Lateral air cavities for microfluidic pumping with the use of acoustic energy
- First Online:
- Cite this article as:
- Tovar, A.R., Patel, M.V. & Lee, A.P. Microfluid Nanofluid (2011) 10: 1269. doi:10.1007/s10404-010-0758-1
- 1.3k Downloads
An acoustically activated micropump is fabricated and demonstrated using a single step lithography process and an off-chip acoustic energy source. Using angled lateral cavities with trapped air bubbles, acoustic energy is used to oscillate the liquid–air interface to create a fluidic driving force. The angled lateral cavity design allows for fluid rectification from the first-order pulsatile flow of the oscillating bubbles. The fluid rectification is achieved through the asymmetrical flow produced by the oscillating interface generating fluid flow away from the lateral cavity interface. Simulation and experimental results are used to develop a pumping mechanism that is capable of driving fluid at pressures of 350 Pa. This pumping system is then integrated into a stand-alone battery operated system to drive fluid from one chip to another.
KeywordsMicropumps Acoustic Simulations Microstreaming
Over the past 20 years, great strides have been made in developing a portable micro total analysis system using microfluidic technology. However, one major challenge has been developing an adequate cross-over mechanism between the macro and micro environments that exists when trying to introduce samples of μl and ml volumes into devices that manipulate one to three orders smaller. The main component that bridges these environments is the fluidic driving mechanism or pump. Since the advent of microfluidics, there have been great advances in this field (Laser and Santiago 2004) that have ranged from using pneumatic and peristaltic pumping (Unger et al. 2000; Chiu and Liu 2009), electrokinetic pumping (Zeng et al. 2001), magnetohydrodynamic pumping (Lemoff and Lee 2000), to using diodes embedded in channel walls (Chang et al. 2008), and using electrolysis to generate bubbles to pump blood (Chiu and Liu 2009). One major limitation that persists in the pumping mechanisms developed is they often require complex fabrication techniques, severely limiting their applications in portable, low-cost systems. In the case of portable diagnostic systems, a pumping mechanism that is simple to operate, fabricate, and compatible with biological samples is essential. Though many pumping technologies exist, an often used fluid transport mechanism in these systems is the use of capillary forces (Gervais and Delamarche 2009; Zimmermann et al. 2009) to achieve fluid delivery. This form of fluid transport is advantageous in that it (1) requires little to no additional fabrication steps, (2) is simple to operate, and (3) requires no additional form of energy applied to the system for operation. The main drawback to this form of fluid transport is the limited control of the fluid flow. Since fluid flow is achieved through surface energy present in the channel materials, it is difficult to start and stop flow on demand, adjust flow rate, and sequentially drive different reservoirs of fluid into a microfluidic system. To address these issues, the coupling of air bubbles with acoustic energy is investigated as a fluidic driving mechanism.
Traditionally, in the field of microfluidics, the presence of air bubbles or air pockets in microchannels is considered as an unwanted byproduct. However, recently there has been a growing trend in harnessing this byproduct for a variety of novel applications (Rife et al. 2000; Tsai and Lin 2002; Marmottant and Hilgenfeldt 2003, 2004; Dijkink et al. 2006; Marmottant et al. 2006; Hettiarachchi et al. 2007; Kao et al. 2007; Tho et al. 2007; Xu and Attinger 2007; Chung and Cho 2008; Ahmed et al. 2009a, b; Chung and Cho 2009; Tovar and Lee 2009). The first practical use of trapped air bubbles in a microfluidic device utilized acoustic energy to rapidly mix two fluids within a chamber (Liu et al. 2002) for the increase of DNA hybridization (Liu et al. 2003a, b). The combination of acoustic energy and trapped air bubbles within these microsystems allowed for the generation of localized microstreaming which produced rapid mixing characteristics within a platform that is predominately laminar. The use of this phenomenon was integrated into a cell trapping and lysing system (Marmottant and Hilgenfeldt 2003) and thereafter harnessed for particle transport (Marmottant and Hilgenfeldt 2004; Marmottant et al. 2006). Particle transport was achieved through the use of asymmetrical flow caused by obstructions near the liquid–air interface of trapped air bubbles within etched surface cavities (Marmottant et al. 2006). The use of acoustic microstreaming was further studied to determine how interface oscillations are related to voltage and frequency input (Xu and Attinger 2007). These observations were used to determine the optimal trapping geometry for a perpendicular-oriented cavity and that fast mixing could be achieved near the cavity interface (Xu and Attinger 2007). In 2006, Dijkink et al. (2006) propelled an “acoustic scallop” within a liquid reservoir using a one-sided sealed capillary tube and acoustic waves. The bubble’s oscillations produced jet-like propulsion as fluid was expelled from the tube normal to the interface when the bubble expanded and pulled in symmetrically when it retracted. More recently, Ryu et al. reported on a technique for micropumping using an acoustically excited oscillating bubble for implantable microfluidic devices (Ryu et al. 2010).
