CHAPTER 1
FABRICATING A THREE-DIMENSIONAL CHANNEL FOR MICRO-FLUIDIC DEVICES BY LASER ABLATION
Yoshikazu Yoshida, Tsutomu Neichi, Retsu Tahara, Jun Yamada, Hiroyuki Yamada and Nobuyuki Terada
Toyo University, 2100 Kujirai, Kawagoe, Saitama 350-8585 Yamanashi Pref. Industrial Technology Center, 2094 Kofu, Yamanashi 400-0055, Japan University of Yamanashi, 1110 Tamaho, Nakakoma, Yamanashi 409-3898, Japan
Abstract
This paper describes the fabrication in resin of micro-channels for micro-fluidic devices such as the µTAS (Micro Total Analysis System) by UV laser ablation process. A number of heat-hardening resin-films are layered on a soda glass. A laser fabricates a part of the channel on each film for every lamination. Then three-dimensional (3-D) confluence channels are fabricated. The fabricated channels are 45-180 µm in depth and 50-300 µm in width. The through holes are made in the laminate film with a laser. An inlet pipe for a micro-pump is inserted into the hole.
Keywords: µTAS, UV laser, lamination, resin-films, blood
1. Introduction
Recently, in various fields the necessity for small and highly sensitive micro-fluidic analysis systems has increase. Therefore a µTAS has received considerable attention. The µTAS is the size of a card, and has miniaturized channels, detectors, and other elements for fluidic analysis. The advantages of this system are the reduced need of fluidic samples, reagents, and hours of detection. The size of the fluidic analysis elements on the µTAS is a few score micrometers. There are many fabrication methods of micro-channels through semiconductor technology, plastic molding, and laser fabrication of resins. The laser fabrication method has recently been receiving much attention. The advantages of this method are: one stroke fabrication of grooves for channels, an easy change of groove patterns, and 3-D fabrication to allow grooves with slopes and differences in levels. We have been proposing the method which uses silicon or quartz as the substrate part of µTAS, build the micro working parts and electrode in advance onto the substrate, then create the flow path and cistern on the resin part formed on the substrate, An ultraviolet pulse laser was used to form such items as the flow path. A number of heat-hardening resin-films were layered on a soda glass. A laser fabricated a part of the channel on each film for every lamination, and then a 3-D micro-channel structure was fabricated. Two types of flow path, a plane and an overpass, are fabricated.
2. Experiment equipment
The substrate is soda glass laminated by heat-hardening resin-film. This film is made of two films, one of 25µm thick polyimide and the other 20µm thick epoxy. Channels are fabricated by a pulse Nd:YAG laser system (Brilliant; Quantel) and a KrF excimer laser system (LPX220; Lambda Physik, AG). For the experiment condition, the YAG has a wavelength of 266nm, pulse energy of 3.1mJ, pulse width of 4.3nsec, and repetition rate of 10Hz. The laser beam is fixed, and the substrate is moved in the XY stage. This stage has a positional bi-directional repeatability of [+ or -] 5pm. The excimer has a wavelength of 248nm, output energy of 8-80W, maximum pulse energy of 450mJ, pulse width of 10-20nsec, and repetition rate of 25-200Hz. A mask is used to shape the laser beam into a square shape to allow fabrication with smooth wall surfaces at low overlap rate conditions. The laser beam is focused to the width of a groove.
