13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Evanescent Molecular Tagging Technique for Electrokinetic Effects
on Velocity Field in the Vicinity of Electrolyte-Glass Interface
Hiroki Fukumura1, Mitsuhisa Ichiyanagi2, Yohei Sato3
1: Department of System Design Engineering, Yokohama, Japan, fukumura@mh.sd.keio.ac.jp
2: Department of System Design Engineering, Yokohama, Japan, ichyanagi@mh.sd.keio.ac.jp
3: Department of System Design Engineering, Yokohama, Japan, yohei@ sd.keio.ac.jp
Abstract A novel velocity measurement technique was proposed for investigation of nanoscale flow
structure in the vicinity of electrolyte-glass interface. The technique enables the velocity measurement of the
ion existing on the order of tens of nanometers from a microchannel wall due to the caged fluorescent dye
and evanescent wave illumination. Caged fluorescent dye is initially non-fluorescent because a chemical
group is attached to quench the fluorescence of the dye. After irradiated by ultraviolet light, the chemical
group is cleaved and the dye restored the fluorescence. In order to excite the uncaged dye which exists in the
vicinity of the interface selectively, evanescent wave illumination is examined. Evanesent wave penetrates
into the electrolyte when total internal reflection occurred on the interface. The velocity of the uncaged dye
was obtained by tracing fluorescence from the uncaged fluorescent dye. A comparison between the fluid
flow measured by using evanescent wave and epi-fluorescent illumination was investigated. The molecular
tagging technique with two types of illumination was applied to a pressure-driven flow and an electroosmotic
flow (EOF) in an I-shaped microchannel with the 50 μm × 400 μm rectangular cross-section. The
microchannel was comprised of a poly(dimethlsiloxane) (PDMS) chip and a ∅50 mm diameter 170 μm
thickness borosilicate cover glass. The working fluid was prepared by dissolving CMNB-caged fluorescein
and DMNB-caged fluorescein dextran into a carbonate buffer at pH 9.3. A pulsed ultraviolet laser beam was
focused as a sheet into the microchannel to cleave the chemical group of the caged dye. The uncaged
fluorescent dye was excited by epi-fluorescent illumination using a mercury lamp or evanescent wave
illumination. For evanescent wave illumination, a 60× magnification objective lens with a numerical aperture
of 1.45 was utilized to generate evanescent wave. The measured velocity of pressure-driven flow by
evanescent wave illumination showed smaller than that by epi-fluorescent illumination, while the velocity
difference between both illumination was not observed in EOF. The present study exhibited that the
evanescent wave molecular tagging technique has the ability to investigate the nano-scale flow structure in
the vicinity of the electrolyte-glass interface.
1. Introduction
Development of micro- and nanotechnology will contribute to great advance of
microminiatureized chemical analysis system whose processes, e.g. mixing, reaction, separation and
so forth, are integrated on a small chip comprised of microchannels with a few hundred μm width.
The small liquid sample volumes in a microchannel are handled by a force on the liquid, which is
generally exerted by an external electric field. This transport mechanism is well known as
electromomotic flow (EOF), in which the motion of ions is induced. The EOF structure is governed
by the formation of positive ions formed close to the solid surface of a microchannel known as
electric double layer (EDL), because the electromigration of ions in the EDL induces the ion drag
on the rest of liquid in a microchannel. Therefore, quantitative measurements of velocity of the ions
that exist on the order of nano meters from channel wall will contribute to the accurate control of
transport phenomena affected by EDL, which is the important issue of development of the
microminiaturized and integrated system.
To date, several novel measurement techniques for flow velocity in a microchannel have been
developed. An application of particle image velocimetry (PIV) to microscale flow was proposed as
micro-PIV, which can measure the flow velocity with a microscale resolution by using submicron
-1-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
particles (Santiago et al. 1998). For a nanoscale resolution near the microchannel wall, nanoparticle image velocimetry (nano-PIV) realizes a near-wall velocity measurement for EOF (Sadr et
al. 2004). In the velocity measurement using submicron particles, the influence of the lift force
acting on a particle has been vaguely discussed.
A dye-based measurement technique, i.e., a molecular tagging technique using caged fluorescent
dye has been developed to measure the fluid velocity in both macro- and microscale. First
experimental efforts using caged fluorescent dye is conducted in macroscale flow (Lempert et al.
