2015 10th International Conference on Design & Technology of Integrated Systems in Nanoscale Era

(DTIS)

3D Integration in Biochips: new proposed architectures for 3D applications

! in ATDA based Digital Microfluidic Biochips
1
Pranab Roy, 2Pampa Howladar, 3Raja Dastidar, 4Hafizur Rahaman, 5Parthasarathi Dasgupta
1,2,3,4
School of VLSI Technology, Indian Institute of Engineering Science and Technology, Shibpur, India
5
Indian Institute of management Calcutta, India
Email: 1ronmarine14@[Link],[Link]@[Link],3rajadastidar2012@[Link],4rahaman_h@[Link],5partha@[Link]

Abstract— Digital microfluidic biochip (DMFB), a latest through the electrodes [10]. Hence, a programmed sequence of the
invention in lab-on-a-chip devices integrates electronics with biology electric potential at the consecutive electrodes creates a strength
for development of customized miniature devices intended for new disparity of this wetting force. This variation results in movement of
application areas namely clinical diagnostics and detection, DNA droplets from one electrode contact to the adjacent one.
analysis, point-of-care applications and so on. All-Terrain droplet Typically, a DMFB is comprised of 2D open array of electrodes
actuation (ATDA) based biochips have (has) recently been emerged covered with a hydrophobic insulation layer and peripheral devices
as the new variant in DMFB devices for application in 3D domain. In namely optical detection sites, dispensing ports, storage and heater
this work, we explored the possibility of design optimization locations and so on. In a EWOD based DMFB architecture, a unit
followed by development of design automation techniques cell in the array is comprised of a pair of electrodes that acts as two
specifically applicable in 3D biochips. We consider the advantages parallel plates. The bottom plate is formed with a patterned array of
and issues involved in development and application of 3D based electrodes, and the top plate is coated with a continuous ground
design. Accordingly we demonstrated few prescheduled bioassay electrode. A droplet rests on a hydrophobic surface over an electrode
execution in both 2D and 3D domain and the following enhancement and is sandwiched between two parallel plates. The sample droplet
in the route performance. carried within a filler medium, such as silicone oil and moves within
Keywords— Digital Microfluidics, ATDA Biochips, 3D Actuation, it. Electrical actuation is carried out through control pins that connect
Biochip architecture, Placement and Route performance, Transition the electrodes addressing either individual or group of unit cells in a
ladders or vias. 2D planar array (Fig. 1). Specifically, the operations namely merging,
I. INTRODUCTION mixing and splitting are executed in the form of microfluidic modules
The current advancement in microfabrication and microfluidic (i.e., virtual devices), that are simply obtained by grouping the
technology has inspired the emergence of a new class of lab-on chip adjacent cells [Link] operations performed by actuating control
device termed as microfluidic biochip. This new class of devices voltages through the electrodes are considered as reconfigurable due
offers a convenient alternative platform for conventional laboratory to their flexibility in resource utilization (electrodes involved) and in
processes. They facilitate on-chip integration of basic microfluidic execution time.
functions necessary for biochemical analysis namely dispensing, B. Droplet actuation in 3D plane and related works
transportation, merging and mixing [1]. Microfluidic biochips found In the present design of DMFB, the droplet motions are restricted
