Microsyst Technol DOI 10.1007/s00542-013-1738-7

TECHNICAL PAPER

Design and fabrication of 3-dimensional helical structures in polydimethylsiloxane for flow control applications Rajeev Kumar Singh • Avinash Kumar • Rishi Kant • Ankur Gupta • E. Suresh • Shantanu Bhattacharya

Received: 3 December 2012 / Accepted: 12 January 2013 Ó Springer-Verlag Berlin Heidelberg 2013

Abstract Soft lithography in 2-dimensional (2-D) was developed for polymer MEMS applications about two decades back. The technique was highly useful for replication of microstructure molds using a soft polymeric material called PDMS (polydimethylsiloxane). From its inception the process has been widely applied to microfluidics, biochips, hybrid biomedical microdevices etc. However, it was limited to only surface microstructures and 3-Dimensional (3-D) soft lithography although performed by some research groups involved some very precise and expensive techniques like stereolithography etc. The exploration of soft lithography in three dimensions by using a replication technique with copper wires with micron size diameters was performed by our group relatively recently (Singh et al. in International conference on MEMS, IIT Madras, Chennai, 2009). In this work we have used the 3-D replication and molding technique to develop concentric solenoid patterns around micro-channels in the bulk of PDMS. The solenoidal paths of various pitches ranging from 0.4 to 1.2 mm have been replicated in PDMS using an innovatively designed fixture. The solenoids have been structurally characterized using an inverted fluorescence microscope (Nikon 80i) for dimensional parameters Electronic supplementary material The online version of this article (doi:10.1007/s00542-013-1738-7) contains supplementary material, which is available to authorized users. R. K. Singh  A. Kumar  R. Kant  A. Gupta  S. Bhattacharya (&) Microsystems Fabrication Laboratory, Department of Mechanical Engineering, Indian Institute of Technology, Kanpur, India e-mail: [email protected] E. Suresh Genpact Ltd., Bangalore, India

like pitch, length etc. Further, the solenoidal path designs have been simulated, optimized and fabricated around a central channel of 80 l diameter and we have observed the repeatability of this fabrication process multiple times. The purpose of this architecture is to initiate valving action wherein fluid movement in the central channel can be restricted by filling the surrounding solenoidal track with compressed air at high pressure so that it can squeeze the centrally located micro-channel carrying the liquid. This valving structure may find a lot of applications in lab on chip devices, PCR biochips, biomedical micro-devices etc.

1 Introduction Elastomeric polymers are widely used in realizing lab on chip architectures due to their simple and inexpensive fabrication techniques. Of special significance is the polymer PDMS which is widely used to develop micro-fluidic devices catering to a variety of lab on chip applications. PDMS is most often used for realizing physical phenomena like pumping (Kant et al. 2012), valving (Lee et al. 2006), mixing (Choudhary et al. 2011) etc. at the microscopic length scales PDMS possesses a lot of strengths like biocompatibility, optical clarity and accuracy of conformation to varied shapes, structures, crevices etc. while being replicated onto such structures of sizes varying up to the nanometric length scales. The processes of replication and molding are widely used for fabricating 2-D and very recently some efforts have been made to fabricate 3-Dimensional (3-D) objects like buried micro-coils (Lee et al. 2009), channel network, helical channels etc. such processes however are complex like laser based precise exposure of UV curable resists (Geschke et al. 2004), photopolymerization in 3-D (Wang et al. 2002),

123

Microsyst Technol

Lithographie, Galvanoformung, Abformung (LIGA) (Rajaraman et al. 2007), localized electrochemical deposition (Madden and Hunter 1996) etc. Apart from this varied bulk and surface micromachining methods have been repeatedly used to fabricate 3-D structures on silicon, polymers and PDMS although they are limited, height wise (in the 2D). Some examples include use of laser patterning with variable foci to obtain high aspect ratio 3-D micro-channel arrays (Marcinkevicius et al. 2001), use of photo-polymerization by two photon absorption with a Ti–Sapphire laser to obtain helical coils of diameter 6–7 lm (Maruo et al. 1997), use of two photon driven photopolymerization in modified PDMS (containing a photoinitiator) by using hydrosilylation and radical initiated cross linking mechanisms (Coenjarts and Ober 2004), using reactive ion etching on silicon (Rao et al. 2004), using synchrotron radiations over PTFE (poly tetra fluoro ethylene) (Katoh et al. 2001) using thermosensitive resins and selective modification in the bulk of the resin by using variable foci lasers (Yamakawa et al. 2004) etc. All the methods illustrated above involve precise and expensive optics, beam sources and sometimes controlled environmental conditions like high vacuum and inert atmosphere and therefore are not very amenable to high throughput applications. A much easier technique of replication around thin threads of nylon has been extensively demonstrated by researchers (Verma et al. 2006) although the accurate dimensional control and of the basic diameter of the replicating nylon wire was not controlled. In order to address this dimensional control aspect our group had developed a Fig. 1 Schematic of solenoid micro-valve (number of turns = 5, R = 1,250 lm, extrusion length = 10,000 lm and rectangular block dimension 25,000 9 20,000 9 12,000 lm)