We recently reported on the first proof-of-concept of a novel acoustic pumping platform that exploits trapped air bubbles within angled lateral cavities off a main microchannel to address the shortcomings of traditional pumping technologies (Tovar and Lee 2009). Acoustic energy is supplied to the system via a piezoelectric transducer (PZT) located beneath the microfluidic chip that can be powered through a battery or USB powered circuit. The acoustic energy actuates the angled, lateral cavities producing a rectified, first-order pulsatile flow to drive fluid within the main microchannel. The system is initially primed using capillary forces or a slight negative pressure at the outlet to draw fluid in and trap air bubbles within the lateral cavities. The main advantages of this pumping mechanism are its ability to drive fluid using only an external acoustic energy source and its easy integration with existing microfluidic devices. Since the main driving mechanism is the liquid–air interface of the trapped air bubbles, fabrication is achieved through a one-step photolithography process with lateral cavities situated in the same plane as the microchannels. The fact that lateral cavity acoustic transducer (LCAT) pumps are dead-end side channels they can be implemented in a simple passive chip rendering it amenable to very low-cost manufacturing processes with potential for applications requiring disposable devices. In this article, we expand the acoustic pumping operational understanding and improve the pumping platform efficiency and operation through simulation and experimental results. We also implement a practical pumping system that is able to drive fluid from one chip to another demonstrating a real-world application.
2 Acoustic pumping: principle and operation
3 Design and fabrication
3.1 LCAT pump designs
The behavior of the acoustic pumping mechanism using the design in Fig. 1a only demonstrates recirculatory pumping operation. To determine the pumping behavior with an open inlet and outlet, a scaled-up device is designed and shown in Fig. 1b. The 80 cavity pair serpentine design is fabricated with the same dimensions as the recirculation design (channel height of 100 μm, main microchannel width of 500 μm, lateral cavity lengths of 500 μm, widths of 100 μm, and a cavity to cavity spacing of 100 μm). The device footprint is 9 × 12.3 mm with seven bends and eight straight sections with 10 cavity pairs per straight section. The inlet reservoir is fabricated using a 15 mm hole-punch and the outlet is fabricated using a 5 mm hole-punch. The total polydimethylsiloxane (PDMS) thickness is approximately 10 mm. An image of the lateral cavities fabricated within the PDMS device is shown in Fig. 1c.
3.2 Microfabrication of LCAT chip
The microfluidic acoustic pumping devices are fabricated using standard soft lithography techniques (Duffy et al. 1998). The recirculation and 80 cavity serpentine designs are fabricated out of PDMS as the main polymer microchannel carrier with a glass microscope cover slip as the base substrate. The PDMS (Sylgard 184, Dow Corning Corp., USA) microchannels are molded from a SU-8 50 (Microchem Corp., USA) photoresist layer which is UV exposed and developed on a 3″ Si wafer to produce the desired microfluidic pattern. A standard 10:1 ratio of PDMS monomer to curing agent is used for all microdevices tested in these experiments. The PDMS is cured at 60°C for at least 4 h and are bonded to microscope coverslips (150-μm thick) after a 2 min plasma activation of the surfaces. The inlet and outlet holes are punched in the PDMS before bonding using a 5-mm diameter hole-punch. Upon bonding, the devices are kept stored for at least 24 h at room temperature before experimental operation to allow the PDMS surface to return to a hydrophobic state.
4 Experimental setup
4.1 Operation of LCAT
4.2 Fluid flow velocity and pressure measurements
The characterization of the acoustic pumping devices is done through measuring the fluid flow velocity of the recirculation designs as the voltage amplitude applied to the PZT is increased. The fluid flow velocity is measured through the use of 2–6 μm polystyrene, non-fluorescent microbeads (F13838, Invitrogen, USA). The polystyrene beads are mixed with DI water with a final density of 3 × 105 beads/ml and are drawn into the recirculation design through the inlet while applying a slight negative pressure at the outlet. As the fluid is drawn into the channels, visual inspection is done to confirm that the channels are filled with the microbead/DI water mixture. The device is placed on the custom made PZT platform using a thin layer of ultrasound gel and is visualized through a Nikon Eclipse L150 upright microscope. A high speed camera is used to capture the fluid flow within the device upon activation of the PZT. The high speed cameras used in the experiments are a FASTCAM PCI (Photron USA, USA) that is capable of video capture up to 10,000 fps and a Phantom 7.3 (Vision Research, Inc., USA) capable of video capture up to 500,000 fps. Fluid flow velocity is measured through the tracking of multiple microbeads within the fluid flow over a specified period of time. The distance of each bead is measured through the stacking of images and the use of Image J, an image analysis program by NIH Image.