numerous applications in the areas of clinical diagnostics (specifically into one plane comprised of 2D array of electrodes. This in turn
for point-of-care diagnosis of diseases), drug discovery, DNA reduces the device capacity for a given device footprint and restricts
analysis, immunoassays and environmental monitoring. the integration of different physical or chemical environments that on
Earlier, the generation of microfluidic biochips was based on a single device. In order to overcome this limitation, some initial
continuous fluid flow through permanently etched channels. Liquid techniques have been devised towards the actuation of 3D droplets
flow was actuated by external pressure sources, integrated [11, 12]. In [11], the alignment of two flexible electrode plates is to
micropumps or by electrokinetic mechanisms such as electro- be restricted for the definitive actuation methods employed for 3D
osmosis. However, their integrated micro-structures make actuation. This in turn restricts the range of architecture that can be
continuous-flow biochips unsuitable for more complex applications, designed for bioassay application. In [12], the droplet motion is
requiring complicated fluid manipulation [2]. restricted to vertical jumps between two horizontal platforms.
A. Digital Microfluidic Biochips However, the contribution reported in [13-14] provides a major
A new generation of device involving manipulation of discrete breakthrough in the droplet actuation in 3D plane. The method was
liquid droplets has emerged only in the last decade. Digital named as All Terrain Droplet Actuation(ATDA), where the discrete
microfluidic biochips (DMFB) are capable of manipulating droplets droplets are manipulated on inclined, declined, vertical, twisted and
in nanolitre volumes on a 2D array of individually controllable upside down architecture [13]. It has been shown that these new
electrodes. The droplet manipulation can be achieved through any of geometries enable flexible, straightforward integration of distinct
the mechanisms, namely electro-wetting [3]–[5], di-electrophoresis physicochemical environment on monolithic devices [14].
[6], thermo-capillary transport [7] and surface acoustic wave ATDA device is based on flexible, copper-clad polyimide substrates.
transport [8]. In the DMFB architecture, the unit liquid volume is They are formed by patterned arrays of the copper electrodes and
determined by the geometry of the system (fluid quantization). The coating them with poly-dimethylsiloxane (PDMS) and Teflon-AF. In
use of unit volume allows the bioassay execution to be divided into a all geometries, the droplet movement is facile and fast, with no
set of basic microfluidic functions. significant differences relative to conventional, planar microfluidics
One of the widely used methods for droplet actuation in DMFB is [14]. In all geometries, the droplet movement in 3D plane is found to
Electrowetting on dielectrics (EWOD). The mechanism of electro be facile and fast, with no significant difference with conventional,
wetting is based on control of surface tension of a liquid-solid planar microfluidics. ATDA is demonstrated to be capable of
interface through the application of the electrical potential at the actuating a wide variety of liquids, including aqueous buffer, solution
interface. At the nanolitre scale, the forces of interfacial tension of proteins and DNA, and undiluted bovine serum [14].
between three phases dominate a droplet’s hydrodynamic behavior
[9]. So whenever a droplet is kept in contact with a solid electrode, a
wetting force can arise upon application of an electric field through
these electrodes. This wetting force acting on the triphase contact line
can be altered by applying varied electrical potential to the droplet