123

technique wherein etched copper wires with round sections and controlled dimensions (up to 30 l) (Singh et al. 2009) were used to replicate micro-channel arrays within bulk of PDMS. Here we have tried to explore another aspect of this methodology wherein we realize 3-D solenoidal structures around the micro-channels placed at various depths within the bulk of the PDMS material. The requirement of developing this process was stimulated by an idea of a microvalving architecture wherein the central channel would be used as the main fluid channel and the helical loop around this central channel would be filled with compressed air and would act as a gating mechanism to control the flow along the central channel (Fig. 1). This paper explores the building of such an architecture at the microscopic length scale in a repeatable manner in details. The design is first characterized using COMSOL MultiPhysics and the optimized design is fabricated using replication. The replica is then investigated by optical micrographs and a MATLAB code that evaluates the concentricity of the microchannel with respect to the solenoidal tracks.

2 Experimental 2.1 3-Dimensional fabrication of micro-channel in PDMS We had developed a technique prior to this to fabricate 3-D micro-channel array in PDMS by replication and molding

Microsyst Technol

2.1.1 Characterization of the channel diameter The PDMS with embedded micro-channel was cut into thin sections in a manner so that the channel could be visualized cross-sectionally. An optical microscope (Nikon 80i) mounted with a CCD camera with a brightfield option was then used to visualize this channel cross-section multiple times at different magnifications and an image analysis software (ImagePro, M/s Cybernetics, USA) was used to scale the image at the same magnification that was used to acquire the image data. The various thin slices of a single diameter were heated for 1–6 h at 85 °C in a gravity fed convection oven (M/s Khera Instruments, Delhi) in the step of 1 h and were then analyzed using the software. Fig. 2 Straight wire in mold box using fixture

processes (Singh et al. 2009). As one part of the helical valve assembly is a central channel a quick review is made of this process so as to realize the central channel at a suitable depth within the PDMS. The process starts with a plastic box (of size 40 9 45 9 19 mm) wherein the drilling of holes on opposite walls of this box are accomplished for placing a finely etched copper wire of circular cross-section and diameter of 80 l. A CNC drilling machine DT100 (M/s Mikrotools, Singapore) drills holes on opposite faces of the box by reading a CAD file containing the exact positional information of the holes and both ends of the copper wire are inserted into the drilled holes with the help of a special tweezers. The wire is further tied in a jig on both ends and a suitable value of tension is provided to this wire (Fig. 2). The standalone wire is first replicated individually by the PDMS by pouring the prepolymer (M/s Dow Corning, Midland) after mixing it in a ratio of 10:1 by weight with a curing agent inside the plastic box submerging the copper wire. This plastic box is then kept in desiccator for degassing following which it is heat cured at 85 °C for 45 min in a gravity fed convention oven. The casting with the embedded copper wire dangling on both sides is removed from the plastic box and this assembly is dipped in toluene for 12 h to swell the PDMS matrix thereby releasing the grip on the copper wires and taking the wire out to create open micro-channels embedded within the PDMS. This process is further followed by drying of the swelled PDMS piece (temperatures used range from 25–85 °C) with the channel cavity. The fabrication flow chart is shown in supplementary figure S1. Here we would like to illustrate that while the standalone channel was replicated using PDMS the drying step incorporated a change in the overall channel diameter which has been characterized in detail. The final assembly of micro-valve was realized in a separate step wherein both the central wire and the helical wire were assembled together and this assembly was replicated as a whole (Sect. 2.2).

