength is tuned by the refractive
index of the liquid core flow stream, and the output energy
can be varied by changing the volume ratio between the two
flow streams. It has wide applications in biomedical,
biological and chemical analyses in the near future.
INTRODUCTION
Optofluidics dye lasers are microfabricated dye lasers
integrated in the microfluidic channels [1]. Besides
providing the optical gain, the liquid medium in the
microchannel plays a more flexible role in varying the
optical properties such as changing the lasing wavelengths,
tuning the spatial modes, and controlling the output energy,
which do not exist in solid-state or bulk dye laser systems.
Different kinds of optofluidic dye lasers have been
developed, such as microcavity dye lasers, microdroplet dye
lasers [2], DBF dye lasers [3] and waveguide dye lasers [4].
Among of these devices, liquid waveguide dye laser is
typically attracting many interests because of the achievable
low threshold and high efficiency.
Liquid waveguides retain all the functionalities associated
with the waveguiding capability while providing a natural
synergy with lab-on-a-chip systems. 2D liquid-liquid
waveguide using three flow streams was published and is
subsequently a breakthrough research [5,6]. However, the
2D liquid-liquid waveguide cannot be regarded as a
complete 3D liquid waveguide. The absence of the liquid
cladding on the vertical direction produces serious
restriction to the optical property of the liquid waveguides.
First, the refractive index of the solid substrate is usually
much higher than the liquid claddings. Therefore, some
important optical parameters such as cut-off frequency or
band gap of the liquid-liquid waveguide are still determined
by the solid materials but not as expected by the liquid
materials. Many propagation modes should be kept in the
Pump laser
(a)
(b)
L
L'  ncore L  q
Φ1
F

2
2
Φ2
(c)
o
i
F
1
Outlet
F
Outer
inlet
Inner
inlet
2
Laser
output
Figure 1: The schematic illustration of 3D liquid
waveguide laser. (a) The work principle: the lasing
wavelength is determined by the length of the
waveguide and the refractive index of the liquid core. (b)
The output energy can be tuned by the volume ratio
between liquid core and cladding. (c) The microchip
design: 3D dye dissolved liquid-liquid waveguide as
gain medium, a pump laser to produce stimulated
emission, and two mirrors to form a cavity.
liquid-liquid waveguide will leak from the vertical direction,
making the liquid cladding in the parallel direction is not
fully functional. Second, the actual profile of the crosssection in the 2D liquid-liquid waveguide can only tuned in
one dimension with limited flexibility as compared to a pure
fluidic systems [7]. As a result, traditional liquid-liquid
waveguide is still a solid–fluid hybrid. Thus, the efficiency
of the liquid waveguide dye laser is also reduced by this
drawback.
This paper reports a tunable 3D liquid-liquid waveguide
laser constructed by two flow streams in a microchannel.
The 3D liquid-liquid waveguide is achieved based on the
centrifugal force [8, 9] and is highly efficient to overcome
the optical loss as compared with liquid-liquid 2D
waveguide.
DESIGN AND THEORY
Liquids traveling through curved microchannels
experience forces which are more complex than that in
straight microchannels [9]. Besides inertial forces acting in
the axial motion, the centrifugal force is also acted along the
radius of curvature of the conduit. The interplay between
them establishes a radial pressure gradient whose magnitude
can be sufficient to generate a transverse flow field (Fig. 1
inset), which is known as the Dean flow [10 - 12]. It can be
characterized in terms of a dimensionless Dean number
(De) that expresses the relative magnitudes of the inertial
and the centrifugal forces to the viscous forces
De=δ0.5Re
(1)
where δ = d / Ris a geometrical parameter and R is the flow
path radius of curvature. Re = Vd / ν, V is the average flow
velocity, d is the channel hydraulic diameter and v is the
kinematic viscosity of the fluid.
