Packaging of closed chamber PCR-chips for DNA
amplification
C.G.J. Schabmueller
Department of Electronics and Computer Science, Southampton University,
Southampton, UK
A.G.R. Evans
Department of Electronics and Computer Science, Southampton University,
Southampton, UK
A. Brunnschweiler
Department of Electronics and Computer Science, Southampton University,
Southampton, UK
G. Ensell
Department of Electronics and Computer Science, Southampton University,
Southampton, UK
The polymerase chain reaction (PCR) is a well described method for the selective identical replication of DNA molecules (Ehrlich, 1989). By an enzymatic in-vitro amplification process, the concentration of a DNA species is nearly doubled in a process, stepping through three different temperatures. In this way, the DNA concentration can be multiplied more than a million-fold by 20 to 30 cycles of temperature.
Since the first report of specific DNA amplification using the PCR in 1985, the number of different applications has grown steadily. It has become an important analytical method for the detection of diseases, pathogens or in forensics where the amount of sample is often small and direct detection would be impossible.
Two important innovations were responsible for automating PCR. First, a heat-stable DNA polymerase was isolated from the bacteriumThermus aquaticus, which inhabits hot springs. This enzyme, called the ``Taq DNA polymerase'', remains active despite repeated heating during many cycles of amplification. Second, DNA thermal cyclers have been designed in which a computer controls the repetitive temperature changes required for PCR.
The first micromachined PCR chip was presented in 1993 (Northrupet al., 1993). It used a relatively large reagent volume of 25-50l and it needed the use of silicone adhesive for the sealing of the chambers. The latest device developed by this group uses two bonded silicon wafers with matching, anisotropically etched grooves to form a tunnel into which a plastic reaction tube is inserted (Belgraderet al., 1998). The PCR chip presented by Daniel et al.(1998) has a reagent volume of 1l and is very fast (2.5 min for 30 cycles) but lacks the possibility of optical detection of the PCR products. Instead of a micro-chamber, the group of A. Manz (Koppet al., 1998) and Koehleret al. (1998) employed capillaries on a silicon chip to construct a ``continuous flow PCR'' device.
Given the capacity of PCR to synthesize millions of DNA copies, contamination of the sample reaction with either products of a previous reaction (product carryover) or with material from an exogenous source is a potential problem, particularly in those reactions initiated with only a few templates. Therefore, for industrial application the single shot PCR devices are still preferred.
The advantages of miniaturized PCR chips are many, such as rapid temperature cycling, smaller quantities of rather expensive reagents needed, low cost fabrication and portability, which is of great use for environmental screening as well as in the medical sector.
In this paper the design, fabrication and packaging of miniaturized reaction chambers etched into silicon are described. Advantages of the PCR chip are the small size of the chip (6 mm64 mm61.5 mm) and the sealing of the chamber with a Pyrex wafer using the anodic bonding method. The transparent surface of the Pyrex makes it possible to incorporate optical readout methods. The microchambers of 1l volume can be easily filled with the reagents using hypodermic needles and the fabrication process is simple. For fast thermocycling and control, platinum heaters and a temperature sensor have been integrated onto the chip. This ensures increased performance due to the closeness to the reagents and reduction of power and size. The packaging and the external fluidic connection are for the first time addressed in this paper.
Principle
To perform a PCR reaction, a small quantity of the target DNA is added to a buffered solution containing Taq DNA polymerase, short oligonucleotide primers, the four deoxynucleotide building blocks of DNA and the cofactor MgCl2. then the PCR mixture is taken through replication
cycles.
Figure 1 shows the principle of the DNA multiplication. The sample is heated to 94-968C to denature the target DNA (separate into single strands) ± (Figure (a) to (b)). The temperature is then lowered to 50-658C allowing the primers to anneal to their complementary sequences ± (c). The temperature is now raised to 728C, allowing the Taq DNA polymerase to attach at each priming site (where primers have annealed) ± (d) and extend (synthesize) a new DNA strand ± (e).
As amplification proceeds, the DNA sequence between the primers doubles after each cycle. Following 30 such cycles, a theoretical amplification factor of one billion is attained.
The research register for this journal is available at
http://www.mcbup.com/research_registers/mi.asp
The current issue and full text archive of this journal is available at
http://www.emerald-library.com
[ 11 ]
Microelectronics International 17/2 [2000] 11±14
#MCB University Press [ISSN 1356-5362]
Keywords
Packaging, Chips, Silicone, Reactors
Abstract
Reports the design, fabrication and packaging of a
micromachined silicon/Pyrex based chip for the polymerase chain reaction (PCR). The anodic bonding method is used for sealing the chambers of 1l volume with a Pyrex glass wafer. Platinum resistors on the back of the wafer are used as heaters and temperature sensors. The chip is externally cooled by forced air to achieve rapid temperature cycling. The transparency of the Pyrex makes it possible for using optical readout methods. The packaging is especially designed for easy handling, filling, power connection, temperature regulation and optical readout. The mass production of such silicon reactors could make single-shot throwaway devices economically viable.