978-1-4799-1999-4/15/$31.00 ©2015 IEEE

 !

inclination. In [14] the experiments have been carried out using

droplets of the different volumes and the maximum angle has been
! recorded at which the actuation is feasible. The potential required for
droplet actuation by ATDA has been found to be in the range of 500
VRMS on horizontal surfaces and 700 VRMS for climbing vertical
walls [13]. Higher actuation potential was required to overcome the
additional resistance resulting from droplet weight. The experiments
showed close conformance with the prediction.
The rest of the paper is organized as follows. Section II states the
Fig.1. Schematic diagram of DMFB using 2D array of electrodes major motivation for design optimization using ATDA based 3D
design of biochips. Section III represents the advantages of using 3D
geometry for design of ATDA based biochips. Section IV describes
three different 3D architectures and the point to point route distances
between two planes. Section V demonstrates the application
advantages for specified bioassays on a prescheduled layout in 3D
based design in comparison to 2D counterparts. Finally, the
a) Temperature sensor application
concluding remarks with future scopes for further enhancements are
discussed in section VI.

II. MOTIVATION FOR DESIGN OPTIMIZATION IN 3D PLATFORM

Early research on CAD for DMFBs has been focused on device-level
b) Oxygen sensor application physical modeling of single components. In modern days, a top
Fig. 2. ATDA manipulation of droplets -Sequences of frames from movies down system design and test methodology have been used for the
(top to bottom) depicting multi-environmental ATDA devices. (a) Droplet design of DMFBs. This methodology speeds up the design cycle,
containing thermoresponsive dye (black < 29◦C < white) is driven up and reduces human effort, and increases dependability [19]. A behavioral
down an aircooled structure on a device positioned on a hot-plate held at 40◦C.
(b) Droplet containing methylene blue dye and glucose is driven from an
model for a biomedical assay is obtained from the bioassay protocol.
aerobic chamber to an anaerobic chamber (submerged in silicone oil). In Next, architectural-level synthesis is applied to generate a
anaerobic conditions, the dye is reduced and the droplet becomes colorless. macroscopic structure of the biochip. The macroscopic model
t (courtesy Lab-on-chip Journal,issue5,2008, page. 673)[14] provides an assignment of assay functions to biochip resources as
In the EWOD based planar devices, the two forces regulate droplet well as a mapping of assay functions to time-steps based in part on
motion, namely, an actuation force FEWOD, resulting from difference the dependencies between them. Based on the scheduling results, the
in the contact angles caused by the strength disparity in wetting force placement algorithms are applied to generate a suitable layout for
on either side of a droplet, and a resistive friction force, Ff, resulting allocating droplets and modules [20] followed by necessary
from contact angle hysteresis [15] and viscous friction [16, 17]. compaction. Finally, the droplet-routing algorithms are used to
However, a major force to be considered for ATDA based actuation formulate droplet behaviors in a reconfigurable manner for
is the force due to gravity Fg that resists the droplet motion in concurrent time multiplexed routing within a 2D planar array of
inclined plane. The Fg is a linear function of droplet volume. Thus, it DMFB.
limits the maximum droplet size that can be moved up a plane for a In recent years, several automated design tools are proposed for