2.2 COMSOL simulations used for designing the helical micro-valve We have used COMSOL (version 4.1) for finding out the interaction between the actuating helical track and the fluid containing micro-channel for modeling the micro-valving action of the assembly. The micro-valving for different combinations of central channel radius (r) and radius of the helix (R) are simulated in a scaled up ratio of the solenoid valve. The scaling up of the valve dimensions and also the operating pressure is necessary for the convergence of the simulated output as detailed in the results and discussion section. The scaled up operating pressure of compressed air within the helical track is kept at 550,000 Pa (79.77 psi) by looking at the reduction in the cross-sectional area of the central channel by virtue of the squeezing action generated by the helical track. All simulations have been performed using a HP Compaq dx2480 Business PC with 8 GB RAM and Intel(R) Core(TM)2 Quad CPU Q8200 at 2.33 GHz processor running on Windows-64 platform. The different combinations of ‘r’ and ‘R’ are used to look at the reduction of cross-sectional area and the optimized combination for which reduction in cross-sectional area is the highest is used as the final geometry for fabrication of the valve. A 3-D geometry, with real dimensions indicated in Fig. 1. The helical radius (R) = 1,250 lm with 5 turns and ends circumferential and a central channel of radius ‘r’ = 600 lm parallel to the co-ordinate planes. Both ends of the helical tracks extruded to the length of 10,000 lm are placed in a rectangular block of dimension 12,000 9 20,000 9 25,000 lm as drawn in the geometrical module of COMSOL. Although the actual PDMS block size is 45 9 40 9 19 mm the circumscribing smaller block would be mostly compressing towards the helical centre of the assembly when enclosed in the bigger block and thus the smaller block gives a rigid estimate of the micro-valving effect which if contained within the bigger block would have a more valving because of limited expansion of the

123

Microsyst Technol

helical track in an outward direction. Unstructured meshing containing different mesh elements with a point wise variability of the material property is used for solving the micro-valving phenomenon. Our interest in this simulation is to evaluate the deformation and stresses generated in the helical track because of inducing compressed air which would cause the fluid within the central channel to get guided by the deflecting channel wall. Thus a change in the properties of the mesh on a point to point basis is obviously required. A 3-D geometry with the dimension illustrated in Fig. 1 and meshed using moving mesh physics is simulated wherein the structural analysis equation has been coupled to Incompressible Navier-Stoke’s equation to model thus fluid structure interaction leading to estimation of volume force (body force) in the fluid. Equations 1, 2 and 3 below represent the structural mechanics relation, conservation of momentum and continuity equation respectively. The structural mechanics Eq. 1 is coupled to the conservation of mass and momentum and (Navier-Stoke’s equation) in 3-D continuity (Eqs. 2 and 3) using the arbitrary Lagrangian–Eulerian formulation (ALE) in which the solver works on a freely moving deformed mesh which constitutes the boundary of the fluid domain. q

o2 X  ðr  rÞ ¼ F ot2

q


 oU þ qðU  rÞU ¼ r  pI þ l rU þ ðrUÞT þ F ot

ð1Þ

ð2Þ qr  U ¼ 0

ð3Þ

where F is volume force (body force), r is stress tensor, X is displacement vector, l is the kinematic viscosity, q is the density, U = u(u, v, w) is the velocity field, p is the pressure, I is the identity matrix. The boundary conditions used for this simulation are given as U = 0.001 m/s at inlet and p = 0 at outlet. The deformation of this mesh relative to the initial shape of the domain is computed using Laplace smoothening (User guide COMSOL MultiPhysics version 4.1). If the number of variables in any of the above equations is more than the number of equations then we cannot get a unique consistent solution numerically. For such a solution we have to develop a set of constitutive equations wherein two property variables of a particular material or particular flow situation have a mathematical relationship. We have also used a coupled form of solid stress strain equation for a linear isotropic elastic material. The results of these simulations are summarized in Fig. 3 and Table 2. Table 2 shows a summary of percentage compression in cross-sectional area as a function of R and corresponding R/r ratio. Figure 3 shows the representative snapshot of the

