Miniature lenses for tiny microarray spots

Hank Hogan, hank.hogan@photonics.com

Although microarrays may seem small, they’re too big, at least for portability. Miniaturizing them further is difficult because they consist of many sensing spots. The individual spots are small, but most arrays manufactured today are millimeters or centimeters on a side and require expensive scanners for reading.

Researchers have developed ways to shrink the spots a thousandfold, which could greatly decrease microarray dimensions or increase the number of spots in a given area. But a fluorescence signal from such small spots would be a million times fainter than that from current ones. Now researchers at Ben-Gurion University of the Negev in Beersheba, Israel, have come up with a solution: microlenses that magnify the fluorescence signal and can be placed where needed.

According to Levi A. Gheber, current technology produces single spots of 50 to 200 μm, with an equivalent space between them. Features on a microarray have a pitch — the distance from one repeating feature to the next — of 100 to 400 μm. Thus, an array that is a centimeter on a side can hold only roughly 50 x 50 spots.

Some years ago, a research team that included Gheber used a nanopipette as the probe at the end of an atomic force microscope to print protein dots 200 nm in diameter. Another group, of which he also was a member, showed that nanochannels and wells could be carved into a protein layer using small amounts of the enzyme trypsin to do the chopping. These two investigations showed that it was possible, from a basic research point of view, to manufacture microarrays with spot dimensions 1000 times smaller than current ones. Fine tuning is still needed (and is under way) to transform these concepts into commercial reality, but even so, the detection of these nanospots is a big challenge.

These 3-D representations of polymer microlenses, based upon atomic force microscopy data, show how various physical parameters and the resulting focal length can be controlled by varying the deposition time (1 s on the left and 5 s on the right). Images courtesy of Levi Gheber and Ben-Gurion University of the Negev.

In microarrays, fluorescent labels are attached to target molecules, those that will bind to the arrayed material. Mapping the fluorescence pattern against the known distribution of spots on an array shows what specific targets are present.

However, reducing the linear dimensions of a spot by three orders of magnitude decreases the number of molecules by roughly six orders of magnitude, and consequently the fluorescence signal from such a spot will be a million times weaker, Gheber said.

The Ben-Gurion research team demonstrated a solution for this by constructing lenses using the same nanopipette employed to deposit the nanospots.

The technique uses a nanopipette, which is a hollow glass or quartz capillary with a tapered tip and an opening of a few hundred nanometers. Solution loaded into the nanopipette is drawn toward the tip by capillary forces. The solution will flow from the tip only when it is in contact with a material for which it has a greater affinity than it does for the nanopipette.

Constructed from a hollow glass or quartz capillary, a nanopipette, seen here end on, is mounted as the probe of an atomic force microscope. When filled with a UV-curable polymer, the nanopipette can deliver it in precise locations to create microlenses, which could be useful in the readout of miniaturized microarrays.

As reported in the April issue of Nano Letters, the researchers mounted the nanopipette on the end of a cantilever of an atomic force microscope from Nanonics of Jerusalem. They loaded the nanopipette with a monomer mixture that would polymerize when exposed to ultraviolet radiation. They deposited the monomer on glass coverslips, varying the contact time, and then exposed the droplets to 254-nm radiation from a UV lamp until the molecules cross-linked into a polymer, which was several minutes.

The result, verified by atomic force and near-field scanning optical microscopy, was smooth spherical droplets, with surface roughness of 1.6 nm. The diameter of the lenses was between 4 and 9 μm for a deposition time of between 1 and 20 s, with smaller lenses produced by shorter deposition times. The lens diameter didn’t increase linearly with contact time. Rather, it increased rapidly at first, and the growth slowed as the diameter neared a maximum value.

The contact angle between the droplet and the surface was determined by the properties of the two, and the refractive index of the materials was fixed. Therefore, the deposition time governed the optical performance of the lenses.

The researchers characterized the lenses using bright-field microscopy. For more precise characterization, they mapped the point spread function of the lenses as well.

Finally, to see how much the lenses could improve fluorescence measurement, they placed a microlens and a spot of fluorescein on the opposite sides of a coverglass. They imaged fluorescence from the spots through the lenses, and the measured intensity profile indicated that a detector placed at the right location would have collected 50 percent more light as a result of the lens.

Although the nanolens technique works, Gheber pointed out several areas where ongoing research aims to improve performance. Achieving this will involve a balancing act between competing demands.

For example, the lenses have a low numerical aperture because of the relative flatness of the drops and the refractive index of the polymer. Increasing the numerical aperture requires changing one or both of those parameters, which implies the use of a different material. On the other hand, any material employed still has to flow out of the nanopipette on demand, which favors higher weta-bility, a smaller contact angle and a flatter lens.

“We would like to have drops with high contact angle, but then they will be less likely to leave the pipette,” Gheber remarked.

Contact: Levi A. Gheber, Ben-Gurion University of the Negev, Beersheba, Israel; e-mail: glevi@bgumail.bgu.ac.il.