1995). As the application to microscale flow, a few previous works have reported the velocity
distribution of a pressure-driven flow and an EOF in capillaries and microchannels (Paul et al. 1998,
Ross et al. 2001, Sinton and Li 2003). However, it is difficult to detect the flow structure in the
vicinity of the electrolyte-glass interface, especially, electrokinetic effects on velocity field, because
their measurements remain restricted to a spatial resolution on the order of a few μm.
The present study proposes the measurement technique for velocity of ions, which enables to
detect ions selectively in the vicinity of electrolyte-glass interface, termed as evanescent wave
molecular tagging (EWMT) technique. Caged fluorescent dye is initially non-fluorescent because a
chemical group is attached to quench the fluorescence of the dye. After irradiated by ultraviolet
light, the chemical group is cleaved and the dye restored fluorescence. In order to excite the
uncaged dye which exists in the vicinity of the interface, evanescent wave illumination is employed.
The evanesent wave penetrates into the electrolyte when total internal reflection occurred at the
interface. The velocity of the uncaged dye was obtained by tracing fluorescence from the uncaged
fluorescent dye. A comparison between the fluid flow measured by using evanescent wave and epifluorescent illumination was investigated.
2. Measurement technique
2.1 Evanescent wave illumination
Total internal reflection occurs when a laser beam propagating through transparent medium of
high-refractive index reaches on an interface with one of low-refractive index with higher incidence
angle than critical angle (Axelrod et al. 1984), as illustrated in Fig 1, which defined as:
θc = sin−1
n2
,
n1
(1)
where θc is the critical angle, n1 and n2 are the refractive indices of the first medium and the second
medium, respectively. While the beam is totally internally reflected at the interface, evanescent
wave, which penetrates into the medium of lower refractive index, is emerged. The evanescent
wave intensity, Ieva, decays exponentially with the distance, z, from the interface, given by:
⎛ z ⎞
I eva ( z ) = I eva (0)exp⎜⎜ − ⎟⎟,
⎝ zp ⎠
(2)
(b) Z
(a)
Evanescent wave
Z
n2
n1
zp
Electric double
layer
θ > θc
Laser
0
I0 /e
I0
Ieva
Fig. 1. Schematic of (a) total internal reflection and (b) the evanescent wave intensity
-2-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006 #1334
where Ieva(0) is the evanescent wave intensity at the interface and zp is the characteristic penetration
depth of he evanescent wave, defined as follows:
zp =
λ
4π
n12sin2 i
θ − n22
,
(3)
where, λ is the laser beam wavelength in vacuum. Therefore, evanescent wave illumination can
excite selectively fluorescent dye at the interface region on the order of the laser beam wavelength
(from tens nanometers to hundreds nanometers).
2.2 Principle of evanescent wave molecular tagging technique using caged
fluorescent dye
Caged fluorescent dye is initially non-fluorescent because a chemical group is attached to
quench the fluorescence of the dye. After irradiated by ultraviolet light, the chemical group is
cleaved and the dye restored the fluorescence, known as an uncaging event. This is irreversible and
requires on the order of milliseconds at least (Lempert et al. 1995).
Figure 2(a) illustrates the schematic of the molecular tagging technique with epi-fluorescent
illumination. Just after the pulsed ultraviolet laser beam induces an uncaging event for the dye, the
uncaged fluorescent dye is excited by an illumination light, e.g., a mercury lamp. Evanescent wave
molecular tagging technique utilizes the evanescent wave, as exhibited in Figure 2(b). The
evanescent wave enables us to detect the fluorescence from the dye that exists in the vicinity of wall
surface. The dye is diffused by the flow convection. The instantaneous fluorescence intensity
profile is fitted to Gaussian profile (Yamamoto et al. 2002). The position of the fluorescent intensity
peak-value is obtained by Gauss fitting. The observed dye velocity is calculated from the peakvalue displacement and the interval of the images.
Pulsed ultraviolet laser beam
(b)
Pulsed ultraviolet laser beam
(a)
Wall
Wall
Z
Flow
direction
~ 1 mm
Flow
direction
X
Z
few hundreds nanometers
X
Illuminated region
Illuminated region
Fig. 2. Schematic concept of (a) epi-fluorescent illumination and (b) evanescent wave illumination.
uncaged fluorescent dye
(a) t = t
(b) t = t + Δ t
0
Flow
direction
Y
X
Displacement of uncaged
Δ x fluorescent dye
Y
X
Fig. 3. Schematic concept of the calculation of the dye velocity.