The pressure generation by the 80 cavity serpentine device is calculated through measuring the height difference of the liquid columns between the inlet and the outlet (Wang et al. 2009). A large reservoir is used as the inlet and the outlet is connected to tygon tubing extending perpendicular to the surface of the device. The device is primed with DI water and allowed to equalize so there is no residual flow before operation. The activation of the PZT driven acoustic pump drives the fluid up the tygon tubing until a maximum pressure head is reached. The PZT is activated at specified voltages for 30 min to allow the maximum height to be reached. A Rebel XSi DSLR with a 100 mm f/2.8 macro lens (Canon Inc., Japan) is used to record images throughout the experiment. The height difference is measured using Image J and a known dimension in the image for scale reference. The experimental setup can be seen in the Supplementary material 1.
5 Results and discussion
5.1 Pumping velocities and pressure output
5.2 Acoustic pumping flow paths
Figure 8 shows a time lapsed image of a simulated acoustic pumping device with 4 μm polystyrene beads spread throughout the main channel. The paths of the beads are visualized using the particle trajectory history tool in CFD-VIEW. As seen in the figure, the beads located near the liquid–air interface travel a much greater distance than the particles fore and aft of the cavity interface. The average distance travelled for each particle in the upstream and downstream channel is approximately 1 μm over 8.29 ms. This gives an average particle velocity of approximately 120 μm/s for a single cavity pair. When comparing this to experimental data, it can be seen that the simulated velocity is approximately 3.5 times larger. However, the current 3D model is only 11 mm in length, while the experimental design is approximately 30 mm in length in a recirculation system. Since for a given flow rate, pressure is directly proportional to fluidic resistance and the resistance is proportional to channel length, it can be estimated that the flow velocity is approximately three times smaller than the simulated value. The corrected velocity would be in good agreement with the experimental data.
5.3 Chip-to-chip pumping
Expanding on the observations made through both experimental and simulation studies, the 80 cavity pair serpentine design is used to drive fluid from one chip to another. This pumping demonstration shows positive pressure generation from an acoustic microstreaming operated device. The rectified flow allows for fluid to be pumped from the LCAT pumping region out through a 1 mm ID tygon tube connected to an external straight channel microfluidic chip. The chip is operated in the same manner as described in Sect. 4 with the DI water dyed red for visualization purposes. The video can be seen in the Supplementary material 4 and shows the pump being actively turned off and on various times throughout its operation.
6 Future works and improvements
This report has demonstrated and investigated the use of trapped air bubbles and acoustic energy as an effective fluidic driving mechanism. The simplistic design and ease of operation lends itself to a variety of microfluidic applications that would need an integrated and portable pumping source. One major hindrance in the development of microfluidic pumps is the ability to demonstrate its operation in a portable system while still maintaining simple fabrication. The demonstration of portability was done by developing a prototype acoustic driving platform that is powered by 4 AA batteries or a USB 2.0 port. Two 80 cavity pumps are fabricated and bonded to a glass coverslip which is placed on the PZT. Each pump has a reservoir of approximately 100 μl, which are filled with either a red or green dye solution. The fluid is initially drawn in with capillary forces or a slight negative pressure at the outlet for priming and the pumps are then activated using the PZT. The square-wave driving frequency is 35 kHz at an amplitude of 25 Vpp. As seen in the video (Supplementary material 5), each pump drives fluid to the outlet of the system and can be independently controlled. This prototype device demonstrates the use of the acoustic pumping platform and its extendibility into a portable system for serial delivery of reagents.
This article demonstrates a novel acoustically activated microfluidic pumping platform that is both simple to fabricate and operate. The pumping is generated through rectified flow from oscillating liquid–air interfaces within lateral cavities oriented 15° to the main microchannel. The device is fabricated using only a single lithography step and adds no additional fabrication processes when integrating into existing microfluidic systems. Lateral cavities of dimensions 100 μm wide by 500 μm long and a height of 100 μm are activated at a pumping frequency of 35 kHz and PZT voltages ranging from 0 to 25 Vpp. The pumping platform is characterized through determining a relationship between cavity pairs and voltage input and is predicted beyond 10 pairs to 80 cavity pairs. The maximum pumping pressure generated with the 80 cavity pair device is approximately 350 Pa at 25 Vpp. Simulation work is also done to gain a better understanding of the fluid behavior near the liquid–air interface and to determine if the model can be used as an aid in future designs and iterations. The liquid–air interface oscillations in the CFD model are matched through experimental observations and fluid flow velocity is found to be in good agreement to a single cavity pair design. To further demonstrate the practicality of the LCAT pumping mechanism, a two fluid pumping device is fabricated and used to show multi-fluid pumping from large reservoirs at the inlets to a single channel outlet. To extend the applications that could be made possible through the use of the LCAT platform, a portable acoustic generation device is designed that can be operated on either 4 AA batteries or a standard USB 2.0 port.
The authors acknowledge financial support from the DARPA S&T Program (HR0011-06-1-0050) through the Micro/nano Fluidics Fundamentals Focus (MF3) Center.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.