123

simulated reduced area for a certain R/r value for the optimized dimensions and number of turns. 2.3 Fabrication of 3-D helical structure circumscribing the microchannel The fabrication of the helical actuator track needs the placing of the helix around the central microchannel following by replication of this assembly with PDMS. The most critical dimension needing process accuracy is the pitch of the helical coil for which the fabrication step necessitates an instrumentation which can be used for making constant pitch helices at different pitches for the optimal R/r value. The optimum R/r as discussed later is observed as 15.6 for a R of 1,250 lm respectively. 2.3.1 Fixture for coil making For spinning the helix around the central channel we have designed and developed a fixture which can layout the wire at various pitches starting from 0.4 to 1.2 mm in the steps of 0.2 mm for the various R and r combinations including the optimum combination (Fig. 4a). The helix is realized by a combination of linear movement of the slide in the XX’ direction which dispenses the wire with a tension ‘T’ in the ‘dcb’ direction as indicated by the arrow and the spinning of the wire over a needle of varied diameters (1.0–3.0 mm) clamped in a universal fixture and rotated by the gear train towards the side of the fixture assembly along the axis QQ’ at an angular velocity ‘x’. The opposite extremity of the wire is clamped at ‘a’ so that the tension is maintained while the spinning process is continued. The wire is spun at a pitch defined by the pitch of the lead screw of the carriage fitted at location ‘E’ and gets varied as the lead screw is changed from trial to trial. The rotation power to the fixture is manually provided to the crank wheel and the gear train placed on the right side of the fixture assembly which also defines the carriage speed in the XX’ direction. The wire so spun is a copper wire and once the helix is spun under a critical tension force the wire does not recoil back substantially even if it is removed from its spindle (needle) owing to the small diameter of the wire. 2.3.2 Fixture for concentric placement of helical coil with straight wire The wires on the flanges of the helix are given a bend of 90° by holding it over an edge under tension with tweezers and these flanges are then mounted on a second fixture containing the plastic box holding the central wire in tension with the solenoid mounted around this central wire (Fig. 4b). The level of the helix placed around the central wire is adjusted by moving the stage P in the ZZ’ and the

Microsyst Technol

Fig. 3 a Cross-section image of channel, b Plot of various channel diameter obtained corresponding to wire of 33–88 l, c Possible explanation of the reduction in diameter

123

Microsyst Technol

YY’ direction using two set of lead screws while visualizing the whole assembly by a set of two digital cameras. The cameras take the real-time pictures of the valve assembly and we use image analysis and keep shifting the relative position of the solenoid with respect to the central wire so that it comes in the geometric center of the helical assembly. Figure 5a and b shows the front and top elevation of a representative sample having problems in alignment of the helix mold with respect to the central wire and Fig. 5c and d show similar images after correction in alignment using our process. Figure 6 shows the front and top elevation pictures acquired by two cameras (M/s iBall) which is further stored in a computer (HP Compaq dx2480 Business PC with 8 GB RAM and Intel(R) Core(TM)2 Quad CPU Q8200 at 2.33 GHz processor running on Windows-64 platform) as digital images. The mold is fabricated iteratively by repeated imaging and image analysis by an inhouse developed MATLAB code. The code identifies the boundaries of the solenoid assembly and manual the positioning of the central wire with respect to the outside solenoid is carried out iteratively until the maximum offset is below a threshold value of less than 20 % (Seth 2012). The corresponding adjustment values to make helical coil concentric with central channel are given

Fig. 4 Snapshot of reduced area for a certain critical R/r value

123

in Table 2. The supplementary table S1 represents the MATLAB code used for helical coil alignment. 2.3.3 Replication of PDMS The assembly is replicated in PDMS using a mixture of curing agent and pre-polymer (M/s Sylgard 184, Dow Corning) in a ratio 1:10. The degassed mixture is poured inside the casing containing the mold assembly of metal wires. This mixture is cured in room temperature for 48 h to cure the PDMS mix. The post cured sample is taken out of the casing and dipped in toluene for 24 h following which the wires are pulled out using a special tweezer and the matrix is shrinked back by heating at 85 °C for 4 h as per the earlier developed protocol (Singh et al. 2009).

3 Result and discussion 3.1 Microchannel fabrication in PDMS The device containing the helical and central micro-channels is fabricated by using replication and micro-molding processes. The micro-molding process has already been

Microsyst Technol

Fig. 5 a Fixture for coil making, b Fixture for alignment of coil with central wire Fig. 6 a Non-aligned from side view, b Non-aligned from top view, c Aligned from side view, d Aligned from top view

described earlier and we replicate the mold using PDMS and follow the method described earlier. The image of the fabricated micro-channel is captured using an inverted fluorescence microscope (Nikon 80i) by both fluorescence and bright/dark field option. The cross-sectional area of the

fabricated micro-channel in PDMS is shown in Fig. 7a which are circular in shape. The area of circular hole is computed using ImageJ (Courtesy: NIH, USA) and the diameter is numerically estimated. Figure 7b shows a plot of the various channel diameters obtained corresponding to

