Not applicable.
Not applicable.
The invention relates to a process for manufacturing a microfluidic device with buried channels, devices with buried channels, and uses thereof.
In general, chemical microreactors are provided with a microfluidic circuit, including microchannels. In the most advanced microfluidic devices the microchannels are buried in a substrate and/or in an epitaxial layer of a semiconductor chip. Substances to be processed, which are dispersed in a fluid medium, are supplied to one or more inlet reservoirs of the microfluidic circuit and are moved there through. Chemical reactions take place along the microfluidic circuit.
As is known, microfluidic devices may be exploited in a number of applications, and are particularly suited to be used as chemical microreactors. Thanks to the design flexibility allowed by semiconductor micromachining techniques, devices have been made that are capable of carrying out individual processing steps or even an entire chemical process.
For example, microfluidic devices are widely employed in biochemical processes, such as nucleic acid analysis. Such microreactors may also be called “Labs-On-Chip.” The discussion herein is simplified by focusing on nucleic acid analysis as an example of a biological molecule that can be analyzed using the various devices of the invention. However, the various devices may be used for any chemical or biological test, although typically molecule purification is substituted for amplification and detection methods vary according to the molecule being detected. For example, another common diagnostic involves the detection of a specific protein by binding to its antibody or by a specific enzymatic reaction. Lipids, carbohydrates, drugs and small molecules from biological fluids are processed in similar ways.
DNA amplification involves a series of enzyme-mediated reactions resulting in identical copies of the target nucleic acid. In particular, Polymerase Chain Reaction (PCR) is a cyclical process where the number of DNA molecules substantially doubles at every iteration, starting from a mixture comprising target DNA, enzymes (typically a DNA polymerase such as TAQ), primers, the four dNTPs, cofactor, and buffer.
During a cycle, double stranded DNA is first separated into single strands (denatured). Then the primers hybridize to their complementary sequences on either side of the target sequence. Finally, DNA polymerase extends each primer, by adding nucleotides that are complementary to the target strand. This doubles the DNA content and the cycle is repeated until sufficient DNA has been synthesized. RNA amplification is similar, but is typically preceded by copying the RNA into DNA.
Although PCR allows the production of millions of copies of the target sequence in few hours, in many cases its efficiency and speed might be improved by increasing the concentration of the reagents. Similarly, end-point detection of amplified DNA (amplicons) by hybridization is highly concentration dependant.
As already mentioned, in the most advanced microfluidic devices the channels are “buried” in a substrate and/or in an epitaxial layer of a semiconductor chip. However, processes for manufacturing microfluidic devices with buried channels are quite complicated. In particular, several steps are required once the buried channels have been completed and alignment of subsequent masks is often critical. Usually, additional steps are required to reveal the alignment signs of the wafer being processed, which would otherwise be hidden.
A known technique is described in “PROCEEDINGS OF THE IEEE,” Vol. 86, No. 8, August 1998, page 1632, and essentially envisages the creation of a cavity or air gap by means of anisotropic chemical etches made using potassium hydroxide (KOH), tetramethyl ammonium hydroxide (TMAH), etc., and employing a sacrificial polycrystalline-silicon layer.
This technique is schematically illustrated in
As used herein “buried channel” is defined as a channel or chamber that is buried inside of a single monolithic support, as opposed to a channel or chamber that is made by welding or otherwise bonding two supports with a channel or two half channels together.
According to one embodiment of the invention, a process for manufacturing a microfluidic device with buried channels is provided, as well as devices made by such process, and the various uses for such devices.
Generally speaking, true buried channels are made by antisotropically etching a substrate using a holed mask with apertures whose sides form an angle of 45°±1° with respect to the “flat” of the wafer. The apertures are arranged so that the longitudinal axes A of the buried channels are perpendicular to the flat of the wafer. Hence, each buried channel has a trapezoidal longitudinal section and a triangular cross section when etched to completion, or a trapezoidal cross section when etching is terminated short of completion.
Next, as a dielectric diaphragm is formed above each channel to close it, and heating elements are formed directly on said dielectric diaphragm. The dielectric diaphragm is made by depositing a coating film of a semiconductor material, partially occluding said apertures, thermally oxidizing said coating film, thereby narrowing said openings, and depositing a closing layer of a dielectric material, to completely close the openings.
The method thus avoids the use of epitaxial or pseudo-epitaxial layers and the need for a second mask, making it simpler and easier to fabricate a variety of devices.
For a better understanding of the various embodiments of the invention, preferred embodiments thereof are now described, merely to provide non-limiting examples, with reference to the attached drawings, in which:
a-1c show cross sections of a semiconductor material wafer in successive steps of a known forming process.
a-3d show cross sections of the wafer of
a and 6b show cross sections of the wafer of
As is known, a crystal of a semiconductor wafer has a number of crystallographic planes, among them <110>, <100>, <111>. As shown in
Alternatively, some semiconductor wafers do not contain the flat 15. Instead, they use other methods for identifying the crystallographic orientation of a plane. Thus, instead of using the flat of the wafer 15, some other method may be used to ensure that the orientation of the lattice structure is at the desired angle, relative to the selected plane.