Figure 1 shows the schematic of the 3D liquid-liquid
F The working principle of the 3D liquid
waveguide laser.
waveguide 1laser is shown in Fig. 1 (a). High concentration
of fluorescent dyes is dissolved in the liquid core of the 3D
waveguide as the gain medium and light is confined to
enhance the efficiency. The lasing wavelength is selected by
the effective cavity length, which is finally determined by
the refractive index of the core flow stream. Fig. 1(b) shows
the flexibility of the 3D waveguide laser. The output lasing
energy can be tuned by the volume ratio between the two
flow streams. Fig. 1 (c) shows the chip design of the 3D
waveguide laser, which consists of a 3D dye dissolved
liquid-liquid waveguide, a pump laser to produce the
stimulated emission, and Au mirrors to provide feedback.
The 3D liquid waveguide consists of two flow streams in a
microchannel as shown in Fig. 1(c). The two flow streams
are pumped into the microchannel from the inner and the
outer inlets. When the two flow streams are transported in
the curved section of the microchannel, the inner flow
stream is surrounded by the outer flow stream by centrifugal
force. The refractive index of the inner flow streams is
higher than that of the outer flow stream and light is
projected into the 3D liquid waveguide from the outlet of
the microchannel. This 3D liquid waveguide is divided into
two regions. One is the reconfiguration region (I), and the
other one is the waveguide region (II). Based on the
designed microchannel geometry and the chosen liquids, the
3D liquid-liquid waveguide can be formed by tuning the
flow rates according to the Eq. (1).
In traditional liquid core-liquid cladding waveguides,
the core liquid was horizontally sandwiched by the two
cladding liquids. However, the core flow is contacted with
the channel wall in the vertical direction (top and bottom),
so that the waveguides could not be fully characterized as
pure liquid waveguides. More importantly, the refractive
index of the cladding liquid was usually lower than that of
the polydimethylsiloxane (PDMS), so that optical loss was
enhanced at the interface between the core liquid and the
PDMS channel wall. Fig. 2 shows the simulated field
intensity in the common 2D liquid-liquid waveguides [5,6]
PDMS
Liquid Cladding
Liquid
Cladding
Liquid
Core
Liquid
Cladding
Liquid
Core
PDMS
PDMS
(a)
(b)
1
(c)
Side view of 2D liquid waveguide
0
(d)
Side view of 3D liquid waveguide
Figure 2: Simulation results of a common 2D liquid
waveguide and 3D liquid waveguide. The cross-sectional
intensity filed of (a) 2D waveguide, and (b) 3D waveguide.
The intensity field along the waveguide length of (c) 2D
waveguide, and (d) 3D waveguide (d). The refractive index
of the core and the cladding is 1.410 and 1.332. The
refractive index of PDMS is 1.410.
and the 3D liquid waveguides. The refractive index of the
liquid core and the one of the liquid cladding are 1.410 and
1.332, respectively. The intrinsic loss of the 2D liquid
waveguide is shown in Fig. 2 (a) and (c), which poses a
fundamental problem in the laser application. Compared
with the 2D liquid waveguide, the 3D liquid waveguide
overcome the drawback and light is well kept in the liquid
core as shown in Fig. 2 (b) and (c) [13].
Different from the bulk liquid handing system, the
mixing, replacing and transporting of liquids in the
microchannel can be achieved in a highly flexible fashion.
This feature can make the liquid waveguide laser as a
tunable device by tuning the refractive index contrast
between the liquid core and the claddings. In liquid
waveguide lasers, the refractive index of the core flow
stream determines the optical path in the microcavity.
Different effective cavity length L can select different lasing
modes. However, in traditional liquid waveguides, the
tuning range is limited by the condition that the refractive
index of the core liquid must be higher than the PDMS to
confine light in the vertical directions. In 3D liquid
waveguides, it can keep light well even the refractive index
in smaller than PDMS as demonstrated in Fig. 2. Thus, a
more flexible waveguide laser is produced. Fig. 3 shows the
relationship the longitudinal mode spacing (∆v) and the
refractive index of the core flow stream. The liquid flowing
into the core can be tuned from DI water (n = 1.332) to
benzylalcohol (n = 1.530).