123

Microsyst Technol Table 1 Percentage shrinkage of the diameter of each of the microchannel with respect to the molding wire diameter corresponding to all the heating times Channel diameter (lm) 88

Major radius of solenoid valve (R) (lm)

Radius ratio R/r

Percentage area reduction

6h

1,200

2.00

88.89

0.5

1,250

2.08

88.89

2.33 2.50

82.63 69.75

Percentage reduction in diameter 1h 15

2h 8

3h 4

4h 2

5h 0.5

Table 2 Percentage area reduction of central micro-channel estimated by COMSOL multiphysics simulation

78

15

8

4

2

0.5

0.5

69

15

8

4

2

0.5

0.5

1,400 1,500

60

15

8

4

2

0.5

0.5

1,750

2.92

54.44

0.5

2,000

3.33

33.98

2,250

3.75

30.56

51

15

8

4

2

0.5

41

8

4

3

1

0.2

0.2

33

8

4

3

1

0.2

0.2 Table 3 Adjustment needed and adjustment done for coil alignment

wires of 33–88 l and their shrinkage characteristics, i.e., in case of each of the wire diameter used after heating for 4 h respectively. We observe convergence of the channel diameter data after 4 h of heating and therefore no further heating is needed. Table 1 shows the percentage shrinkage of the diameter of each of the micro-channel with respect to the molding wire diameter corresponding to all the heating times. The process of making the PDMS replica is extensively studied in an earlier work (Singh et al. 2009). Here we will only include the necessary data for fabricating the micro-valve. 3.2 COMSOL multi-physics simulation COMSOL multi-physics is used to simulate and optimize the polymeric solenoid valve. The geometries of the solenoid micro-valves have been scaled up for performing simulations as the characteristic dimensions of the features that are actually used are very small and resolution of the coupled model described above is low. We have modelled a geometry as indicated above with the helix diameter varying between 1,200 and 2,250 l. The simulations are performed over 92,839 elements. The solenoidal channel is filled with compressed air at a pressure of 550,000 Pa and the helix radius (R) of the solenoid channel around the central channel (r) is varied to study the effect of R/r ratio on the reduction in the cross-sectional area of the central channel. The total force exerted by the outer helix on the central channel is also a function of the helix pitch. The deformation of the central channel due to air pressure in the helix happens both radially and longitudinally although the longitudinal thrust is does not contribute to valving. The thickness of the layer formed between solenoidal channel and central channel plays a crucial role and better valving efficiencies are obtained at minimum thickness values and at higher thicknesses the resistance to deformation is more resulting in less percentage compression. So we aim to

123

Trial no.

Image no.

Adjustment done in degree revolution

Adjustment done in mm

Adjustment needed in RU

Z

Z

Y

Z

Z

–

–

0.27

0.27

0 0.5

1.27 0

0.19 0.21

0.19 0.21

Y

1.

Image 1

–

–

2. 3.

Image 2 Image 3

0 180

360 0

4.

Image 4

70

70

0.1946

0.22946

0.11

0.11

5.

Image 5

30

30

0.0834

0.06835

0.08

0.08

keep the helical radius ‘R’ at a minimum value that can be realized using the micro-fabrication technique illustrated above. The percentage compression is defined as follows %compression ¼

A0  Av A0

ð1Þ

where, A0 is original cross-sectional area and Av is crosssectional area after valving. The percentage compression is measured from the crosssectional images of a radial plane passing through the center of the solenoidal valve in compression mode. The area of the compressed state is irregular and thus we take the simulated cross-sectional area values by using postprocessing to the basic compression data generated. The percentage compression obtained for various R values for PDMS material with different R/r ratio are listed in the Table 2, 3. It is observed that for radius values varying between 1,200 and 2,250 l the percentage compression ratio varies between 88.89 and 30.56 % with a peak compression at 1,250 l of 88.89 %. Below this diameter we have observed a convergence in the percentage compression. The radius ratios as obtained vary between 2.08 and 3.75. The representative simulated cross-sections after deformation is shown in supplementary figure S2 which illustrates the cross-sectional area before and after compression. The compression pressure used for all our experiments is taken as 550,000 Pascal. At the minimum