-3-
t = t
0
X
Fluorescent
intensity
Flow
direction
Fluorescent
intensity
0
Δx
t = t
0
t = t + Δt
0
X
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Table 1. Prperties of caged fluorescent dye
CMNB-caged
DMNB-caged
fluorescein
fluorescein dextran
Molecular weight
826.81
10000
Absorption wavelength* [nm]
333
353
Absorption wavelength
[nm]
494
494
Emission wavelength
[nm]
518
518
(Appproximate maximum values)
* For uncaging event
3. Experimental setup
3.1 Caged fluorescent dye
The DMNB-caged fluorescein dextran (Molecular Probes Inc.) and CMNB-caged fluorescein
(Molecular Probes Inc.) were used as the caged fluorescent dye. The properties of the dyes are
compiled in table 1. The DMNB-caged fluorescent dye is selected for the pressure-driven flow
because of redusing diffusion due to large molecular weight (MW = 8000 - 12000), however, Paul
et al. (1998) reported that the undesired separation effect was observed due to its multicomponent
nature. Thus, the CMNB-caged fluorescein as a single component dye is employed for the EOF.
The DMNB-caged fluorescein dextran and CMNB-caged fluorescein was prepared by dissolving
with a 20 μmol/l and 0.2 mmol/l concentration, respectively, in 10 mmol/l carbonate buffer at pH
9.3.
3.2 Microchannel
An I-shaped microchannel shown in Fig. 4 was comprised of a poly(dimethylsiloxane) (PDMS)
chip which was fabricated by soft lithography and a ∅50mm diameter 170 μm thickness
borosilicate cover glass plate (Matsunami glass ind., LTD. ). Cylindrical reservoirs were attached
on the PDMS at the end of the microchannel. The pressure-driven flow was generated by the water
head between two reservoirs. For the EOF, a DC electric field 34.5 V/cm was applied with platinum
electrodes immersed in the reservoirs.
3.3 Imaging apparatus
A schematic of the present measurement system is illustrated in Fig. 5. A Peltier-cooled CCD
camera (Hamamatsu Photonics, K.K., C4880-80) with a 656 × 494 array was mounted on an
inverted microscope (Nikon Corp., TE2000U). A continuous ultraviolet laser beam from an
Nd:YVO4 laser (DPSS Inc., 3510−30) operating at a wavelength of 355 nm and a power of 1 W was
used for the uncaging event. The beam, whose power was controlled to be 0.36 W by an attenuator ,
was pulsed by a shutter (Vincent Associate, VMM-T1) in 10 ms pulse duration. The pulsed beam
was focused with a plano-convex cylindrical lens (f = 50 mm) into a laser sheet in a microchannel.
In order to generate evanescent wave illumination, a 60× magnification oil-immersion objective
lens (Nikon Corp., CFI Plan Apo. TIRF) with a numerical aperture of 1.45 was employed as
illustrated in Fig. 6. For epi-fluorescent illumination, the 60× magnification or a 20× magnification
(Nikon Corp., CFI S Fluor) was employed. The measurement region is illustrated in Fig. 7. The
objective lens was focused on the XY plane of z = 0 μm and z = 25 μm for evanescent wave and epifluorescent illumination, respectively.
-4-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Mirror
(a)
Lens
Microscope stage
29 mm
400 μm
(b)
PDMS
400 μm
Z Y
Outlet
Objective lens
Mercury lamp
Filter block
Filter block
170 μm
Inlet
50 μm
X
Y
0.6× TV lens
Mirror
Microchannel
Y
Shutter
X
Attenuator
Mirror
Lens
Laser (355nm, 1W)
Beam expander
Laser (473nm, 30mW)
Cooled CCD camera
Borosilicate glass cover slip
Fig. 4. Schematic illustration of the
microchannel in (a) top view and
(b)cross-sectional view.
Fig. 5. Schematic illustration of imaging apparatus.
