Biological Sensors Take Advantage of Silicon Processes

One mission of Purdue's Birck Nanotechnology Center (BNC, West Lafayette, Ind.) is to integrate micro and nanotechnologies with biotechnology to create novel solutions to important problems in the fields of biology and medicine. The 187,000 ft² center, completed in 2005, includes the Scifres Nanofabrication Laboratory, a 25,000 ft² cleanroom rated at 1, 10 and 100 microparticles per cubic foot; and electronic and biological materials characterization laboratories. Here, interdisciplinary teams from electrical and bioengineering, biosciences, food science and agriculture come together to share equipment and develop new concepts. One area of particular interest is the application of microelectronics and silicon processes to the development of chip-based diagnostic devices.

Unlike silicon processes, bottom-up fabrication is often used in nature as chemical and biological self-assembly processes that make up biological entities at the nanometer scale and beyond (Fig. 1 ). For example, viruses are on the order of 0.1 μm or smaller, and DNA have diameters of 2–3 nm. With chip geometries approaching 45 nm, "it is possible to make structures that are essentially on the same scale as biological elements, such as protein, viruses and DNA," said Rashid Bashir, professor of electrical and computer engineering and biomedical engineering, and director of the Laboratory of Integrated Biomedical Micro/Nanotechnology and Applications (LIBNA) at the BNC. One outcome is detectors that are orders of magnitude more sensitive than any available today. Moreover, these detectors can probe and detect even single molecules.

A further advantage of applying well-established silicon processes to these diagnostic sensors is the possibility of mass production and cost reductions. With E.coli and other bacteria making their way into water supplies and vegetables fields, there is an urgent need for point-of-use disposable sensors for environmental and food field testing, as well as personal point-of-care applications.

1. Advances in the silicon chip industry are making it possible to create detectors that are on the same scale as biological elements, such as protein, viruses and DNA.

Identifying bacteria

For identifying the presence of a particular bacterium, one of the distinguishing features of this project from other lab-on-a-chip projects is a true multidisciplinary approach that adds greater functionality and versatility, noted Bashir. Microfluidic devices typically have very small cross-sectional channels — from a few microns to tens of microns in diameter through which the samples pass. The channels can only handle tens to hundreds of microliters in tens of minutes. For real-world samples, such as the testing of drinking water, current specifications require volumes 100 mL or more to be used. Samples this large would take days to flow through the microfluidic chip. LIBNA uses an off-chip concentration method to process large fluid volumes into small volumes using novel filtration devices, developed with collaborators from Biomedical Engineering and Agriculture. These small volumes are then injected into the chip, where the sample is further concentrated using electrical and mechanical filters.

One microscale filtration approach is based on the fundamental characteristics of bacteria and antibodies, and applies the dielectrophoretic effect to trap bacteria. Bacteria have electrical and dielectric properties with an associated dielectric constant. Antibodies are proteins that recognize and capture specific bacteria based on their surfaces. By attaching antibodies on a surface, specific bacteria can be identified and collected.

Dielectrophoresis is the movement of a particle, for example, the bacteria of interest, in an AC field when the dielectric constants of the particle differs from that of the medium. For this to occur, the electric field has to be spatially non-uniform to create a non-uniform dipole in the sample. A net force (dielectrophoretic force) is then formed, which can be used to concentrate or trap bacteria in a fluid through microchannels in a chip. Antibodies are attached to the inner surface of the microchannels, capturing and identifying the bacteria of interest.

Rapid detection of live bacteria is a major issue. Today, live bacterial detection is commonly accomplished in a Petri dish containing agar, which is a growth medium. Bacteria generally double in number every 20–40 min, and require days to reach the colony size needed for detection. LIBNA has found a way of reducing the threshold of detection by orders of magnitude down to just a few bacteria.

During growth, bacteria use specific nutrients, like glucose and sugar, and excrete byproducts that are ionic and acidic in nature. These ionic byproducts can result in changes in the electrical resistance, conductance or impedance of the growth medium. Based on this idea, the team developed Petri dish-on-a-chip (Fig. 2 ), which electrically monitors bacterial growth on-chip using electrodes and microfluidic channels after the bacteria are concentrated using the nanoliter volume channels. Changes in current are monitored as a function of time, and testing is complete after one or two doubling cycles, reducing bacterial growth times from several days to 20–60 min.

2. A Petri dish-on-a-chip is used for bacterial culture, and the necessary nutrients for bacterial growth are provided via fluidic and electrical connections.1

The Petri dish-on-a-chip is a cartridge consisting of a microfluidic chip ~1 × 1 in. that is mounted on a PCB with electrical interface and fluidic connections. The cartridge is plugged into a box, as is the electrical filtration cartridge, for concentrating a large amount of fluid.

Anomalies in DNA sequencing underlie nearly every medical problem and every disease. DNA has four bases, A T G C, which are repeated for different sequences. The sequence in all human cells remain unchanged over time. Related DNA information is translated into a protein, which in turn performs various functions. Mutations in sequencing can result in diseases, including cancer, for instance, so knowledge of the sequence of DNA is critical. Bashir's group again applied silicon-related nanotechnology to probe the sequence of short strands of DNA molecules.

Nanopore channels are formed in an oxidized silicon membrane using a combination of electron beam lithography, electron beam-induced melting of the oxide, and focused ion beam (FIB). One channel, with a minimum diameter ranging from 5 to 20 nm, is positioned in the membrane on a silicon chip. The membrane separates two compartments of fluid, and the only path between the two compartments is through that nanochannel. When a DNA molecule is put in a fluid, such as phosphate-buffered saline solution, it has a net negative charge. Therefore, when an electric field is applied across the negatively charged fluid, DNA moves toward the positive electrode, a phenomenon known as electrophoresis of DNA (Fig. 3 ).

3. DNA molecules flow through and are identified in the 20 nm nanopore channel (TEM), which is positioned between two membranes (schematic). 2

To add selectivity to this technique, the researchers relied on an important property of DNA — the existence of a complementary strand. The implication is that a specific sequence can be recognized by its complement. Base A binds to T and G binds to C. Therefore, if an unknown sequence binds to a complementary sequence, called a probe sequence, it can be identified.

Short sequences of DNA were attached to the inside of a nanochannel, a cylinder where the inside is lined with silicon dioxide, which has probe molecules attached to it. The unknown DNA is put on one side of the membrane, and a voltage between 5–300 mV is applied across the membrane. A single DNA moves through the channel, which has a diameter small enough to allow one molecule to go through at a time.

When the target DNA is not going through, ions from the potassium chloride medium flow through the channels, and a current on the order of tens to hundreds of picoamps can be measured. However, when the DNA passes through the channel, it physically blocks the background ionic current. The subsequent drop in current is indicated by a downward pulse. In essence, it can be said that perturbations in the current indicate molecular transport. Bashir's group found that in these functionalized nanochannels, target DNA strands sequence could be identified by the pulse width of the translocation events. Where the pulse widths are shorter, the target molecules are recognized by the molecules attached to the inside of the nanochannels.

This approach is similar to that used to count cells that are ~10 µm in diameter. However, what makes this technique powerful is added selectivity of the nanopore channels. This approach is the first report of a single molecule solid-state silicon-based DNA sensor using nanochannels that have selectivity and provide information on selectivity and sequence.

The application of the silicon process to the development of biological diagnostic tools seems endless. New studies continue to emerge from LIBNA, such as silicon on insulator (SOI) transistors on a chip for detection of DNA and proteins, and silicon cantilevers with attached antibodies to capture target viruses. All are showing promise in a field where the demand for fast, easy and disposable detectors are of increasing interest.