Lab on a Chip Paper
the acoustic streaming-induced drag force. This is reflected
in the saturation of R for the 1 μm and 0.6 μm particles seen
for the data points obtained at the two lowest flow rates. The
smaller the particle diameter, the more influence the streaming
will have in comparison to the primary acoustic radiation
force. This is also reflected in the fact that the focusability
was generally lower for the 0.6 μm diameter particles than
for the 1 μm diameter particles. Increasing the acoustic
energy or decreasing the flow rate through the channel will
not increase the focusability, since both the acoustic streaming
and the acoustic radiation force depend linearly on the
acoustic energy density.
For one-dimensional acoustophoretic focusing in this
system, the acoustic streaming-induced drag force limits the
focusability of particles less than 1.6 μm in diameter. This is
caused by the streaming, counteracting the radiation force
in the top and bottom regions of the channel, whereby the
particles are pushed outwards from the center of the channel
instead of inwards.28 This can be avoided by using twodimensional
focusing, without compromising the channel size
or sample throughput, as presented in the following section.
Two-dimensional dual-frequency focusing in a channel
of rectangular cross-section
To enable focusing of the particles in the vertical direction
as well, a second piezo-ceramic ultrasound transducer was
added to the rectangular channel with a half-wavelength
matched to the height of the channel. This resulted in a significantly
improved focusability of the 1 μm and 0.6 μm
particles of R = 0.87 ± 0.10 and R = 0.92 ± 0.34, respectively
(Fig. 5). The voltage Upp,2 applied to the second transducer
was varied at an interval from 0 V to 4 V, while maintaining
the settings for the flow rate and voltage of the first transducer
in the corresponding one-dimensional focusing experiment
at the lowest flow rate, indicated by dashed rings in
Fig. 4(a). The relative focusability R increased steadily as the
voltage Upp,2 approached the maximum achievable in the current
system configuration. The increase in the value of R for
the small particles demonstrates the benefit of introducing a
second orthogonal acoustic standing wave. Increasing the
voltage Upp,2 above 4 V may result in higher focusability, but
it also caused the temperature of the system to rise above the
dynamic range of the temperature regulator. An improvement
in the focusability of the small particles was seen visually
when the flow rate was reduced further.
Two-dimensional single-frequency focusing in a channel
of square cross-section
A more straightforward way to generate two-dimensional
focusing in an acoustophoresis microchannel is by using a
square cross-section geometry. In this way, the same transducer
operated at a single frequency can excite both the vertical
and horizontal component of the standing waves. Even
though the strict square symmetry is broken slightly, e.g. due
to fabrication inaccuracies, the two resonances can still be
Fig. 5 Two-dimensional dual-frequency focusing in the rectangular
channel. The relative focusability R plotted against the voltage Upp,2 on
the second (5 MHz) piezo transducer focusing the particles vertically
in the rectangular channel. The voltage Upp on the first (2 MHz)
transducer and the flow rate were kept constant at the same value as
used to focus the particles giving the data points surrounded by the
dashed red rings in Fig. 4 for the 1 μm (10 μL min−1) and 0.6 μm-particles
(3 μL min−1), respectively.
excited simultaneously due to their finite width of approximately
10 kHz.38
In the square channel (230 μm × 230 μm in cross section),
which supports a two-dimensional resonance, again the large
particles with diameters of 7 μm, 5 μm, and 3 μm reached
high focusability of R = 1.01 ± 0.02, 0.94 ± 0.04, and 1.07 ± 0.004,
respectively (Fig. 4(b)). The smaller particles with diameters
of 1 μm and 0.6 μm also reached high focusability
(R = 0.95 ± 0.08 and 1.04 ± 0.10, respectively), thus demonstrating
improved focusability compared to the one-dimensional
focusing experiment. This is evident from the fact that the
normalized focusability data for all the different particles
now collapsed onto a single line (Fig. 4(b)) as compared to
the one-dimensional focusability data (Fig. 4(a)).
The square channel cross section offers a simpler system
configuration with only one frequency. In contrast, the rectangular
channel required the use of two different piezoceramic
transducers and therefore of two electronic driving
systems (signal generators and power amplifiers), adding
both cost and complexity to the system. Also, two transducers
complicate the design of the temperature controller and are
more likely to cause overheating, leading to a shift in frequency
of the acoustic resonance and therefore poor focusing
performance.
To investigate the performance of the three systems for
particles less than 0.6 μm in diameter, fluorescence
Lab This journal is © The Royal Society of Chemistry 2014 Chip, 2014, 14, 2791–2799 | 2797