Laser beam (355nm)
Microchannel
Cover glass
Measurement region
by 60× objective lens
Microcscope stage
X
Y
Objective lens
(60 ×, N.A. = 1.45)
400 μm
Back focal plane
Laser beam (473nm)
Microchannel
Fig. 6. Schematic illustration of the total
internal
microscope
using
a
60×
magnification objective lens.
Measurement region
by 20× objective lens
Fig. 7. Schematic illustration of measurement
region.
A laser beam from a diode pumped solid laser (Laser-Compact Co.Ltd, LCS-DTL-364)
operating at a wavelength of 473 nm and a power of 30 mW was used to generate evanescent wave
and excite the uncaged fluorescent dye. The beam was collimated and expanded by a beam
expander, comprised of a 10× magnification objective lens, a pinhole (d = 20 μm) and a planoconvex lens (f = 500 mm). The expanded beam was focused on the back focal plane at the
peripherals region of the 60× magnification objective lens, passing through a plano-convex lens (f =
600 mm), a plane mirror and a dichroic mirror. The total internal reflection angle was adjusted to be
70° by the X-direction translation of the plane mirror, which resulted in 74 nm of the characteristic
penetration depth. The collimated laser beam from the objective lens passed through the immersion
oil and total internal reflection occurred on an interface between the borosilicate cover glass of 170
μm thickness and the buffer solution. The present system can be easily switched to epi-fluorescent
illumination using a mercury lamp by changing the optical path in the microscope.
Fluorescence from the uncaged dye was captured onto the CCD pixel sheet through the
objective lens, an emission filter transmitting wave length longer than 520 nm and 0.6×
magnification TV lens. The size of the CCD pixel sheet along with the objective lens corresponded
-5-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
to a measurement area of 180 μm × 136 μm for the 60× objective lens and 408 μm × 541 μm for the
20× objective lens in the object plane. A frame grabber (Coreco Inc., IC-PIC) captured time-series
images with 37 ms frame interval.
For pixel error reduction, an average per 5 × 5 pixels was performed. In the present experiment,
a spatial resolution of molecular tagging technique was 1.4 μm in the Y-direction for the 60×
objective lens and 4.1 μm in Y-direction for the 20× objective lens. As the captured images were
depended on the excitation light intensity and background noise of the CCD camera, a
postprocessing was conducted to correct the fluorescent intensity. The luminous intensity of the
image captured without the excitation light. The background noise was subtracted from the detected
intensity. After the subtraction, the intensity was normalized by the fluorescent intensity of images
captured under the excitation light with the long exposure time where small amount of uncaged dye
homogeneously dispersed in the aqueous solution.
4. Results and discussion
4.1 Molecular tagging technique with epi-fluorescent illumination
For the investigation of the bulk flow structure in the microchannel, dye velocities of the
pressure-driven flow and the EOF were measured by using 20× objective lens and epi-fluorescent
illumination. Figure 8 shows the time-evolution of instantaneous image of uncaged fluorescence in
the pressure-driven flow. The fluorescent intensity at y = 200 μm in Fig. 8 is plotted in Fig. 9. The
fluorescent intensity profile broadened with migration to the X-direction keeping Gaussian shape.
The position of the peak-value of fluorescent intensity profile was obtained by Gaussian fitting. The
fluorescent intensity profiles of 51 images in time series resulted in 50 velocities at each Y-direction
positions. The RMS value of the velocity at y = 200 μm was estimated to be ± 15.1 μm/s (16.8 % of
the velocity). This error could be derived from Gaussian fitting error caused by undesired
fluorescent intensity profiles and the fluctuation of the CCD camera frame interval whose RMS
value was estimated to be ± 0.52 ms (1.42% of the averaged frame interval). The averaged velocity
at each Y-position is plotted in Fig. 10. The trapezoidal shape of the velocity profile agrees with the
theoretical result (Brody et al. 1996). This result shows that the velocity profile has uniform
velocity region at the center of the microchannel (at y = 100 - 300 μm) in terms of the Y-direction.
In the EOF, the time-evolution of instantaneous image of uncaged fluorescence is shown in Fig. 11.