For forming the cavity 20, according to what is illustrated in
Next, using a resist mask (not shown), dry etching is carried out on the uncovered portions of the silicon-nitride layer 14 and the silicon-dioxide layer 12, and the resist mask is then removed. In this way, the portions of the silicon-nitride layer 14 and the silicon-dioxide layer 12 that have remained after the dry etching form the holed mask 16 as shown in
As is illustrated in detail in
Other mask configurations and angles may be used when the flat of the wafer, or other indicia, is not aligned with the <110>plane. For example, the angle may be between 30° to 60° for other orientations. In general, the angle range depends on the crystallographic orientation of the wafer relative to the mask.
Using the holed mask 16, the substrate 11 is then anisotropically etched under time control in tetramethyl ammonium hydroxide (TMAH), thus forming the cavity 20, which substantially has the shape of an isosceles trapezium turned upside down and a uniform depth of between 50 μm and 100 μm (
In particular, the shape of an upside-down isosceles trapezium of the cavity 20 is obtained due to the combination of the following factors: execution of an anisotropic etch; use of a holed mask 16; and orientation at 45°±1° of the openings 18 with respect to the flat of the wafer 10. Also, length of etching time controls the bottom shape because a short etch time will lead to a flat bottom cavity, but if desired the etch can be continued until a triangle shaped cavity in cross section is achieved (see e.g.,
In fact, with the particular combination described above, the individual etches having their origin from the openings 18 of the holed mask 16 are performed on particular crystallographic planes of the silicon which enable the individual etches to “join up” laterally to one another, thus causing removal of the silicon not only in the vertical direction (i.e., in the direction of the depth of the substrate 11), but also in the horizontal direction (width/length), thus leading to the formation of the cavity 20 having the shape shown in
If, instead, the mask were oriented such that the openings 18 of the holed mask 16 had sides parallel or orthogonal to the flat of the wafer 10, the individual etches having their origin from the opening 18 of the holed mask 16 would be performed on crystallographic planes of the silicon that would not enable the individual etches to “join up” laterally to one another, thus leading to the formation of a set of cavities, equal in number to the openings 18 of the holed mask 16, separate from one another, and each having a cross section shaped like an upside-down triangle, of the same type as that shown in
One factor in determining the configuration and the angle of orientation of the lattice structure is that as the etch progresses in the substrate underneath the lattice structure from one opening it must eventually meet up with another opening, as can be observed in
The use of TMAH for carrying out anisotropic etching of the substrate 11 is particularly advantageous in combination with the structure of the holed mask 16 described above for leading to the formation of the cavity 20 having the shape illustrated in
With reference again to
The closing layer 24 is preferably formed of the same material as the coating layer 22, as part of a continuation of the same step such as CVD of TEOS. Namely, as the TEOS layer is formed on the individual side walls of the mask 17. As the coating layers build up, the deposited material between one mask portion 17 and another mask portion 17 will bridge over, so as to provide a complete block and provide for the formation of a top wall or dielectric diaphragm 26. A suspended structure, such as an inductor or a resistor can then be made, in a way in itself known and not illustrated.
Forming cavities as above described does not entail the deposition of a special sacrificial polycrystalline-silicon layer. Thus, the fabrication process is simpler and more economical due to the reduction in the number of the steps required, and in particular to the elimination of the mask necessary for the definition of the sacrificial polycrystalline-silicon layer.
The process described also enables the fabrication of a cavity 20 having a uniform depth beneath the dielectric diaphragm 26. In contrast, the prior art techniques shown in
In addition, the present process can be employed for the formation of cavities having, in plan view, any shape whatsoever, and even elongated cavities defining true buried channels, as shown in
The holed mask used in the process could also present a different pattern of the openings. For instance, it is possible to use the pattern shown in
In addition, the openings 18′ are arranged in parallel rows, and the openings 18′ belonging to adjacent rows are staggered with respect to one another.
Furthermore, the openings 18′ could present a shape slightly different from that illustrated in
The same process can be used to make buried channels connected with the outside world at communication openings, for example elongated channels having two opposite ends and being connected via communication openings set at the ends of the channels themselves. In this case, the openings 18, 18′ of the holed mask 16, 16′ (see
Next, as shown in
A variant of the above described process may be advantageously exploited in manufacturing microfluidic devices including buried channels, such as microreactors for nucleic acid analysis. An example of application of the process to the production of a microreactor will be now described, with reference to
Initially, a plurality of parallel buried channels 50 (e.g. twelve) are formed in a substrate 11″ of a semiconductor wafer 10,″ wherein nucleic acid amplification reaction, such as PCR (Polymerase Chain Reaction), is to be carried out. The buried channels 50 are first etched using a holed mask 16,″ having squared apertures 18,″ sides whereof form an angle of 45°±1° with respect to the flat of the wafer 10.″ The apertures 18″ are moreover arranged such that longitudinal axes A of the buried channels 50 are perpendicular to the flat of the wafer 10.″ Hence, each buried channel 50 has a trapezoidal longitudinal section (
Then,
Heaters 58 and temperature sensors 60 are subsequently formed directly on the closing layer 54 and the diaphragms 56, across the buried channels 50. Moreover, an array 61 of electrodes 62 is formed on the closing layer 54, adjacent to longitudinal ends of the buried channels 50.
