Conference-ICSFS-14-Effect of Surface Modification on Laminar Flow in Microchannels Fabricated by UV-Lithography

The laminar motion of fluid in microchannels is the necessary criteria for integrated microfluidic system as lab-ona-chip. Experiments were conducted to investigate laminar flow characteristics of dyed water through photoresist microchannels with square pillars. We found that the pillar dimensions on the channel surfaces have significant impact on the flow rate. The evaluated Reynolds number was less than unity in each microfluidic flow. The compatibility between the pillar sizes and corresponding air-water meniscus movement in microfluidic flow has been reported. [DOI: 10.1380/ejssnt.2009.330]


I. INTRODUCTION
The Reynolds number is useful parameter to estimate the tendency of turbulence in microfluidic flow [1].Kenis et al. [1] have used "fabrication using laminar flow" or FLO to fabricate microstructures inside microchannels using conventional photolithography.Ichikawa et al. [2] have experimentally verified the theoretical prediction of the laminar fluid flow through rectangular microfluidic channels.According to Cheow et al. [3], the presence or absence of a liquid slug is the true or false respectively in two-phase flow for digital microfluidics.The working principle of the microfluidic logic devices is based on the change in hydrodynamic resistance by variation of surface hydrophilicity for a rectangular microchannel having constant height containing droplet in low Reynolds number regime.Damean et al. [4] have experimentally demonstrated the fabrication of posts by using ferromagnetic photoresist.Shirtcliffe et al. [5] fabricated the high aspect ratio photoresist pillars as superhydrophobic by measuring the static contact angles of water for different pattern height.He et al. [6] experimentally demonstrated the contact angle hysteresis on PDMS (Polydimethylsiloxane) coated SU8 (a photopatternable epoxy) pillars as rough hydrophobic surface.They concluded that the Cassie drop is more preferable than Wenzel drop in applications involving moving droplets.Suk et al. [7] have varied the area of hydrophobic patterns in their microchannels to find the variation of passive capillary flow with respect to hydrophobic nature of the channel surface.Nordstrom et al. [8] have presented a method to reduce the contact angle of the SU8 surface to facilitate the fabrication method of devices for 'lab-on-a-chip'.Lee et al. [9] gave the first demonstration of the wettability switching mechanism using surface roughness modification.They used the theoretical analysis to explain the experimental results.Stephan et al. [10] fabricated array of square pillars and round pillars of photoresist by soft-lithographic technique for microfluidic applications.
In the present study, we have experimentally demonstrated the effect of photoresist pillars and wettability of rectangular microfluidic channel surface on laminar fluid flow determined by Reynolds number to set the platform for integrated microfluidic systems as lab-on-a-chip.The flow rate of air-water meniscus and transit time were estimated and compared in all microfluidic devices.

II. EXPERIMENTAL PROCEDURES A. Fabrication of microchannels
Various microfluidic channels were fabricated on glass microscope slides coated with SU8/polyimide photoresist using maskless lithography.The photoresists were coated by a spin coater, with 1000 RPM for polyimide and 3000 RPM for SU8.In brief, the microfluidic devices, having rectangular microchannels, were fabricated having the SU8 side walls, polyimide bottom surface containing polyimide square pillars of side lengths (at the top surfaces of the pillars) of 100 µm, 150 µm, 200 µm and 350 µm and subsequently sealed by a glass slide coated with PMMA (poly methyl methacrylate) dissolved in acetone as adhesive material to SU8.Each microchannel was fabricated with one particular side length of pillars on the bottom surface.The heights of the pillars were around 6 µm in all the microchannels.Details of the fabrication procedures can be found elsewhere [3,[11][12][13].

B. Thermal bonding and fluidic measurements
The thermal bonding has been done by applying pressure during the hard baking with a temperature ramp of 150 • C to 195 • C. We have recorded and measured the characteristics of dyed water flow through the channels in all the above devices by a CMOS camera catching 25 frames per second.The Reynolds number was estimated in each device type, for the flow of water through different channels using the density, viscosity, saturated meniscus velocity of the liquid and the channel height.The temperature of the dyed water was 26 • C during the microfluidic flow in all channels.