The fluorescent intensity at y = 200 μm and the calculated velocity profile from 51 images are
shown in Fig. 12 and Fig. 13, respectively. The RMS value of the velocity at y = 200 μm was
estimated to be ± 6.46 μm/s (4.67 % of the velocity). The velocity profile of the EOF in Fig. 13 has
a flow pattern that superimposed a reverse-trapezoidal and a plug flow pattern. When the external
electric field is applied, the electromigration of positive ions in the EDL exerted the bulk solution
flow in the microchannel because the flow is governed by viscosity. It is shown theoretically that
the flow shows a plug flow pattern (Probstein, 1994) in the EOF. Taylor and Yeung (1993) and Sato
et al. (2002) reported that a reverse-parabolic flow pattern was observed at a depth-wise midpoint of
a microchannel. They measured the eloctroosmotic flow velocity with the pressure-driven flow
which was induced by elevating the water head in the cathode-side inlet due to the electroosmotic
flow. The result in the present experiment has good agreement with the flow pattern reported in the
previous experimental works.
-6-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
541 μm
(a)
(b)
(c)
(d)
408 μm
Y
X
2.5
100
0 ms
Col
8 vs Col 9
370
msCol 10
Col
8 vs
Col
8 vs
740
msCol 11
Col
8 vsms
Col 12
1110
2.0
1.5
Dye velocity [μm/s]
Fluorescence intensity [-]
Fig. 8. Time-evolution of instantaneous image of uncaged fluorescence in the pressure-driven
flow at (a) 0 ms, (b) 370 ms, (c) 740 ms and (d) 1110 ms.
1.0
0.5
0.0
0
80
60
40
20
0
100 200 300 400 500
X-direction position [μm]
Fig. 9. Fluorescence intensity profiles at y = 200
μm in the pressure-driven flow with epifluorescent illumination.
0
100
200
300
400
Y-direction position [μm]
Fig. 10. The calculated velocity profile in the
pressure-driven flow.
541 μm
(a)
(b)
(c)
(d)
408 μm
Y
X
Fig. 11. Time-evolution of instantaneous image of uncaged fluorescence in the EOF at (a) 0 ms,
(b) 370 ms, (c) 740 ms and (d) 1110 ms.
-7-
3.0
Col
6 vs Col 7
0 ms
Col
6 vs
370
msCol 8
Col
6 vs
740
msCol 9
Col
6 vsms
Col 10
1110
2.5
2.0
Dye velocity [μm/s]
Fluorescence intensity [-]
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
1.5
1.0
0.5
0.0
0
100 200 300 400 500
X-direction position [μm]
Fig. 12. Fluorescence intensity profiles at y =
200 μm in the EOF with epi-fluorescent
illumination.
160
140
120
100
80
60
40
20
0
0
100
200
300
400
Y-direction position [μm]
Fig. 13. The calculated velocity profile in the
EOF.
4.2 Molecular tagging technique with evanescent wave illumination
Figures 14 and 15 exhibit the time-evolution of instantaneous image of uncaged fluorescence in
the pressure-driven flow with epi-fluorescent and evanescent wave illumination captured by the 60×
objective lens, while the fluorescence intensity profiles at y = 200 μm of Fig. 15 is plotted in Fig. 16.
The velocity profile calculated from 21 images is depicted in Fig. 17. The RMS value of the
velocity at y = 200 μm was estimated to be ± 4.52 (6.14 % of the velocity) μm/s for epi-fluorescent
illumination and ± 9.89 μm/s (37.4 % of the velocity) for evanescent wave illumination,
respectively. The both of the velocity profiles have a flat pattern because the measurement region is
the center of the microchannel (y = 136 – 272 μm) where the uniform velocity patterns at y = 100 –
300 μm present in the both flows as verified in the section 4.1. It is obvious that the velocity
obtained by evanescent wave illumination was much smaller than that by epi-fluorescent
illumination. The velocity profile in the Z-direction in the present microchannel is predicted
theoretically to form a parabolic pattern (Brody et al. 1996) and the velocity near z = 0 μm in the
pressure driven flow is expected to be much decreasing compared to the bulk velocity because of
the non-slip condition at wall; Therefore, the result indicates the present technique can detect the
velocity in the near wall region.