Microsyst Technol

Fig. 7 Different trials for concentricity coil

123

Microsyst Technol

helix diameter as can be observed there is convergence of the percentage area which can be explained by a maximum compression of the central channel followed by a higher longitudinal deformation which happens mostly if the radius ratio is lower. 3.3 PDMS replication process The liquid PDMS after pouring in the mold box is allowed to cure by necessary heating and then the procedure used to release the mold is performed as detailed earlier. We have observed that after the toluene induced expansion of the PDMS replica and its subsequent heating occurs the PDMS shrinks to a lower diameter than the original mold wire diameter. The reducing diameter of the embedded microchannel within PDMS as the expanded replica is heat cured at 85 °C for different instances of time is plotted in Fig. 7b. The diameter reduces by 15 % on a mold wire diameter of 88 l after a heating time of 1 h is elapsed. The percentage reduction changes by almost an order of magnitude (2 %) as the mold wire diameter is below 60 l. The heating is subsequently carried out from 2–6 h and it is observed that the percentage reduction in diameter reduces substantially with the progress of time so that after around 4.50 h of heating a convergence in the diameter value is reached. So, we conclude that the process reports stable values of microchannel diameters after a 4.50 h heating protocol is executed. We have tried to find out a possible reason for this shrinkage in overall diameter. PDMS comprises of a monomer unit [CH3–SiO–CH3]n and has a structure with long and short chains that are positioned and entangled randomly (Bhattacharya et al. 2005). The distance between these highly entangled network of molecules is substantially altered as liquid toluene gets into this structure which results in an overall dimensional expansion in bulk form.

In this condition as the expanded PDMS is heated at 85 °C the toluene evaporates and the polymeric chains try to redistribute and thus the overall void space which was initially present before the expansion step is also redistributed and probably the total amount of void space present in the pre-expanded configuration reduces. This model seems to explain the reduction in the overall diameter of the micro-channels from the diameter of the molding copper wire. The volume change induced by this process seems to be homogeneous as there is hardly any change in the aspect ratio of the micro-channels. The overall bulk volume of the PDMS also reduces thus reducing the length of the micro-channels. In order to ascertain the variation in void space a BET test was performed on the original PDMS sample and the post heated sample. The total pore volume of these samples for a pore size smaller than 3.0 l is reported as 1.965 9 10-3 and 8.235 9 10-4 cc/g, respectively for the original PDMS sample and the post expanded and heated sample respectively. This shows that there is an almost order of magnitude change in the total pore volume per gram of the material which is also confirming an overall reduction in the void space. Figure 7c shows a schematic of the process of this reduction mechanism of the void space. The supplementary figure S3 and S4 represents the N2 adsorption and desorption during the BET test. 3.4 Characterization of solenoidal micro-valve The mold of the helix is prepared with the mechanism described above in details. The helix mold is made with pin diameters of 1.0, 1.5, 2.0, 2.5 and 3.0 mm and a pitch of 0.4 mm are realized. Although eventually we intend to fabricate and test the micro-valve for a certain optimized configuration (based on maximum squeezing) as reported

Fig. 8 Representative image of a valve assembly for a pitch of 1.2 mm and a pin diameter of 2.0 mm

123

Microsyst Technol

by the simulation data but we have tried a variety of sizes of the valve assembly on the fixtures successfully. The helix wire mold is taken out of the fixture and is positioned with respect to the central wire following which the assembly is replicated with PDMS and the mold is pulled out as per the methodology detailed earlier. We next inject a fluorescent dye (acridine orange) into the helical channel and dark ink in the centrally located micro-channel and use the Nikon 80i to digitally acquire the image. Figure 8 shows a representative top and side elevation of a valve assembly with a pitch of 1.2 mm and a pin diameter of 2.0 mm. Although there is some spring back of the helically spun wire after releasing from the pin (because of the elastic nature of Copper) we have taken care of this problem by taking out the helix wire mould in a spun condition on the top of the pin and heating this assembly in an over at 100 °C for about 1 h. Once this heat treatment is done then the only deviation in dimensions are observed because of buoyancy forces of liquid PDMS but as the coil is hung from the fixture in Fig. 4b by straightened ends of the copper wire and also by virtue of the weight of the helix assembly the final deviation due to buoyancy is negligible.