3. Results and discussion
3.1 Three-pronged channel
Combining of laminar flows in a micro-channel makes possible the study of blood cell analysis. Figure 1 shows an optical photomicrograph of three-pronged grooves without cover film fabricated by the excimer laser. Channels have a width of 50pm and a depth of 45µm in three-pronged parts, and a width of 150µm in a confluence part. Blood is injected at a low flow rate between two rapidly flowing streams of physiological salt solution. The width of the stream of blood can be controlled by the height difference between a blood reservoir and solution reservoir, that is potential energy. The width narrows as it climbs to the speed of the neighboring streams. The width decreases with the increasing height difference. The cells velocity increases with the increasing height difference. In this experiment, the channel substrate is placed horizontally on a microscope stand, and the reservoir made from a connector between a syringe and a needle is connected to the channel inlet with Silicon tube. This tube has an external diameter of 1mm and an inside diameter of 0.3mm. Figure 2 shows a focused picture of blood when the blood reservoir is 200mm high and the solution is 400mm high from the channel. The cells velocity is almost 15.5mm/s. It is almost 9mm/s when the height difference is nearly zero. As shown in Fig.2 (b), blood cells can be measured individually.
3.2 Three-dimensional channels
Figure 3 shows the optical photomicrograph of a channel made on the second film by the YAG laser. The laser scanning distance is 150µm. The second film is peeled off from the first one by the scanning. The film placed between the scanning is removed from the substrate. The space caused by removal of the film is used as a channel space. The film peelings on the channel side are removed with the following laminating process.
Figure 4 shows the production process of the steric mixture flow path used to branch. There are 3-D pattern diagrams and optical photomicrographs each time a lamination is done. First, the channel element of the first layer is made for the film on the glass by the laser (Fig.4 (a)). After that, the second film is laminated on the first one, and the channel element is made on the film (b). These processes are iterated several times, and the 3-D confluence channel is fabricated. The channel is closed because the groove is treated with laminate processing, and liquid can't enter into the groove. Therefore it is necessary to make a perforated hole in the flow path to insert liquid inside, and connect the tube. The diameter of the hole is 150pm, and formed on the laminate film by laser drilling. Deionized water is injected into the channels with a microinjection pump. The flow rate is 5µL/min. There is no damage to the channel.
4. Conclusions
(1) The groove with a width from several dozen pm to several hundred pm is created on the resin layer without any damage on the substrate by ablation processing with an ultraviolet laser.
(2) A heat-hardening resin film can be used to maintain 100% of the channel space for fluid flow.
(3) Fresh blood flows easily through the channels.
CHAPTER 2
PHOTOPOLYMERIZED POLY (ETHYLENE) GLYCOL DIACRYLATE (PEGDA) MICROFLUIDIC DEVICES
Amy Butterworth, Maria del Carmen Lopez Garcia and David Beebe Department of Biomedical Engineering, University of Wisconsin 1550 Engineering Drive, 53704 Madison WI, USA
Abstract
As microfluidic applications in cell biology move beyond diagnostic assays to long term culture and production, alternative materials will be needed. Poly (ethylene) glycol (PEG) has been widely utilized as a biocompatible polymer due to its hydrophilicity and non-fouling behavior. Diacrylated, PEG can be photopolymerized using the microfluidic tectonics platform (µFT) and provides a more biocompatible alternative to previous polymers used. The ability of this monomer to be polymerized and patterned into channels for micro-cell culture was evaluated. Also, the biocompatibility of the polymer was assessed using FT-IR and cell interaction studies with the unpolymerized components.
Keywords: Microfluidic tectonics, biocompatibility, photopolymerization, poly (ethylene) glycol (PEG)
1. Introduction
Poly (ethylene) glycol (PEG) has been widely utilized as a biocompatible polymer due to its hydrophilicity and non-fouling behavior. PEG resists protein absorption and has been used as a coating or as a polymer substrate to prevent or control cell adhesion and adsorption of proteins for over a decade. PEG has been used previously in bioMEMS-related technologies as a coating or as a co-monomer for purposes such as polymerizing cells in gels. Diacrylated, PEG can be photopolymerized using the microfluidic tectonics platform (µFT) and provides a more biocompatible alternative to previous polymers used. The ability to incorporate PEG as a construction material for microfluidic systems will allow the unique properties of PEG to be exploited for a variety of cell-based experiments. Examples include using in-situ polymerized porous PEG gels as selective diffusional barriers to replace media changes during cell culture, or copolymerizing with a hydrolytically degradable monomer for controlled release of biomolecules of interest.
