Biomedical Engineering Reference
In-Depth Information
a
H(t)
t = 0
R(t)
t = t1
θ
static
θ
dynamic
t = t2
t = ∞
a(t)
R
w
R
p
P1
Flow
b
c
d
t = 0 ms
t = 233 ms
t = 367 ms
FIGURE 3.60
Surface. tension-driven. passive. micropump.. (a,. from. Erwin. Berthier. and. David. J..
Beebe,. “Flow. rate. analysis. of. a. surface. tension. driven. passive. micropump,”.
Lab Chip
. 7,. 1475-
1478,.2007;.b-d,.from.Ivar.Meyvantsson,.Jay.W..Warrick,.Steven.Hayes,.Allyson.Skoien,.and.David.
J..Beebe,.“Automated. cell.culture.in. high.density.tubeless.microluidic.device.arrays,”.
Lab Chip
.
8,.717-724,.2008..For.the.irst.paper.on.passive.pumping.see.G..M..Walker.and.D..J..Beebe,.“A.
passive.pumping.method.for.microluidic.devices,”.
Lab Chip
.2,.131-134,.2002..Reproduced.with.
permission.from.The.Royal.Society.of.Chemistry.)
high-density, large-area addressing with robotic multipipetters and even multistream laminar
lows for short periods of time, but (unlike Delamarche's scheme) it is sensitive to evaporation
and it is not possible to load an arbitrarily large number of units at once.
he obvious advantages of the two surface tension-based pumping schemes shown in this section
are their low cost, straightforward implementation, and compatibility with “tubeless microluidics”
(in which inlets are already integrated into the device in the form of reservoirs, see
Figure 3.22
),
which makes them extremely valuable for resource-poor settings. However, the pumping rate decays
with time (in microchannels; paper wicking is a form of capillarity-based pumping whose pumping
rate stays constant) and it is not amenable to computerized control (i.e., the rate cannot be changed).
3.8.4.2 Gas-Permeation Micropumps
PDMS is extremely permeable to gases—one can inject air into a dead-ended microchannel with
a syringe and the air goes into … the PDMS! his same concept, in reverse, has been applied to
pump luids: remove all the residual gas that is present in PDMS (by irst placing the device in
a vacuum jar for 15-20 minutes) and it will create a local vacuum in the microchannels. his
simple idea was irst applied by Mizuo Maeda's group from RIKEN in Japan to run biochemical
assays (
Figure 3.61
) in 2004 but it is now used by many groups elsewhere. Unfortunately, the
pump's pumping rate decays exponentially with time as the air reservoir gets gradually reilled
(operation times of >15 minutes were reported).
In 2006, Bruce Gale's group at the University of Utah designed a diferent pump based on
similar principles. Here, the idea was to inject air into microchannels at very small rates such
that the air can in turn push plugs of luids around (
Figure 3.62
). To add a high resistance
to air low, they simply placed a PDMS membrane of the right thickness (the study com-
pared membranes of 100, 45, and 25 μm thicknesses). his pumping scheme is very robust
but is incompatible with applications that are sensitive to bubbles (e.g., cells die if exposed
to a bubble or even in the proximity of an air bubble in which the CO
2
concentration is not
adequate).