4 Conclusions We have designed and fabricated a solenoidal micro-valve by using the processes of replication and molding. The valve design was simulated and optimized for maximum compression. This valve can be used for a variety of applications in the field of micro-fluidics and lab on chip etc. The valve eventually constructed contains a helical track of helix diameter 1,250 l which is concentrically including a micro-channels of diameter 80 l. We have developed a novel replication technique using wires of small diameter to realize high aspect ratio 3-D microchannel array and helical structures. The resolution of our replication process is 80 l as below this value the wire molds are strength-wise insufficient and the replication process fails. Simulations have been performed to optimize the design of the valve and the testing of valve efficiency is currently ongoing. We envision this micro-valve to have a lot of applications in the biomedical industry which requires dispensing of minuscule volumes of fluids.

References

oxygen plasma treatment and correlation with bond strength. J Microelectromech Syst 14(3):590–597 Choudhary R, Bhakat T, Singh RK, Ghubade A, Mandal S, Ghosh A, Rammohan A, Sharma A, Bhattacharya S (2011) Bilayer staggered herringbone micromixers with symmetric and asymmetric geometries. Microfluid Nanofluid 10:271–286 Coenjarts CA, Ober CK (2004) Two photon three dimensional micro fabrication of polydimethylsiloxane elastomers. Chem Mater 16:5556–5558 Geschke O, Klank H, Telleman P (2004) Microsystem engineering of lab-on-chip devices. Wiley, Germany Kant R, Singh H, Nayak M, Bhattacharya S (2012) Optimization of design and characterization of a novel micro-pumping system with peristaltic motion. Microsyst Technol. doi:10.1007/s00542012-1658-y Katoh T, Nishi N, Fukagawa M, Ueno H, Sugiyama S (2001) Direct writing for three dimensional microfabrication using synchrotron radiation etching. Sensor Actuator A 89:10–15 Lee S-W, Kim D-J, Ahn Y, Chai Y G (2006) Simple structured polydimethylsiloxane microvalve actuated by external air pressure. Proc IMechE J Mech Eng Sci 220(C):1283–1288 Lee D, Mekaru H, Hiroshima H, Matsumoto S, Itoh T, Takahashi M, Maeda R (2009) 3D replication using PDMS mold for microcoil. Microelectron Eng 86:920–924 Madden JD, Hunter IW (1996) Three dimensional microfabrication by localized electrochemical deposition. J Microelectromech Syst 5(1):24–32 Marcinkevicius A, Juodkazis S, Watanabe M, Miwa M, Matsuo S, Misawa H, Nishii J (2001) Femtosecond laser assisted three dimensional micro-fabrication in silica. Opt Lett 26(5):277–279 Maruo S, Nakamura O, Kawata S (1997) Three dimensional microfabrication with two photon absorbed photopolymerization. Opt Lett 22(2):132–134 Rajaraman S, Choi S-O, Shafer RH, Ross JD, Vukasinovic J, Choi Y, DeWeerth SP, Glezer A, Allen MG (2007) Microfabrication technologies for a coupled three-dimensional microelectrode, microfluidic array. J Micromech Microeng 17:163–171 Rao MP, Aimi MF, MacDonald NC (2004) Single mask three dimensional microfabrication of high aspect ratio structures in bulk silicon using reactive ion etching lag and sacrificial oxidation. Appl Phys Lett 85(25):6281–6283 Seth AW (2012) Optimization of laser machining process for preparation of photo mask and its application to microsystems fabrication. Dissertation, Indian Institute of Technology Singh RK, Ghubade AB, Basu B, Bhattacharya S (2009) A novel replica molding process for realizing three dimensional microchannels within soft materials. In: International Conference on MEMS, IIT Madras, Chennai Jan 3–5, 2009 Verma MKS, Majumder A, Ghatak A (2006) Embedded template assisted fabrication of complex microchannels in PDMS and design of a microfluidic adhesive. Langmuir 22:10291–10295 Wang I, Bouriau M, Baldeck PL, Martineau C, Andraud C (2002) Three dimensional microfabrication by two photon initiated polymerization with a low cost microlaser. Opt Lett 27(15):1348–1350 Yamakawa S, Amaya K, Gelbart D, Urano T, Lemire-el-more J (2004) Development of three dimensional microfabrication method using thermo-sensitive resin. Appl Phys B Laser Optic 79:507–511

Bhattacharya S, Dutta A, Berg JM, Gangopadhyay S (2005) Studies on surface wettability of polydimethylsiloxane and glass under

123