2. Fabrication and Biocompatibility Analysis
The biocompatibility of this polymer will be partially dependent on the complete polymerization of the monomer while using a minimal concentration of photoinitiator. The typical concentration of photoinitiator used in these experiments was 0.05 wt%, although lower percentages (below 0.01 wt%) can be polymerized but exhibit more swelling. To verify the degree of polymerization, FTIR studies were done, comparing the spectra of the polymer with 0.1 wt% photoinitiator (Fig. la) to that with 0.05 wt% (Fig. 1b). The polymerization of diacrylates reduces the magnitude of the carbon-carbon double bond peak (shown in Fig. 1) and is expected to decrease with increasing photoinitiator concentration as shown. FTIR measurements allow one to find a balance between concentration of photoinitiator and degree of polymerization that minimizes the cytotoxicity of the devices, while maintaining good patterning capabilities. After UV sterilization before use in cell culture, this peak decreased slightly. Straight channels were patterned to test the patterning capabilities, with widths ranging from 175pm to 1,000pm in 250 pm high devices (Fig. 2). Good resolution of less than 10pm was achieved which is comparable to that achieved with poly (IBA). A valve mask currently in use for creating the substrate for a hydrogel actuated valve was also patterned in PEGdA with relative ease [5]. These two materials show similar capabilities although the PEGdA devices produce more rounded features.
Since our intended use of the PEG channels is long term (weeks) cell culture, the long term capability of the material was tested to ensure no failure occurred due to the PEGdA swelling. Swelling of the PEGdA material did cause device failure more frequently as the percentage of photo initiator decreased (from 0.05 wt% to 0.01 wt%) and exposure intensities decreased (from 20 mW/cm2 to 10 mW/cm2). The molecular weight of the PEGdA was reduced from 575 MW to 258 MW, which showed significantly reduced swelling and produced devices which could be incubated at 37°C without immediate failure. Multiple photoinitiators 4-(2-hydroxyethoxy)phenyl-(2 -hydroxy-2-propyl) ketone (Irgacure 2959, Ciba, Inc.) and biacylphosphine oxide (BAPO, Iragure 819, Ciba, Inc.) were also studied. The former has commonly been used for photopolymerization of cells in gels and has shown to be more biocompatible than many photoinitiators. The latter has a higher efficiency (the absorption band extends to 400nm) at the wavelengths of exposure, so lower concentrations were needed. Successful polymerization was demonstrated with both compounds, although BAPO-initiated devices proved to be more resistant to swelling most likely due to the faster reaction kinetics causing denser gels.
4. Biocompatibility
Devices were created with the optimized prepolymer mixtures and exposures with 1,000µm straight channels and incubated with DPBS at 37°C. When externally reinforced with adhesive, these devices are suitable for cell culture, surviving for more than one week without failure due to swelling of the polymer. The cellular response to the presence of the monomer and photoinitiator in the media was evaluated. A concentration of 10 µM PEGdA caused significant reduction in NMuMG cell adhesion, while 1µM did not prevent adhesion, although cell morphology was slightly different than the controls. Cells with PEGdA in the media that did attach to the surface remained rounded in colonies rather than spreading as expected for epithelial cells. Due to very low solubility of BAPO in the media, quantitative results were not obtained, although the presence of BAPO did cause cell death in media with the maximum soluble amount of BAPO. It is clear that further optimization of the polymerization technique and prepolymer mixture is needed to ensure minimal concentrations of prepolymer components remain after polymerization and washing in order to maximize biocompatibility.