given angle of inclination. The droplet movement by ATDA in an design of DMFBs in 2D plane. However, no such attempt has been
inclined plane can be modelled through the estimation of the three reported for design optimization of ATDA based devices operating
forces FEWOD , Ff and Fg. on 3D plane. The design optimization techniques are predicted to be
FEWOD has been estimated in [14] using the Carre and Shanahan different due to variation in transit path as well as architecture for 3D
model [18] for calculating capillary forces resulting from contact transition. In this paper, we attempt to describe three different
angle hysteresis. architectures for ATDA based design in 3D plane. We further
demonstrate few layout based routing applications to show
FEWOD ≈ F CAPILLARY ≅ 1 πr0 γ LV (cos θ adv − θ rec ) advantages of routing in 3D based design over its 2D counterparts.
2 III. . 3D DESIGN ADVANTAGES
where the r0 is the radius of the droplet base, γLV is the liquid–vapour ATDA based Biochips use multiple layers (at least two) for droplet
surface tension, and θadv and θrec are the advancing and receding actuation and transportation between two planes. Avoidance of cross
contact angles, respectively. Frictional resistance to droplet motion - contamination and zone based reduction in routing paths has a
Ff has been calculated from the following relation. number of geometrical advantages of using multiple planes in DMFB
Ff = C 6 πμ UR [16][17] architecture.
θ
Where, μ is the liquid viscosity, U is the droplet velocity, R is the A. K5 and K3,3 connectivity and crossover issues
radius of droplet base, θ is the liquid static contact angle in radians, For a K5 graph (a fully connected graph with 5 vertices) or a K3,3
and C is a numerical constant (C ≅ 10). graph (a complete bipartite graph with 3 vertices on each partition)–
Ff is estimated using a standard curve of the resistive force acting there remain at least two edges that intersect each other.
on static droplets that prevents them from sliding down a tilted plane. According to Kuratowski’s theorem [19] a graph is planar if and
Droplets of the size range tested in the study of [14], the measured only if it contains no sub graph that is isomorphic to or is a
static friction force is found to be more than 30 times larger than subdivision of K5 or K3,3 graph. However, it can be shown that for a
viscous friction forces acting on moving droplets. Hence, viscous graph with K5 or K3,3 configurations, it is possible to place them in
friction was neglected in formulation of the ATDA based model. two parallel planes with no edges intersecting one another. Figure 3
Finally, the third force affecting droplet movement in ATDA, Fg, was and 4 represent two plane based K5 and K3,3 connectivity with all
calculated simply as the projection of droplet weight down the vertices being located at either of the two planes.
inclined plane. With the estimation of the three forces described In figure 3 a) K5 connectivity is represented with all 5 vertices being
above, the feasibility of droplet actuation by ATDA is predicted in located in a single plane. Here it has been found that at least one
[14]. It has been predicted through the estimated model that droplets crossover between any two edges is inevitable. However using two
with volumes less than ~7.3 μL can be moved up a 900 incline, and planes the connectivity involves inter planar edges and coplanar
droplets with higher volumes can only be driven up on a reduced edges, As shown in figure 3b) it is possible to achieve K5 connectivity
 !