RESULTS AND DISCUSSIONS
114
∆v (GHz)
110
106
102
98
1.3
1.35 1.4
1.45
Refractive index
1.5
1.55
Figure 3: The relationship between the longitudinal
mode spacing ∆v and the refractive index of the core
flow stream.
For the experimental study, the microchannel system
was fabricated using soft photolithography processes. First,
photoresist-on-silicon master was prepared in a clean room
facility with photolithography (Micro-Chem, SU-8) using
transparent glass masks (CAD/Art Services, Inc. Poway,
CA). Then, microchannels were molded using PDMS and
sealed against flat PDMS sheets after plasma oxidation.
After fabrication, the microchannel dimensions of W = 100
µm and H = 100 µm. The flow path radius of curvature R is
2 mm.
A Fabry-Perot cavity with metallic mirrors is fabricated
on the surface of the fibers by electron beam evaporator.
The right mirror H1 is made of 30 nm Au, and have a
reflectance R1 = 70% and a transmittance T1 = 5%. The left
mirror H2 is made of 80 nm Au, and has a reflectance more
than 80% as shown in Fig. 4. The cavity length is 500 µm.
A dye solution of 2mM Rhodamine 6G in ethyl glycol (EG)
is pumped from the inner inlet as the gain medium. In the
experiment, EG (n = 1.410) and DI water (n = 1.332) are
used to form the 3D liquid-liquid waveguides and injected
from the inner and outer inlets, respectively.
Figure 5:
Cross sectional view of the confocal
microscopy results at the at the flow rates v = 1000
µl/min from the beginning to the end of the curvature (af). The positions between ethyl glycol (EG) and DI water
are transposed by Dean flow.
Figure 5 shows the experimental confocal microscopy
results of the dean flow at the cross section of the
microchannel. Besides inertial forces, the centrifugal force
acts along the conduit’s radius of curvature establishes a
radial pressure gradient whose magnitude can be sufficient
to generate a transverse flow field, which is known as the
Dean flow. The positions between EG and DI water are
transposed as shown in (a-f) at the flow rates v = 1000
µl/min. It implies that if the flow rates are sufficient, the
liquid pumped from the inner inlet can be totally surrounded
by the liquid pumped from the outer inlet.
Figure 6 shows the confocal microscopy images of the
3D liquid waveguide laser at different volume ratio between
EG and DI water. The flow rate of EG is about 400 µl/min
and the volume ratios are (a) Φ = 30, (b Φ = 40, (c) Φ = 50.
The EG flow stream moves into the middle of the straight
microchannel and is surrounded by DI water to form a
tunable 3D liquid circular waveguide, resulting different
Outlet
W = 100 µm
H1
H2
Φ=30
Outlet
R = 2 mm
Figure 4: Microscopy image of the microchannel for
liquid waveguide laser. A Fabry-Perot cavity with
metallic mirrors is fabricated on the surface of the fibers.
(a)
Φ=40
Φ=50
(b)
(c)
Figure 6: The 3D confocal microscopy images of the 3D
liquid core/liquid cladding waveguides at the volume
ratio are (a) Φ = 30, (b) Φ = 40, (c) Φ = 50. The output
lasing energy can be tuned by the volume ratio.
Intensity (a.u.)
7000
[3]
5000
[4]
3000
[5]
1000
560
580
600
620
640
Wavelength (nm)
Figure 7: The laser emission spectra of the 3D liquid
waveguide laser (red) and the emission spectra of the
RH 6G in EG without Au mirrors (blue). The inset is the
output laser spot image after the fiber (H1).
[6]
output energy. The 3D liquid waveguide structure is steady
in the straight microchannel because of the absence of the
Dean flow.
Figure 7 shows the results of the 3D liquid waveguide
laser. The pump source is a 532 nm pulse laser. If there is
no Au mirrors to support the feedback, the spontaneous
emission of the RH 6G in EG solution is shown in blue line.
The full width at half maximum (FWHM) is 40 nm.