CHAPTER 3
A FLOW-THROUGH SHEAR-TYPE MICROFILTER CHIP FOR SEPARATING PLASMA and VIRUS PARTICLES FROM WHOLE BLOOD
Levent Yobas, Ee-Ling Gui, Hongmiao Ji, Jing Li, Yu Chen, Wing-Cheong Hui, Siti Rafeah Binte Mohamed Rafe, Sanjay Swarup, Sek-Man Wong, Tit-Meng Lim Chew-Kiat Heng
Institute of Microelectronics, 11 Science Park Road, Singapore Science Park II, 117685, Singapore Nanyang Technological University, School of Materials Engineering, 639798, Singapore National University of Singapore, Department of Pediatrics, 119074, Singapore National University of Singapore, Department of Biological Sciences, 117543, Singapore
Abstract
A flow-through shear-type microfilter chip has been proposed for the purpose of separating plasma and virus particles from whole blood. The microfilter chips have been fabricated in three different design configurations by silicon micromachining and tested for their percent efficiency of separating plasma from diluted blood samples. One of the designs has been further demonstrated to be capable of isolating virus particles from a spiked sample of whole blood.
Keywords: Microfilter, shear filter, virus, plasma, sample preparation
1. Introduction
Recent epidemics such as Severe Acute Respiratory Syndrome (SARS) have highlighted the importance of an automated sample preparation for virus and pathogen detection. Detection of extracelullar viruses from blood usually requires separation of plasma or serum containing virus particles from cellular components. This is because hemoglobin from red blood cells (RBC) is known to inhibit nucleic acid amplification while nucleic acids in white blood cells (WBC) can contribute to background noise during detection. Typically, plasma is obtained from whole blood by a centrifugation step. Nevertheless, centrifugation is not amenable to automation. An alternative approach is filtering plasma based on size exclusion of cells. Most viruses are less than 1 µm while most RBC and WBC remain larger than 2µm.
2. Microfilter Chip
The proposed micro filter concept and structure are diagrammatically shown in Fig. 1. The chip contains a chamber etched about 65-pm deep into silicon by deep reactive ion etching and capped with a glass wafer by anodic bonding. Plasma can be collected through anisotropically-etched backside holes in silicon located at two diagonal comers. At the other comers, backside holes allow blood to flow in and out of the chip through a meander type channel defined by silicon pillars. As blood flows inside the channel, plasma can escape through narrow slits between pillars due to combined action of capillary forces and pressure gradient. Nominal gap between the pillars is about 1.6pm wide, which can retain most blood cells but allow passage of virus particles. The microfilter chips have been fabricated in three design configurations mainly differing in chip size and shape of the meander-type channel (Table I).
3. Experimental Results & Discussion
Fig. 2 shows on-chip collection of plasma escaping through the slits between pillars as the anticoagulant-treated whole blood flows through the meander-type channel. Anticoagulant-treated blood was pumped through the chips at 10µl/min and at different dilutions of phosphate buffered saline (PBS) solution. RBC counts in the blood pumped in (RBCblood) and the plasma collected (RBCplasma) were obtained by a hemocytometer. Table I shows volume of the collected plasma samples and percent efficiency of each microfilter chip (% EF) as calculated by:
EF = [1 - (RBCplasma/RBCblood)]x 100 (1)
As shown, chips based on any of the three designs had higher than 99% efficiency for the Blood:PBS ratio of 25:75. The efficiency deteriorated with an increase in the blood:PBS ratio but stayed above 90% for all three microfilter chips.
Further, experiments were conducted to test whether the plasma filtered by the microfilter chips can be used for detection of virus particles in blood. Anticoagulant-treated whole blood at a volume of 140µl was spiked with virus (Cymbidium Mosaic Virus) suspension in water at a volume of 70µl and concentration of 0.26µg/µl. Approximately, 180|il of the spiked blood was pumped through microfilter #1 at 10µl/min. The plasma filtrate was used for extraction of viral RNA via a commercial kit [3] and amplified by reverse transcription polymerase chain reaction (RT-PCR). The amplified products were separated by agarose gel electrophoresis and ethidium bromide-stained products were visualized on a UV transilluminator. As can be seen in Fig. 3, viral RNA from the plasma filtered by microfilter chip #1 could be amplified, demonstrating a successful substitute for the conventional centrifugation step.