without any crossover using two planes.

a) b)
a)K5 connectivity – at least two edges are intersecting
b)K5 connectivity – use of two planes results in no crossover a) Design with two parallel ladders in y plane with separate ITZ
Figure 3. K5 connectivity issues – resolved using two planes

a) b)
b) Design with two opposite ladders in x plane with central ITZ
a) K3,3 connectivity – at least two edges are intersecting on a single plane
b) K3,3 connectivity – use of two planes results in inter planer edges with no
crossovers
Figure 4. K3,3 connectivity issues – resolved using two planes
In figure 4 a) K3,3 connectivity is represented with all 6 vertices
being located in a single plane. Here again it is observed that at least
one crossover between any two edges is inevitable. However using
two planes the connectivity involves inter planar edges only. As
shown in figure 4b) it is possible to achieve K3,3 connectivity without
any crossover using two planes. c) Design with 4 ladders in x and y plane with central ITZ
It has also been observed that if distance measured between two Figure 5. Three different configurations used for ATDA based DMFBs
points in a single plane is compared with two points in different .
planes with similar dimensions – sometimes coplanar distances are Figure 6b represents the 3D representation of the planar array
found to be higher than that measured between two separate [Link] distributed into two floors. Each floor has a dimension of 6a x 6a and
this paper we have proposed three specific configurations namely are distributed into 9 zones (each with dimensions of 2a x 2a).The
architecture 1 ,architecture 2 and architecture 3 and measured the zones are named in a similar manner as shown in figure 6 a. The
corresponding 2D and 3D distances between every two possible pair zones B4 and A4 are considered as ITZ. Two parallel transition
of points located in two different segments. The comparison of ladders with opposite inclination are arranged between these zones
distances for all three configurations is displayed in later sections. for connecting the two floors to enable droplet transportation between
IV. PROPOSED ARCHITECTURE FOR ATDA BASED DESIGN the two floors.(refer to figure 6 a). Each ladder has ground coverage
In this section we propose three different configurations for ATDA of 2a and actual inclination length h = (2a+1).
based biochips. All configurations are designed with two layers Procedure for distance calculation between pair of points each lying
namely upper and lower floors .Each layer is divided into a set of between either segments or floors are as follows:
zones. In the lower layer for configuration 1 and in configuration 2 1. Assign corner points for each zone to formulate the
one centralized zone or two parallel zones termed as interlayer boundaries
transition zone (ITZ) is used to carry out the interlayer transition. The 2. for 2D directly calculate Manhattan distance between one
ITZ has two specified parts - namely 1) Transfer corridor – this point in segment 1 and another point in segment 2.
represents the corridor for droplet movement along the lower plane at Continue selection of points between each zone and
the foot of the ladders. 2) Transition ladder – provides the compute the overall sum.
connectivity between two layers. The ladders are assumed to be 3. For 3D design assign entry and exit points for each ladder
inclined at an angle θ with the lower floor. This implies a ladder with and assign zones accordingly to each ladder based on their
length of y cells corresponds to ground width of x cells in the lower proximity with the corresponding ladder.
plane. i.e. cos θ = x /y . 4. Calculate distance of any one cell in floor1 to the
Figure 5 a) , b) and c)show three different configurations that vary corresponding entry.
in the orientation of ladders in the ITZ. 5. Calculate distance from exit of the ladder to the target
location
A. Zone based and overall distance calculation for 2D and 6. Add up the two calculated distances stated above with that
3D based DMFB Design of the ladder height. Thereby sum up all the distances
computed between each cell in floor 1 and that in floor 2.
A.1. Parallel ladder based architecture
7. Compute zone wise distances for both 2D and 3D and
We consider a 2D planar array with dimension 12a x 6a and partition
finally obtain the ratio between overall 2D and 3D
it into 2 parts (6a x 6a) viz. AZone and BZone , each providing an
distances for given parameter value ‘a’.
equivalent representation of the 2 floors in a 3D based design. Here
‘a’ is considered to be some base value – which can be varied to
modify the dimensions symmetrically( figure 6). Each partition is
further divided into 9 zones namely B0, B1 ….B8 for 1st floor
representation and A0,A1,…A8 for 2nd floor representation. Here
each zone is considered to be of equal dimension (2a x 2a).Zone B4
and A4 are considered as ITZ . B4 has been used for placement of a) The 2D layout with two partition each representing one floor in 3D for
transition ladders parallel ladder architecture
 !