III. RESULTS AND DISCUSSIONS
We have used the surface profilometer to measure the height, side length at the top and bottom surfaces of the square pillars and distance between the pillars at the bottom surface in each microchannel.A Surface profilometer scan is shown in Fig. 1 which is on the polyimide pillars of square shaped horizontal cross-section having side length of 200 µm at the pillar top surface.It was around 250 µm at the bottom of the square pillars.The side length of the square pillars was not uniform along the height, as the UV beam used in maskless photolithography was not perfectly collimated.The height of the pillars is around 6 µm.The measured side lengths at the top and bottom surface of the square pillars were listed in Table I.The aspect ratio (A) of the pillars was evaluated using the relation [14], A = l 1 /l 2 , where l 1 = side length of the square pillars at the pillar top surface; l 2 = side length of the square pillars at the bottom of the pillar.
The capillary driven flow phenomenon in the microchannel is considered in this study.The capillary driven flows are predominantly governed by the forces associated with surface tension [2] and offer passive flow enhancement effect.Microfluidic imaging is used to experimentally visualize the interface movement in the microchannel.The four images in Fig. 2 represent the dyed water flow through a microchannel with pillars of 200 µm (l 1 ).
Several interesting observations can be made from the topological transformation profiles in Fig. 2. They are mainly: (a) the capillary driven flow produces a convex meniscus profile at the entrance of the chamber, (b) meniscus profile attains a linear shape when the channel aspect ratio (channel height/channel width) decreased to a constant value, (c) once a straight-line profile was created and that interface profile constantly developed until the fluid completely filled the chamber.As the channel width increased in the chamber region, the cross-sectional surface area increased, consequently the surface energies of the air-water interface and water-channel wall interface are increased which in turn reduced the velocity of the air-water interface in the chamber region than that of the neck region of the microfluidic channel.The Reynolds number were estimated using the relation: R e = ρvL/µ, where ρ is the density of the liquid, v is the saturated velocity of liquid flow in the chamber region, L is the characteristic length which is the height of the rectangular microchannel and µ is the coefficient of viscosity of the liquid.The evaluated Reynolds numbers from the laminar flows in different microchannels are listed in Table I.Reynolds number is linearly proportional with the velocity of the liquid flow.As the saturated velocity increased with the decreasing polyimide pillar side length, the Reynolds number increased.The Reynolds number was less than 1 in case of all channels having pillars of different side lengths.So in this study the microfluidic flow was always laminar in all the microchannels.
According to Fig. 3, the total transit time (time taken by the meniscus to travel from the inlet to outlet) gradually increased with the increasing side length of the square pillars on the bottom surface of the channels.The total transit time for the microfluidic flow was 0.87 sec for the microchannel containing square pillars on the bottom surface having 100 µm side length (l 1 ), but it increased up to 23.15sec for the microchannels containing square pillars on the bottom surface having side length of 350 µm (l 1 ).The surface area faced by the dyed water is higher for the larger pillars having the same spacing and height of the pillars.Hence the surface energy of the interface between water and channel wall is higher in case of larger pillar sizes.Hence the increase of surface energy may be a reason for the increase of total transit time in case of microfluidic channels containing larger pillars.The results in Fig. 3 can be explained well using the friction factor in the microfluidic channel as mentioned in the recent numerical reports [14,15].Wang et al. [15] have numerically investigated the effect of friction factor on microfluidic flow by varying the height, size and spacing of surface roughness elements on both of the two parallel plates of the microchannel.In their investigation, when the size of the roughness element increases considering the height and   spacing as fixed parameters then the friction factor also increases.Rawool et al. [14] have numerically investigated the effect of 3D surface roughness elements on microfluidic flow.They have found that the aspect ratio, height and pitch of surface roughness elements have important effects on flow through rough microchannels.According to Rawool et al., for a fixed normalized height of obstruction, the friction factor increases with increasing aspect ratio of obstruction.In the present study, the pillar height, pillar spacing and channel height are maintained at and around 6 µm, 100 µm and 60 µm respectively.Only the pillar side length and aspect ratio changed in the four devices (Table I).So, as the size and aspect ratio of the pillars increased, the friction factor also increased, which in turn increased the total transit time by decreasing the saturated velocity and as a result the flow becomes more laminar, hence the Reynolds number is also less.Figure 4 shows the variation of air-water interface velocity with interface position from the inlet regarding different side lengths of polyimide square pillars on the bottom surface of the microchannels.The air-water interface velocity increased with the distance from the inlet and attained a peak in the transition region of the channel aspect ratio and then decreased to a nearly constant value or with a small constant slope in the chamber region.The nature of the interface velocity variations with respect to the distance of interface from the inlet was similar in different microchannels containing any particular side length of polyimide square pillars.The maximum peak velocity of the air-water interface was 7.41 cm/sec in the microchannel containing square pillars on the channel bottom surface having 100 µm side length at the pillar top surface and we have found minimum peak velocity of 0.08 cm/sec in the case of square pillars of 350 µm side length at the pillar top surface.Our observation is also partially similar with the work of Chen et al. [16].Chen et al. [16] demonstrated the nature of the surface driven microfluidic flow in open microchannels.They have defined the channel aspect ratio as height to width ratio of the microchannels.They fabricated the microchannels by SU8 layer on silicon substrate.They used IPA (Isopropyl alcohol) as the working fluid as this has more wettability than DI water.They have found that the meniscus velocity of IPA in the microchannels is lower for lower aspect ratio of the microchannels.We have demonstrated the microfluidic flow behavior in closed microchannels.The channel aspect ratio varied from 0.04 to 0.012 just after the neck region.We have used dyed water as working fluid in all microfluidic flows.We also experimentally observed that the air-water meniscus velocity attained a peak in the channel aspect ratio transition region and after attending the peak, velocity reduces with the decreasing channel aspect ratio (Fig. 4).

IV. CONCLUSIONS
The air-water interface motion was highly laminar in photoresist microfluidic channels for surface-driven flow.http://www.sssj.org/ejssnt(J-Stage: http://www.jstage.jst.go.jp/browse/ejssnt/) e-Journal of Surface Science and Nanotechnology Volume 7 (2009) This investigation indicated that the square pillars on the bottom surface of the microchannels acted as surface roughness elements.In the case of fixed height and spacing between the square pillars, the microfluidic flow rate decreased with the increasing side length of the pillars due to enhancement of friction factor.The air-water meniscus velocity and transit time were measured.The Reynolds number reduced for larger side length of polyimide square pillars on the channel bottom surface.
FIG. 1: Surface profilometer scan on square polyimide pillars having side length of 200 µm at the top surface of the pillars.

FIG. 2 :
FIG. 2: The snapshot image of the time evolution of meniscus front in the microfluidic channel containing square pillars on the bottom surface of the channel having pillar side length of 200 µm at the top of the pillar with the PMMA, Acetone solution coated underneath the lid as adhesive material.

FIG. 3 :
FIG. 3: Variation of total transition time required for the interface to flow from the inlet.

FIG. 4 :
FIG.4: Variation of interface velocity with interface position from the inlet.

TABLE I :
Laminar flow variation with dimensions of square pillars.