However, if the F-P cavity is coated, the laser emission of
the 3D liquid waveguide laser is emitted around 578 nm
(red). The FWHM of the waveguide laser is only 8 nm,
which is much smaller than that waveguide source. The
inset is the output laser spot image after it coupled from the
fiber mirror H1.
[7]
[8]
[9]
[10]
[11]
CONCLUSIONS
In conclusion, we construct a 3D liquid-liquid waveguide
laser in a microchannel using only two flow streams. The
3D liquid-liquid waveguides is high efficiency to overcome
the optical loss comparison with traditional 2D liquid coreliquid cladding waveguide. The lasing wavelength is tuning
by the refractive index of liquid core flow stream, and
output energy can be chosen by changing volume ratio
between the two flow streams. It has wide applications in
biomedical, biological and chemical analyses in the near
future.
ACKNOWLEDGMENTS
This work was supported by the Environmental and
Water Industry Development Council of Singapore (Grant
No. MEWR C651/06/171).
REFERENCES
[1] Z. Y. Li, D. Psaltis, ―Optofluidic dye lasers,‖
Microfluid Nanofluid, vol. 4, pp. 145–158, 2008.
[2] A, Mekis, J.U. Noeckel, G. Chen, A.D. Stone, R.K.
[12]
[13]
Chang ―Raychaos and Q spoiling in lasing droplets,‖
Phys Rev Lett, vol. 75, pp. 2682–2685, 1995.
Z. Y. Li,, Z. Y. Zhang, T. Emery, A. Scherer, D.
Psaltis, ― Single mode optofluidic distributed feedback
dye laser,‖ Opt. Express, vol. 14, pp. 696–701, 2006.
D. V. Vezenov, B. T. Mayers, R. S. Conroy, G. M.
Whitesides, P. T. Snee, Y. Chan, D. G. Nocera, M. G.
Bawendi
―A
low-threshold
high-efficiency
microfluidic waveguide laser,‖ J Am Chem Soc, vol.
27, pp 8952–8953, 2005.
D. B. Wolfe, R. S. Conroy, P. Garstecki, B. T. Mayers,
M. A. Fischbach, K. E. Paul, M. Prentiss and G. M.
Whitesides, “Dynamic control of liquid-core/liquidcladding optical waveguides,” Proceedings of the
National Acedemy of Sciences, vol. 101, pp. 12434,
2004.
X. C. Li, J. Wu, A. Q. Liu, Z. G. Li, Y. C. Seow, H. J.
Huang, K. Xu and J. T. Lin, “A liquid waveguide
based evanescent wave sensor integrated onto a
microfluidic chip,” Applied Physics Letters, vol. 93,
pp.193901, 2008.
D. Psaltis, S. R. Quake, C. H.Yang, “Developing
optofluidic technology through the fusion of
microfluidics and optcs,” Nature, vol. 442, pp. 381,
2006.
A. P. Sudarsan and V. M. Ugaz, “Multivortex
micromixing,” Proceedings of the National Acedemy
of Sciences, vol. 103, pp. 7228-7232, 2006.
S. A. Berger and L. Talbot, “Flow in curved pipes,”
Annu. Rev. Fluid Mech., vol. 15, pp. 461-512, 1983.
X.L. Mao, J.R. Waldeisen, T.J. Huang ‗‗Microfluidic
drifting
—
implementing
three-dimensional
hydrodynamic focusing with asingle-layer planar
microfluidic device,‖ Lab Chip, vol. 7, pp. 1260–1262,
2007.
K. S, Lee, S.B. Kim, K.H Lee, H.J. Sung, S. S. Kim,
―Three-dimensional microfluidic liquid-core/liquidcladding waveguide,‖ Applied Physics Letters, vol. 97,
pp. 021109, 2010.
J. M. Lim, S. H. Kim, S. M. Yang, ―Liquid–liquid
fluorescent
waveguides
using
microfluidicdriftinginducedhydrodynamic focusing,‖ Microfluid
Nanofluid., vol. 1, pp. 1-7, 2010.
A. W. Snyder, J. D. Love, “Optical Waveguide
Theory,” Institute of Advanced Studies, Australian
National University, 1983.