2. For 2D calculate manhattan distance between one point in

segment 1 and another point in segment 2 in the same
! manner as used for parallel ladder architecture.
3. Finally compute the overall sum of distances in 2D.
4. For 3D design assign entry point for each zone in floor 1.
5. Similarly assign exit point for each zone in floor 2.
6. Based on the entry and exit points assignment for each
b) The 3D layout with parallel ladders and the distributed zones zone the ladders are assigned accordingly.
7. Calculate the distance from any cell in a given zone to the
assigned entry point .
8. Calculate the distance of the destination cell for the
stipulated zone to the assigned exit point.
9. Obtain the landing location in the 2nd floor ITZ on climbing
the ladder from the entry point. Calculate the distance
between the landing location and the respective exit point
for the destination zone.
10. Add the ladder height together with all the three distances
c) Detailed cross sectional view of parallel ladder orientation
Figure 6. Zone based 2D and 3D configuration (parallel ladder based design) calculated in steps 7,8 and 9 . Thereby total distance
A.2 Opposite ladder based architecture between two cells at different floors(segments) are
In this architecture we placed two ladders along x plane with computed.
inclination along y plane as shown in figure 7 .Each ladder has a 11. Then sum up all the distances computed between each cell
height of h < a. in floor 1 and that in floor 2.
Similar to the previous architecture we used a 2D planar array with 12. Also compute zone wise distances for both 2D and 3Dand
dimension 12a x 6a and partition it into 2 parts (6a x 6a) viz. AZone finally obtain the ratio between overall 2D and 3D
and BZone, each providing an equivalent representation of the 2 distances for given parameter value ‘a’.
floors in a 3D based design. A.3 Four ladder based architecture
Each partition is further divided into a set of 11 zones – but the In this architecture we place two ladders along x plane with
dimensions for all zones are not the same. The zone dimensions are inclination in y plane and another two ladders along y plane with
determined for closest access to the respective entry and exit points to inclination in x plane.
the ladders in ITZ as shown in figure 6 b. for the 3D based design. Similar to the previous architectures we used a 2D planar array with
The zone dimensions are assigned as dimension 12a x 6a and partition it into 2 parts (6a x 6a) viz. AZone
–for BZone representing the 1st floor equivalent segment. and BZone, each providing an equivalent representation of the 2
Dimensions for B0, B1, B2, B5, B8, B9, B10 are planned as floors in a 3D based design.
– 2a x 2a Each partition is further divided into a set of 13 zones – but the
and for B3,B4,B6,B7 –a x 2a dimensions for all zones are different. The zone dimensions are
Similarly for AZone represents 2nd floor equivalent segment. determined through closest access to the respective entry and exit
Dimensions for A0, A1, A2, A5, A8, A9, A10 are planned as points to the ladders in ITZ as shown in figure 7 b. for the 3D based
– 2a x 2a and for A3, A4,A6,A7 – 2a x a design.
The zone dimensions are assigned as
–for BZone representing the 1st floor equivalent segment.
Dimensions for B0, B3, B6, B9, B12 are planned as – 2a x 2a
for B1,B2, B10,B11 – 2a x a
and for B4,B5,B7,B8 – a x 2a
Similarly for AZone representing 2nd floor equivalent segment.
Dimensions for A0, A3, A6, A9, A12 are planned as – 2a x 2a
for A1,A2, A10,A11 – 2a x a
a) The 2D layout with two partition each representing one floor in 3D for
opposite ladder architecture and for A4,A5,A7,A8 – a x 2a.
In this architecture we placed four ladders with two along x plane
with inclination along y plane and the other two along y plane with
inclination along x plane as shown in figure 8 a) and b) .Each ladder
has a height of h = a-2 and inclined height (a+1)..

