IMEC research explores the chip/cell interface
A continuing research program at the Interuniversity Microelectronics Center (IMEC) in Leuven, Belgium, is exploring the unknown land at the intersection of nanoelectronics and bioelectronics. Going far beyond conventional probing of living systems with macro-scale needles, the program is exploring direct electrical, mechanical, and chemical interfaces between living cells and operating integrated circuits. Last week, two research-group leaders from that organization described their work at the IMEC annual research review meeting.
One key program, described by group leader Carmen Bartic, is to develop in effect an artificial synapse: a direct interface between the interconnect points on a living neuron cell and sensors and receptors on the surface of an IC. The first step in this process, according to Bartic, is to develop the biochemistry to fabricate a surface coating for an IC that will help a cell live in extended contact with the chip. Today, IMEC is keeping cells alive and interacting with ICs for as long as several weeks, she said.
The next step in the process is to develop interfaces between the cell and the chip. Initial work in this area at a number of institutions has taken a rather statistical approach. Researchers spread a densely-packed array of electric-field sensors across the surface of a chip, encourage a neuron to grow across the surface as it pleases, and then scan the sensor array to look for the electrical activity that occurs in the vicinity of a synapse. This approach allows noninvasive study of synaptic activity, but it is rather inefficient, is limited to one-way communications, and can tell nothing about activity inside the cell except by inference. Also, since living cells tend to wander around a bit, this approach necessitates tracking the motion of the cell across the surface of the sensor array.
IMEC is now exploring a second step, according to Bartic: guiding the growth of a cell onto the surface of the chip in order to immobilize the cell and align its synapses with transducer sites. Mechanical confinement offers one way to do this. Researchers use conventional oxide-processing steps to form channels and wells in the surface of the chip, encouraging the cell to grow into a particular pattern and then holding the cell in place. A more elegant solution adds to the mechanical structure chemicals that stimulate cell growth or adhere to the cell walls, both encouraging the cell to grow into a particular channel and sticking it there. The hope is that with such techniques researchers can create custom networks of neuron cells precisely located over transducer sites on the surface of a chip.
I/O technology is driving the need for the precise location. The research team is also working on both electrical and chemical sensors that can accurately measure the changes at the synapses during neural activity. And they are exploring nanomachine chemical dispensers—ultraminiaturized versions of inkjet printer heads, if you will—that could produce tiny squirts of chemical ions to mimic the action of a synapse in stimulating another synapse. These electrochemical transducers would probably be too complex to fabricate into an array, so they depend upon having the neuron's synapses located right over the transducer clusters. But if that can be achieved, it would allow direct interaction between active neurons and the chip.
The initial aim of this work is laboratory interaction with purpose-grown neuron networks. But the eventual goal is to embed such an interface into a nerve bundle in-vivo, to act as a diagnostic tool, therapeutic tool, or an interface to a computer-controlled prosthesis.
Another effort in this program—this time not aimed exclusively at neuron cells—hopes to establish an electrical interface directly into the inside of a cell. To accomplish this, researchers are fabricating gold/platinum, mushroom-shaped pillars on the surface of a chip, with electrical connections through to the underlying circuitry. The team then coats these protrusions with a peptide that stimulates endocytosis, the process by which a cell surrounds and absorbs a foreign body.
A passing cell, so stimulated, stops to absorb the offending mushrooms and becomes fixed on the surface of the chip, with the tips of the electrodes reaching into the cell wall. In theory, differential measurements would then allow not only a look at the gross electrical properties of the cell's cytoplasm, but a more detailed dynamic mapping of the cell's contents.
While Bartic's team learns to explore inside cells, another project, under group leader Liesbet Lagae, is exploring the use of custom-tailored nanoparticles, attached to cells and manipulated by nanomachines fabricated on the surface of ICs. The team is investigating nanomagnetics, nano-optics, plasmonics, and nanofluidics as alternative technologies for exploring, sensing, and manipulating cells. This effort shows promise for capturing and measuring individual cells of a particular type, and for sorting them as well.
The process starts with attaching a metal nanoparticle to a protein, antibody, or other molecule that has a specific affinity for the cells you are looking for. As much as this may sound like science-fiction, it is apparently not a huge challenge today. The molecule will attach to the wall of its intended type of cell, attaching the nanoparticle as well. The trick then is simply to use nanotechnology on the surface of a chip to manipulate the nanoparticle, and hence the molecule attached to it, and the cell attached to the molecule.
For example, Lagae said, the surface plasmon resonance of metal nanoparticles can indicate the local dielectric constant on the surface of a cell to which the particles are attached, yielding important information about its chemical and electrical state. With a nanoscale laser, researchers can locally heat cells as well, adding another analytical tool. Such techniques could allow a rather thorough analysis of the surface chemistry of an individual cell.
Researchers also see promise in sensing the presence of particular cells. For example, Lagae offered, giant magnetoresistive sensors on the surface of a chip can detect magnetic nanoparticles attached to specific cells—cancer cells in the blood stream, for instance—with great sensitivity and selectivity. Further, by creating a moving magnetic field on the surface of a chip, a chip can collect cells with attached nanomarkers and manipulate them into a chamber for detection, counting, measurement, or treatment before other analytic procedures. Lagae estimated that such an approach could create detectors capable of detecting 30 picograms of a specific cell type per liter of fluid. Such techniques could be a breakthrough in the early detection of metastasis in cancers, for example.
Lagae's research is also exploring nanomechanical sensors, specifically for measuring mass. In effect, researchers are steering individual molecules onto a resonant beam and weighing them by measuring the resonant frequency. But existing nano-mass sensors of this sort are limited to about 10-15 kg sensitivity: not good enough to sense a single molecule. So Lagae's team is scaling down the size of the beam.
This, however, means that the beam no longer moves enough for the motion to be easily detectable by conventional nanoelectronic techniques. So the team is developing a design that places a waveguide near the resonant beam, and measures the motion of the beam indirectly by its effect on the waveguide. In this way the researchers are achieving 50 femtometer/√Hz resolution.
Lagae emphasized that these projects are far from abstract experiments. The individual techniques, taken together, make up a self-contained diagnostic laboratory on the surface of a chip, capable of identifying and analyzing individual cells and making previously impossible measurements of tiny concentrations of cells in a drop of blood. The goal is not just more accurate metrology, but the ability to identify the cells present in a particular diseased person, measure their genetic makeup, and tailor a therapy that works directly on the disease mechanism in that individual.