The authors would like to thank Ken Frampton from the mechanical workshop of the electronics department at Southampton University for his help developing the crimping tool for the needles.
Design
The aim of the work was to design PCR chips with chamber volumes of 1l. The dimensions of the chambers were calculated, taking into consideration the angle of 54.71 degrees between the (111) and (100) crystal planes of the silicon wafer. The depth of the chamber was chosen as 400m, the length and the width of the upper chamber edge are therefore 1.86 mm for the 1l chamber. The square dimensions of the chambers have been chosen because of the easier optical detection where light is shone into the chamber and the fluorescent light is detected.
The chip size is determined by the chamber size, the space for the fluidic connection and the areas for anodic bonding. This gives a chip size of 6 mm64 mm. There are no thermal isolation structures implemented into the chip, as it will be cooled externally by forced air for fast thermo-cycling.
The fluidic connection involves a trench sawn into a 1 mm thick Pyrex wafer and a channel etched into the silicon leading from the sawing cut to the reaction chamber. The trench in the Pyrex is 0.8 mm wide and 0.8 mm deep where a hypodermic needle of 0.6 mm diameter can be glued in. The trenches are located on two sides of the chamber to keep the area above the chamber clear for optical detection and also for not giving additional volume to the chamber.
The connection channel etched into the silicon from the trench in the Pyrex to the PCR chamber is designed to give little pressure loss and at the same time have small dimensions. The triangle shaped channels are 142m wide and 100m deep. The pressure droppis calculated with (Richteret al., 1997):
pQCfrL
A2
whereQis the volume flow,Cfris the friction coefficient
(for KOH etched triangleCfr= 35.12),Lis the length of the
channel,is the viscosity (1.005*10±3Ns m±2for water) andAis the cross sectional area of the channel. For a volume flow of 100l/min and a channel length of 770m the pressure drop is 900 Pa.
The meeting point of the channel with the square PCR chamber creates a convex corner and because of the crystal dependent etch rates of KOH, compensation structures have to be implemented. Various structures have been designed (Mayeret al., 1990; Sandmaieret al., 1991). Some result in rectangular corners and others give larger openings for easy release of gas bubbles that might be trapped in the chamber. Photographs of the compensation structures before and after etching in KOH are shown in Figure 2.
The heaters and the temperature sensor, which are implemented on the chip, are made of Platinum of 200 nm thickness with an adhesion layer of 20 nm Titanium. The sensor is designed to have a resistance of 100 Ohm with four contact pads and is of industrial standard. It is located under the bottom of the chamber. The two heaters are placed directly under the chamber side walls and can be connected in parallel or in series. The electrode design is shown in Figure 3.
Fabrication
The micromachining is based on double sided processing of silicon and Pyrex wafers. For aligning chambers on one side and the electrodes on the other side of the silicon chip, double sided alignment structures are designed. This alignment is done by etching 1m deep structures into the front in order to align the second mask on the back with the help of infrared light shone through the silicon wafer.
The fabrication process is schematically pictured in Figure 4. After etching the double sided alignment structures onto the front and the back of the wafer it is covered with 600 nm of oxide and 160 nm of nitride which is structured by reactive ion etching to give the patterns of the chamber, the channels and the compensation structures (Figure 4a).
Then the wafer is immersed into KOH (30 per cent, 708C) to etch the reaction chambers and the connection channels. The (111) planes are used as an etch stop for the connection channels (V-groove), which are just 100m deep whereas the depth of the chamber is 400m. Thereafter, the nitride and oxide is stripped off with Orthophosporic acid and buffered Hydrofluoric acid respectively (Figure 4b).
After that a 40 nm thick thermally grown oxide is deposited onto the wafer on which a 20 nm thick film of Titanium and a 200 nm thick film of Platinum are deposited. Titanium is used as an adhesion layer. Both, the Titanium and the Platinum are patterned by ion beam milling to form the heater and sensor resistors (Figure 4c).
The next fabrication step involves the sawing of a Pyrex wafer. 800m wide and 800m deep trenches are sawn into the 1 mm thick Pyrex wafer to form the part of the chip where the hypodermic needles can be glued in for the fluidic connection.
Figure 1
Schematic of DNA multiplication
Figure 2
Compensation structures before and after etching in KOH
[ 12 ]
C.G.J. Schabmueller,
A.G.R. Evans, A. Brunnschweiler, G. Ensell, D.L. Leslie and M.A. Lee
Packaging of closed chamber PCR-chips for DNA amplification
The last step before cutting the wafer into single chips is the anodic bonding of the Pyrex wafer to the silicon wafer to seal the reaction chambers (Figure 4d). The bonding conditions are 4008C and 1,000 Volt.
Owing to the small size, several hundred chips can be fabricated on a single 4 inch wafer.
External fluidic connection and
packaging
For easy handling and for the power connection to the heaters and sensors the PCR chip is glued onto a standard 28 PIN DIL ceramic package. A few modifications have to be carried out to the package, some of which are seen in Figure 5.