b) The 3D layout with opposite ladders and the distributed zones

a) The 2D layout with two partition each representing one floor in 3D for Four
ladder architecture

c) Detailed cross sectional view of opposite ladder orientation

Figure 7. Zone based 2D and 3D configuration (opposite ladder based design)
Procedure for distance calculation between pair of points each lying
between either segments or floors for opposite ladder architectureare
as follows:
1. Assign corner points for each zone to formulate the
boundaries. b) The 3D layout for four ladder architecture and the distributed zones
 !

c) Detailed cross sectional view of four ladder orientation Figure 8. Zone

based 2D and 3D configuration for four ladder based design e) Cross sectional view and f) Isometric x ray view (four ladder)
Procedure for distance calculation between pair of points each lying Figure 9. Cross sectional and x ray view of the three proposed architectures
between either segments or floors for opposite ladder architectureare The 3D view for the three proposed 3D layouts are shown in figure 9.
as follows: The curves with variation in distance ratio with ‘a’values are shown
1. Assign corner points for each zone to formulate the In figure 10 a) and 10 b). The proposed four ladder architecture shows
boundaries. best improvements for 3D in comparison to the other two proposed
2. For 2D calculate Manhattan distance between one cell in design. The 2D/3D distance ratio values tend to saturate with increase
segment 1 and another cell in segment 2 in the same in dimensions as indicated by the ‘a’ value.
manner as used for parallel ladder architecture.
3. Finally compute the overall sum of distances in 2D.
4. For 3D design assign entry point for each zone in floor 1.
5. Similarly assign exit point for each zone in floor 2.
6. Based on the entry and exit points assignment for each
zone the ladders are assigned accordingly.
7. Calculate the distance from any cell in a given zone to the
assigned entry point to the respective ladder.
a) 2D/3D Distance ratio comparison for the three proposed architectures
8. Calculate the distance of the destination cell for the
stipulated zone to the assigned exit point from the
respective ladder..
9. Obtain the landing location in the 2nd floor ITZ on climbing
the ladder from the entry point. Calculate the distance
between the landing location and the respective exit point
for the destination zone.
10. Add the ladder height together with all the three distances
calculated in steps 7,8 and 9 .Thereby total distance
between two cells at different floors(segments) are b)2D/3D Distance ratio variation with different ‘a’ value
computed. Figure 10. 2D/3D distance ratio vs. ‘a’ value curves
11. Then sum up all the distances computed between each cell V. APPLICATION OF 3D ARCHTECTURE ON SPECIFIC LAYOUTS
in floor 1 and that in floor 2. Here initially we have shown two different 2D layouts in figure 11 a)
12. Also compute zone wise distances for both 2D and 3D and 11 e).The layouts displays a set of nets (each comprising of
layouts and finally obtain the ratio between overall 2D and source and destinations) having end to end routes. Such routs always
3D distances for given parameter value ‘a’. result in unwanted crossover which can not be avoided using any
Finally for different values of ‘a’ parameter in all three types of kinds of detour. Figure 11 b),c),d)shows application using equivalent
proposed 3D layout the 2D and overall 3D distances as well as the 3D layouts comprising of the same nets with three proposed
corresponding 2D to 3D ratio are calculate. Table 1 shows the results. architectures for layout 1. It has been observed that the 3D
Table 1. total cell distances for both 2D and 3D layout with 2D/3D application reduces the contamination considerably as well as results
distance ratio for the three proposed architectures in major improvements in route performance in terms of latest arrival
time, cell utilization etc. Similar 3D implementation is displayed in
figure 11 f),g) and h).
All sources in both cases for the 3D layout are placed at floor 1 and
targets at floor 2 .However sources being placed in 2D in the second
half (which has been mapped as floor 2 in the 3D layout) is being
arranged to a symmetrical location at floor 1 in the 3D configuration.
The same placement is applied for targets placed in first half in 2D.
Table 2 display the detailed route performance for 2D as well as all
the 3D based application for the same bioassay. Figure 12 ),b) and c)
shows the route performances as bar charts for the given layouts in
terms of Latest arrival time, contaminations and cell utilization.

