8
1 Introduction
1
2
2
Bypass Channel
c
(a) Droplet enters c 1
Bypass Channel
c
1
c
c
(b) Second droplet enters c 2
Fig. 1.3 Bifurcation for passive droplet routing
and c 2 ). When the bifurcation does not contain any droplet, the larger amount of
the flow enters the successor channel c 1 due to the shorter length and, hence, lower
microfluidic resistance of this channel. 3 Hence, a single droplet will flow through
channel c 1 —the so-called default successor.
However, a droplet increases the overall resistance during its flow through a
channel, e.g., by its viscosity, size, and geometry as studied, e.g., in [6, 32, 36,
122]. This has the effect that, during the flow of a droplet through the default
successor c 1 , the larger amount of the flow now enters the non-default successor c 2 ,
i.e., the droplet temporarily blocks the default successor c 1 for following droplets.
This allows to route a closely following droplet into the non-default successor
(cf. Fig. 1.3b). Overall, exploiting this mechanism allows to route droplets even
through complex microfluidic networks in a completely passive fashion, i.e., with
no valves or switches. The underlying physics is described in detail later in Chap. 6
of Part III.
In order to support the designer in determining microfluidic networks using this
passive droplet routing mechanism, Part III of this book proposes dedicated design
methods. More precisely, the methods address the following tasks:
• Designing Application-specific Architectures: For realizing multiple different
experiments on a microfluidic network using passive droplet routing, the required
modules and how they are connected have to be determined—yielding an
architecture with multiple paths through which the droplets can flow.
In this book, an automatic design method is proposed, which generates
application-specific architectures that are dedicated to a set of experiments and
are optimized with respect to various physical constraints and design objectives.
This automatic method is introduced in Chap. 7.
• Generating Droplet Sequences: For correctly routing a droplet along a path
defined by these application-specific architectures, it has to be made sure that,
whenever a droplet is supposed to take a non-default successor at any bifurcation,
another droplet arrives before and blocks the default successor. This requires
the needed droplets to be injected in a temporally coordinated manner, i.e. a
dedicated droplet sequence is required. But determining such a droplet sequence
3 A bypass channel [20] connects the endpoints of the two successor channels. This bypass cannot
be entered by any droplet and is used to make the droplet routing only dependent on the resistances
of the successors.
1 Introduction
1
2
2
Bypass Channel
c
(a) Droplet enters c 1
Bypass Channel
c
1
c
c
(b) Second droplet enters c 2
Fig. 1.3 Bifurcation for passive droplet routing
and c 2 ). When the bifurcation does not contain any droplet, the larger amount of
the flow enters the successor channel c 1 due to the shorter length and, hence, lower
microfluidic resistance of this channel. 3 Hence, a single droplet will flow through
channel c 1 —the so-called default successor.
However, a droplet increases the overall resistance during its flow through a
channel, e.g., by its viscosity, size, and geometry as studied, e.g., in [6, 32, 36,
122]. This has the effect that, during the flow of a droplet through the default
successor c 1 , the larger amount of the flow now enters the non-default successor c 2 ,
i.e., the droplet temporarily blocks the default successor c 1 for following droplets.
This allows to route a closely following droplet into the non-default successor
(cf. Fig. 1.3b). Overall, exploiting this mechanism allows to route droplets even
through complex microfluidic networks in a completely passive fashion, i.e., with
no valves or switches. The underlying physics is described in detail later in Chap. 6
of Part III.
In order to support the designer in determining microfluidic networks using this
passive droplet routing mechanism, Part III of this book proposes dedicated design
methods. More precisely, the methods address the following tasks:
• Designing Application-specific Architectures: For realizing multiple different
experiments on a microfluidic network using passive droplet routing, the required
modules and how they are connected have to be determined—yielding an
architecture with multiple paths through which the droplets can flow.
In this book, an automatic design method is proposed, which generates
application-specific architectures that are dedicated to a set of experiments and
are optimized with respect to various physical constraints and design objectives.
This automatic method is introduced in Chap. 7.
• Generating Droplet Sequences: For correctly routing a droplet along a path
defined by these application-specific architectures, it has to be made sure that,
whenever a droplet is supposed to take a non-default successor at any bifurcation,
another droplet arrives before and blocks the default successor. This requires
the needed droplets to be injected in a temporally coordinated manner, i.e. a
dedicated droplet sequence is required. But determining such a droplet sequence
3 A bypass channel [20] connects the endpoints of the two successor channels. This bypass cannot
be entered by any droplet and is used to make the droplet routing only dependent on the resistances
of the successors.