Here is the abstract you requested from the IMAPS_2016 technical program page. This is the original abstract submitted by the author. Any changes to the technical content of the final manuscript published by IMAPS or the presentation that is given during the event is done by the author, not IMAPS.
|Silicon based Cell Sorting Device: Fabrication, Characterization and Applications|
|Keywords: Silicon microfabrication, micro-bubbles, Collective bonding|
|In this paper we describe the fabrication process for a microfluidics device targeting cell sorting applications. Currently there are numerous commercial products that are available for cell sorting they include Sony SH800S Cell Sorter and BD FACSAria sorter to name a few, however they are quite bulky and expensive and at the same time quite complex to operate. Most cell sorting techniques are based on Fluorescence activated cell sorting (FACS). FACS is an important diagnostic tool in microbiology to separate different type of cells at a very high throughput. In this paper we report on wafer level fabrication technique that will allows for a small form factor device targeting cell sorting. The aim is to develop a system that will move away from centralized facilities and will be available as a point-of-care tool. It will require small sample volumes and more user-friendly than current systems. The basic principle of our device is generation of heat pulse on a heater element that will create a bubble jet flow enabling the sorting. The device fabrication includes: processing of micro metal heaters, definition of polymer microfluidic channels and collective die-to-wafer bonding of glass substrate onto the polymer channels. In our previous paper we report on the fabrication of cell sorter based on Aluminium heaters. However, one of the drawback of the this device was poor reliability due to low melting temperature on Al. The studies showed that during repeated heating cycles the temperature of heater can reach up to 550C. In the current study we report on the improved version where heaters are made of tungsten The passivation stack on top of heater was changed from silicon oxide to silicon nitride and silicon carbide to improve the yield by reducing the thermal expansion mismatch. Tungsten is too resistive to be used a fanout layer for the wire-bonding. For this purpose the passivation was open and Al interconnects the W heater with the bondpads. In the next step, a polymer (PA from JSR ) is coated on to the wafer. The polymer is a negative tone photo-patternable adhesive. It not only acts as bonding material but also defines the microfluidics channels. It has excellent resolution, has low curing temperature and has straight patterning profile essential for microfluidics channels. It can be used for die to die or die to wafer bonding of Si onto glass. In the current device the minimum channels are 30µm wide and thickness of the polymer is also 30µm. We also report on the multichannel sorting using 5 different channels. The last step of the fabrication process is the bonding of glass die onto the si wafer. The glass is pre-punched allowing the polymer microfluidics channels connection to the outside testing equipment. The bonding is done in two step, firstly individual glass dies are bonded onto Si with a pick and place tool. Secondly, after populating the wafer collective bonding under force and temperature is performed. We report on the initial characterization of the cell sorting targeting sorting rate and sorting yield. We have achieved sorting rate of 5,000 cells/s and yield of 90% is obtained in initial investigations.|