As the chip sits upside down on the package because of the wire bonds and the external air cooling, a hole has to be drilled into the bottom of the package to give optical access to the chip. A second hole is drilled on the edge of the package so that the crimping tool can be placed close to the chip and allow the needles to be cut off close to the chip as well. The holes of 6 mm and 8 mm diameter respectively are cut using an ultrasonic drill. For quick drilling and exact location of the drilled holes a special drilling jig was fabricated.
Next, the PCR chip is glued onto the package. Quick set epoxy glue is used as adhesive. As the Pyrex glass is in contact with the package, it would be quite slow transferring the heat from the chip to the package, because of the low heat conductivity of Pyrex glass. Fast thermocycling is only possible using forced air cooling. Therefore no attention was given to a heat-conducting adhesive. A template is used for the repetitive exact placement of the PCR chip.
Gold wires of 25m diameter connect the heaters and the temperature sensor to the package. Three wires are used per contact pad because of the high current (ca. 50 mA) needed for the heaters.
In a final step, modified hypodermic needles of 0.6 mm diameter are glued into the trenches of the Pyrex. Quick set epoxy glue is used as it proved to be biocompatible. The PCR solution can be conveniently filled into the chip with a standard syringe. After filling, the needles can be sealed off using a crimping tool and afterwards, can be cut to reduce the thermal mass.
The so packaged PCR chip can be conveniently placed into a ``black box'', containing the power connection via a ZIF socket, the cooling arrangement and the detection optic. A photograph of a packaged PCR chip can be seen in Figure 5. This system has been tested and the PCR successfully carried out.
Conclusion
Micromachined reaction chambers for the thermal cycling of DNA by the polymerase chain reaction have been designed and fabricated. The miniaturization of bioanalytical reaction chambers brings to bear many benefits that would otherwise be difficult or impossible to achieve.
Obvious advantages are the small size, the low power consumption because of the heaters and sensors incorporated on the chip, the possibility of using optical Figure 3
Layout of the PCR chip
Figure 4
Micromachining of the PCR chip
Figure 5
Picture of a packaged PCR chip
[ 13 ]
C.G.J. Schabmueller,
A.G.R. Evans, A. Brunnschweiler, G. Ensell, D.L. Leslie and M.A. Lee
Packaging of closed chamber PCR-chips for DNA amplification
detection methods, also the easy filling and the portability of the device.
For easy handling, filling, power connection, temperature regulation and optical readout, a standard 28 PIN DIL package has been used with a few modifications.
During first test, PCR was successfully carried out. This shows that the design is working and that the materials used are biocompatible.
Nowadays the PCR is the most common method for DNA amplification in many different applications areas such as medical, forensic and environmental screening. Therefore, fast and small single shot devices, which can be taken into the field offer an enormous market potential.
References
Belgrader, P.et al.(1998), ``Multi-chamber, real time DNA analysis in the field using microfabricated silicon chambers'',ASME, DSC-Vol. 66, pp. 279-83.
Daniel, J.H., Iqbal, S., Millington, R.B., Moore, D.F., Lowe, C.R., Leslie, D.L., Lee, M.A. and Pearce, M.J. (1998), ``Silicon microchambers for DNA amplification'',Sensors and Actuators A, Vol. 71 No. 1-2, pp. 81-8.
Erlich, H.A. (Ed.) (1989),PCR Technology ± Principles and Applications for DNA Amplification, Stockton Press, New York, NY.
Koehler, J.M., Dillner, U., Mokansky, A., Poser, S. and Schulz, T. (1998), ``Micro channel reactors for fast thermocycling'', Proc. Process Miniaturization: 2nd International
Conference on Microreaction Technology, New Orleans, LA, pp. 241-7.
Kopp, M., Mello, A. and Manz, A. (1998),SCIENCE, Vol. 280. Mayer, G.K., Offereins, H.L., Sandmaier, H. and Kuehl, K. (1990),
``Fabrication of non-underetched convex corners in anisotropic etching of (100)-silicon in aqueous KOH with respect to novel micromechanic elements'',J. Electrochem. Soc., Vol. 137 No. 12, pp. 3947-51.
Northrup, M.A., Ching, M.T., White, R.M. and Watson, R.T. (1993), ``DNA amplification with a microfabricated reaction chamber'',Proc. Transducers '93, Chicago, IL,
pp. 924-6.
Richter, M., Woias, P. and Weiss, D. (1997), ``Microchannels for applications in liquid dosing and flow-rate measurement'', Sensors and Actuators A, Vol. 62, pp. 480-3.
Sandmaier, H., Offereins, H.L., Kuehl, K. and Lang, W. (1991), ``Corner compensation techniques in anisotropic etching of (100)-silicon using aqueous KOH'',Proc. Transducers, pp. 456-9.
[ 14 ]
C.G.J. Schabmueller,
A.G.R. Evans, A. Brunnschweiler, G. Ensell, D.L. Leslie and M.A. Lee
Packaging of closed chamber PCR-chips for DNA amplification