In one embodiment, the heaters 58, the temperature sensors 60 and the electrodes 62 are made from a metal layer (not shown), e.g. Al, together with connection lines (not shown). In another embodiment (not shown), the heaters 58 are made of polycrystalline silicon. In this case, a polycrystalline silicon layer is first deposited and delineated and connection lines are subsequently formed from a metal layer, together with the electrodes 62.
Diaphragms 56 with a thickness of 2-5 μm, preferably 3 μm, provide sufficient mechanical strength to hold heaters 58 across buried channels 50 having a cross dimension of 200 μm without any substantial risk of failure.
After depositing and photo-lithographically defining a resist layer 63, the diaphragms 56 are etched for opening inlets 64 and outlets 65 at first and second ends of the buried channels 50 (second ends of the buried channels 50 are adjacent to the array 61 of electrodes 62).
Then, a protective layer 66, e.g. of dry resist, is deposited and covers the heaters 58, the temperature sensors 60 and the electrodes 62, as shown in
With reference to
After sample preparation for extracting DNA from nucleated cells, a biological sample is introduced in the inlet reservoirs 70 and advanced to the buried channels 50 by applying a pressure gradient in a known manner. Once the buried channels 50 have been filled, an amplification reaction (PCR) is carried out by cyclically delivering controlled amounts of thermal energy through the heaters 58. Then, the sample is pushed toward the detection chamber for hybridization of the DNA probes and detection.
In another embodiment, shown in
Forming the heaters 58 and the temperature sensors 60 directly on the closing layer 54 and the diaphragms 56 brings about several advantages.
First, all the process steps may be carried out without almost any alignment problems. In fact, conventional alignment signs on the top surface of the wafer 10″ always remain visible, since they are only covered by the holed mask 16.″ Because of its optical properties, the holed mask 16″ does not hide the underlying structures and, in particular, the alignment signs.
In contrast, growing a pseudo-epitaxial layer from a polycrystalline seed layer, in order to strengthen the diaphragms 56, requires some alignment measures to be taken, because polycrystalline silicon hides the alignment signs. Thus, the polycrystalline seed layer and the holed mask 16″ should be removed from above the alignment signs, so that a monocrystalline epitaxial layer may be grown thereon directly from the substrate 11.″ Accordingly, an additional mask for selectively removing polycrystalline seed layer and the holed mask 16″ would be required.
Microfluidic motion is improved as well, because biological samples are prevented from directly contacting silicon surface, which is hydrophobic. In contrast, inlet and outlet passages formed through a pseudo-epitaxial polycrystalline layer cannot be passivated by thermal oxidation because the heaters and temperature sensors would be destroyed (either oxidated or melted, depending on the material).
In the second place, opening the inlets 64 and the outlets 65 is simple, because only the diaphragms 56 are to be etched. Moreover, the resist layer 63 may be quite thin, because, although it is thinned during the etch for opening the inlets and the outlets, the etch time is short. For example, the structural layer may have a thickness of 2 μm instead of 7 μm.
Finally, it is clear that numerous modifications and variations can be made to the process described and illustrated herein, without thereby departing from the sphere of protection of one embodiment of the invention, as defined in the attached claims.
For example, the microreactor may comprise one or more of various components such as an injection port, reagent tank, dielectrophoresis cell, capillary electrophoresis channel, chambers or channels for various treatments, micropump, valve, heater, cooler, temperature sensor, detection chamber, detector sensor, power source, controls, display, and the like. In particular, the number and arrangement of these components and their connecting components depends upon the type of treatment to which the specimen fluid is to be subjected. These various components may be integral to the various devices described herein, or may be provided by a mother device on which a disposable microfluidic device is docked. For example, in many preferred embodiments the power source, controls and display are housed on a separate mother device.
Number | Date | Country | Kind |
---|---|---|---|
EP 00830148.3 | Feb 2000 | EP | regional |
This application is a continuation-in-part of U.S. Ser. No. 10/712,211 filed Nov. 12, 2003 and published Jun. 3, 2004 as US20040106290; which is a divisional of U.S. Ser. No. 09/797,206 filed Feb. 27, 2001 and issued Feb. 17, 2004 as U.S. Pat. No. 6,693,039; and claims the benefit of EP Application No. 00830148.3 filed Feb. 29, 2000 and issued Sep. 5, 2001 as EP 1130631. Each application is incorporated by reference in its entirety.
Number | Date | Country | |
---|---|---|---|
Parent | 09797206 | Feb 2001 | US |
Child | 10712211 | Nov 2003 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 10712211 | Nov 2003 | US |
Child | 11191325 | Jul 2005 | US |