Measurements of the dye velocity in the EOF were carried out in order to investigate the
electrokinetic effect by the EDL on the velocity field. Figures 18 and 19 show the time-evolution of
instantaneous image of uncaged fluorescence in the EOF with epi-fluorescent illumination and
evanescent wave illumination captured by the 60× objective lens. Figure 20 exhibits the
fluorescence intensity profiles at y = 200 μm in Fig. 19. The calculated velocity profile was shown
in Fig. 20. The RMS value of the velocity at y = 200 μm was estimated to be ± 7.56 μm/s (7.11 %
of the velocity) for epi-fluorescent illumination and ± 8.60 μm/s (6.95 % of the velocity) for
evanescent wave illumination, respectively. It is expected that the large velocity gradient in the Zdirection exists in the EDL to satisfy the non-slip condition at the wall; however, the present
experiment cannot detect an evidence of the velocity gradient, i.e., the significant differences of the
dye velocity between the two illumination methods. It is considered that the observed velocity by
EWMT technique is affected by the magnitude between thickness of the EDL in the buffer solution
and the characteristic penetration depth of the evanescent wave. The Debye length well known as
EDL thickness is given by:
-8-
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
1
⎛ εRT ⎞ 2
λD = ⎜⎜ 2 2 ⎟⎟ ,
⎝ 2 F z c0 ⎠
(4)
where ε is the permittivity of liquid, R is the gas constant, T is temperature, F is the Faraday
constant, z is the charge number of ion and c0 is the bulk ion concentration. The Debye length in the
10 mmol/l carbonate buffer is estimated as to be 3 nm. This value is much smaller than the
characteristic penetration depth of the evanescent wave in the present experiment, i.e., 74 nm.
Therefore, the difference was not observed between epi-fluorescent and evanescent wave
illumination in the EOF since this estimation indicates that the most of illuminated uncaged dye was
expected to exist outside of the EDL. In the future work, further investigation of the flow structure
in the vicinity of the interface will be carried out by adjusting the ion concentration of the buffer
solution to set the EDL thickness to be the equivalent to the characteristic penetration depth of
evanescent wave illumination.
180 μm
(a)
(b)
(c)
(d)
136 μm
Y
X
Fig. 14. Fluorescence images of the uncaged dye in the pressure-driven flow captured by the ×60
objective lens with epi-fluorescent illumination at (a) 0 ms, (b) 185 ms, (c) 370 ms and (d) 555
ms.
180 μm
(a)
(b)
(c)
(d)
136 μm
Y
X
Fig. 15. Fluorescence images of the uncaged dye in the pressure-driven flow captured by the ×60
objective lens with evanescent wave illumination at (a) 0 ms, (b) 370 ms, (c) 740 ms and (d)
1110 ms.
-9-
3.5
120
0 ms
Col370
8 vs ms
Col 10
Col740
8 vs ms
Col 11
ms12
Col1110
8 vs Col
3.0
2.5
Dye velocity [μm/s]
Fluorescence intensity [-]
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
2.0
1.5
1.0
0.5
0.0
0 20 40 60 80 100 120 140 160 180
X-direction position [μm]
Fig. 16. Fluorescence intensity profiles at y =
200 μm in the pressure-driven flow with
evanescent wave illumination.
100
Epi-fluorescent illumination
Evanescent wave illumination
80
60
40
20
0
140 160 180 200 220 240 260
Y-direction position [μm]
Fig. 17. Calculated velocity profile in the
pressure-driven flow.
180 μm
(a)
(b)
(c)
(d)
136 μm
Y
X
Fig. 18. Fluorescence images of the uncaged dye in the EOF captured by the ×60 objective lens
with epi-fluorescent illumination at (a) 0 ms, (b) 185 ms, (c) 370 ms and (d) 555 ms.
180 μm
(a)
(b)
(c)
(d)
136 μm
Y
X
Fig. 19. Fluorescence images of the uncaged dye in the EOF captured by the ×60 objective lens
with evanescent wave illumination at (a) 0 ms, (b) 185 ms, (c) 370 ms and (d) 555 ms.
- 10 -
8
140
Col
5 vs Col 6
0 ms
1855 ms
Col
vs Col 7
Col
vs Col 8
3705 ms
Col
vs Col 9
5555 ms
6
Dye velocity [μm/s]
Fluorescence intensity [-]
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
4
2
0
120
100
80
60
40
20
Epi-fluorescent illumination
Evanescent wave illumination
0
0 20 40 60 80 100 120 140 160 180
X-direction position [μm]
Fig. 20. Fluorescence intensity profiles at y =
200 μm in the EOF with evanescent wave
illumination.