a) Cross sectional view and b) Isometric x ray view (parallel ladder)

a) 2D layout with routes for case 1 b) 3D parallel ladder appln. (Case 1)

c) Cross sectional view and d) Isometric x ray view (opposite ladder)

 !

c) 3D opposite ladder appln. (Case 1) d) 3D four ladder appln. (Case 1)

c) Cell Used
Figure 12. Route performance for 2D and proposed 3D based layout
implementing two specific bioassays
REFERENCES
[1] R. B. Fair. “Digital microftuidics: is a true lab-on-a-chip possible?”Microfluidics
and Nanofluidics, 3(3):245-281, 2007.
[2] K. Chakrabarty and F. Su. “ Digital Microfluidic Biochips: Synthesis,Testing,
e) 2D layout with routes for case 2 f) 3D parallel ladder appln. (Case 2) and Recmifiguration Techniques.” CRC Press, Boca Raton, F L,2006.
[3] M. G. Pollack, R. B. Fair, and A. D. Shenderov, “Electrowetting-
basedactuation of liquid droplets for microfluidic applications,” Appl.
[Link]., vol. 77, no. 11, pp. 1725–1727, 2000.
[4] J. Lee, H. Moon J. Fowler, C.-J. Kim, and T. Schoellhammer,
“Addressable micro liquid handling by electric control of surface
tension,” in Proc. Int. Conf. MEMS, 2001, pp. 499–502.
[5] S.-K. Cho, S.-K. Fan, H. Moon, and C.-J. Kim, “Toward digital
microfluidic circuits: Creating, transporting, cutting and merging liquid
g) 3D opposite ladder appln.( Case 2) f) 3D four ladder appln. (Case 2) droplets by electrowetting-based actuation,” in Proc. TechDig MEMS
Figure 11. Two different layouts – with 2D and 3D application using all Intl. Conf., 2002, pp. 11454–11461.
proposed architectures [6] P. R. C. Gascoyne and J. V. Vykoukal, “Dielectrophoresis based sample
handling in general-purpose programmable diagnostic instruments,”
It has been observed that in terms of each parameter indicating the
Proc. IEEE, vol. 92, no. 1, pp. 22–42, Jan. 2004.
route performance – any of the 3D based design enhances the [7] D. A. Anton, J. P. Valentino, S. M. Trojan, and S. Wagner,
performance in comparison to 2D layouts. As evident from table 2 “Thermocapillary actuation of droplets on chemically patterned surfaces
and figure 12, 3D layout with four ladders architecture handles by programmable microheater arrays,” J. Microelectromech. Syst., vol.
contaminations most effectively as it reduces all contaminations to 12,no. 6, pp. 873–879, 2003.
zero for the given two cases. For cell utilization purpose opposite [8] A. Renaudin, P. Tabourier, V. Zhang, C. Druhon, and J. C.
ladder architecture performs better in comparison to other layouts. Camart,“Plateforme SAW dédiée à la microfluidique discrète pour
applications biologiques,” in Proc. Congrès Français de Microfluidique,
VI. CONCLUSION Société Hydrotechnique de France, 2004, pp. 14–16.
In this work we explored the scopes for application of ATDA based [9] F Su, Jun Zeng,“ Computer aided design and test for digital
3D DMFBs. We considered geometrical advantages of using 3D microfluidics”,IEEE Design and test for computers, pp. 60-70. Feb,
layout over 2D based ones. In the process we proposed three different 2007.
architectures and displayed individual advantages of using each type [10] J. Berthier, Microdropes and Digital Microfluidics, William Andrew
over 2D based layout in terms of overall distances between cells at Inc,Norwich, NY, 2008.
different regions marked in either of the floors (or segments for 2D [11] H. Yang, S.-K. Fan, C.-P. Lin, C.-T. Wu, and W. Hsu, "3D droplet
transportations by EWOD actuations on flexible polymer films,"
designs).We further considered two specific cases where 3D based presented at 2005 ASME International Mechanical Engineering
application efficiently reduces contamination and improves the Congress and Exposition, Orlando, FL, United States, 2005
overall route performance. The results as displayed in table 2 depicts [12] J.-M. Roux, Y. Fouillet, and J.-L. Achard,"3D droplet displacement in
encouraging improvements through the use of 3D based design. microfluidic systems by electrostatic actuation," Sensors and Actuators,
However there remains future scopes for formulating specific A: Physical, DOI:10.1016/[Link].2006.05.012,2006..
placement and routing strategies using different ladder based [13] Mohamed Abdelgawad and Aaron R. Wheeler “3d droplet actuation in
digital microfluidics devices ”,IEEE 14 th international conference on
architectures for 3D design .The architecture can be designed to be
solid state sensors,actuators and Microsystems , France,June 2007,pp.
protocol specific for further enhancement in performance in bioassay 1809 – 1812 .
execution. [14] Mohamed Abdelgawad, Sergio L. S. Freire,Hao Yang and Aaron R.
Table [Link] performance comparison for 2D and 3D based layouts Wheeler“ All-terrain droplet actuation”Lab-on-Chip Journal,issue5,2008
, pp. 672-677.
[15] J. Berthier, Y. Fouillet, Ph. Clementz, Ph. Dubois, P. Claustre, C.
Peponnet. “Actuation potentials and capillary forces in electrowetting
based microsystems.” Sensors and Actuators A: physical, Vol. 134 (2),
pp. 471-479, 2007.
[16] Nadjoua Moumen, R. Shankar Subramanian and John B. McLaughlin,“
Motion of Drops on a Solid Surface due to a Wettability Gradient”,
Langmuir, 2005, 21, 11844–11849
[17] F. Brochard, “Motions of droplets on solid surfaces induced by chemical
or thermal gradients”Langmuir, 1989, 5, 432–438.
[18] A. Carre and M. E. R. Shanahan,” Drop Motion on an Inclined Plane
and Evaluation of Hydrophobia Treatments to Glass”J. Adhes., 1995, 49,
177–185.
[19] Kuratowski, Kazimierz (1930), "Sur le problème des courbes gauches en
topologie", Fund. Math. (in French) 15: 271–283.

a) Latest arrival time b) Number of contaminations