140 160 180 200 220 240 260
Y-direction position [μm]
Fig. 21. Calculated velocity profile in the
EOF.
5. Conclusion
A velocity measurement technique in the vicinity of the electrolyte-glass interface, utilizing the
caged fluorescent dye and evanescent wave illumination, was developed in order to investigate flow
structure affected by the electric double layer. The important conclusions in the present study are
summarized bellow.
(1) The velocity profiles in a microchannel were obtained by a molecular tagging techniques using
epi-fluorescent and evanescent wave illumination. The techniques were applied to the pressuredriven flow and the EOF. The spatial resolution of the measurement was 1.4 μm in the Y-direction
for the 60× objective lens and 4.1 μm in the Y-direction for the 20× objective lens, respectively.
(2) The evanescent wave molecular tagging technique was performed compared to the technique
using epi-fluorescent illumination. The velocity near the wall region was decreasing compared to
the bulk flow in the case of the pressure-driven flow. In the EOF, the difference of the velocity
between the two illumination methods was not observed, because the magnitude between the
thickness of the EDL and the characteristic penetration depth of the evanescent wave affects the
velocity field measured by the EWMT technique.
Acknowledgement
This work was subsidized by the Grant-in-Aid for Young Scientists (A) No.17686017 of Ministry
of Education, Culture, Sports, Science and Technology and TEPCO Research Foundation.
Reference
Axelrod, D., Burghardt, T.P. and Thompson, N.L. (1984). “Total Internal Reflection
Fluorescence,” Annual review of biophysics and bioengineering, 13, pp. 247−268.
Brody, J.P., Yager, P., Goldstein, R.E. and Austin, R.H. (1996) “Biotechnology at Low
- 11 -
13th Int Symp on Applications of Laser Techniques to Fluid Mechanics
Lisbon, Portugal, 26-29 June, 2006
Reynolds Numbers,” Biophysical Journal, 71, pp. 3430−3441.
Lempert, W.R., Magee, K., Ronney, P., Gee, K.R. and Haugland, R.P. (1995). “Flow tagging
velocimetry in incompressible flow using photo-activated nonintrusive tracking of molecular
motion (PHANTOMM),” Experiments in Fluids, 18, pp. 249−257.
Paul, P.H., Garfuilo, M.G. and Rakestraw, D.J. (1998). “Imaging of pressure and
electrokinetically driven flows through open capillaries,” Analytical Chemistry, 70, pp. 2459−2467.
Probstein, R.F. (1994). Physicochemical Hydrodynamics: An Introduction Second Edition, A
wiley-Interscience Publication.
Ross, D., Johnson, T.J. and Locascio, L.E. (2001). “Imaging of electroosmotic flow in Plastic
microchannels,” Analytical Chemistry, 73, pp. 2509−2515.
Sadr, R., Yoda, M., Zheng, Z. and Conlisk, A.T. (2004). “An experimental study of electroosmotic flow in rectangular microchannels,” Journal of Fluid Mechanics, 506, pp. 357−367.
Santiago, J.G., Wereley, S.T., Meinhart, C.D., Beebe, D.J. and Adrian, R.J. (1998). “A particle
image velocimetry system for microfluidics,” Experimets in Fluids, 25, pp. 316−319.
Sato, Y., Hishida, K. and Maeda, M. (2002) “Quantitative Measurement and Control of
Electrokinetically Driven Flow in Microspace,” Micro Total Analysis Systems 2002, 1, pp.
512−514.
Sinton, D.A. and Li, D. (2003). ”Microfluidic velocimetry with near-wall resolution,” International
Journal of Thermal Sciences, 42, pp. 847−855.
Taylor, A. and Yeung, ES. (1993). “Imaging of Hydrodynamic and Electrokinetic Flow Profiles in
Capillaries,” Analytical Chemistry, 65, pp. 2928−2932.
Yamamoto, T., Inaba, S., Sato, Y., Hishida, K., and Maeda, M. (2002). “Measurements in
microchannel by laser induced molecular tagging and micro-PIV,” 11th International Symposium
on Application of Laser Techniques to Fluid Mechanics: Web-site.
- 12 -