Good microscopy results may be possible on the cheap, say researchers from the French government-funded Centre National de la Recherche Scientifique (CNRS). A group from the Institut Fresnel at Aix-Marseille Université in Marseille, France, showed that the combination of a low-cost, low-numerical-aperture (NA) objective and a latex microsphere could work almost as well as a much more expensive confocal microscope in the detection of single molecules by fluorescence correlation spectroscopy. Researcher Jérôme Wenger doesn’t see a microsphere-based system replacing classical microscope objectives, but he does predict that the new method will have its uses. “I see its application rather implemented in combined optomicrofluidics systems for fast biodiagnostics in a lab-on-a-chip.”The result could be an inexpensive and small system suitable for clinical use. The CNRS group has submitted a patent application on the idea. The team also has been contacted by a biotech company, although Wenger characterized the inquiry as merely informational in nature.Fluorescence correlation spectroscopy is a powerful tool for detecting a low concentration of biomolecules. However, the technique requires that a fluctuating signal arising from the molecules of interest be distinguished from background noise. Achieving a suitable signal-to-noise ratio from a small number of molecules requires a high fluorescence count per molecule and a low background. Traditionally, this has been done with a confocal microscope and a high-NA objective. The lens improves the light collected and increases resolution compared with a low-NA objective. This high-NA approach allows detection of single molecules but is expensive and hard to integrate into a lab-on-a-chip.Much of the cost and complexity come from the objective, but simply using a low-cost, low-NA lens won’t work. The observation volume increases as the aperture drops, which lowers the excitation intensity and the fluorescence count per molecule. The larger volume also drives up background scattering and enhances the effect of photobleaching, which extinguishes a molecule’s signal. To compound the problem, a lower NA leads to lower collection efficiency, which cuts the count even more. An inexpensive low-NA lens is used instead of a costly conventional high-NA microscope objective to detect single-molecule fluorescence in a confocal microscope setup, thanks to a latex microsphere set at the lens focus. The insert displays electric field intensity with a 2-μm cylinder in a water environment illuminated by a Gaussian beam at a wavelength of 633 nm with an NA of 0.3. Images courtesy of Jérôme Wenger, CNRS.The combination of these effects can be large. Going from a 1.2-NA water-immersion objective to a 0.4-NA air objective, for example, cuts the count 60-fold, the researchers noted.Borrowing from the pastThe solution to this problem, as reported in the Sept. 1 issue of Analytical Chemistry, is a reworking of a classical compound microscope objective. Typically, these lenses are very close to a glass half-sphere, and the researchers put the concept into the microscale, Wenger said. He added, “What is nice is that a complete sphere still performs well and is very cheap and simple to obtain commercially.”The researchers used a setup that involved latex microspheres between 2 and 5 μm in diameter from Fluka Chemie of Buchs, Switzerland, attached to a glass coverslip. They arranged these so that there was a single bead per 10 × 10-μm area.The group came up with the idea to employ these beads in microscopy because of recent work on the focusing of light by microspheres. Various papers have predicted that when these small latex spheres encounter an appropriately focused beam, the result is another beam, a photonic nanojet, on the opposite side. This second beam emerges with high intensity and low divergence, measures less than a wavelength of the excitation light across and could form a small observation volume, thereby potentially solving the count problem very inexpensively.In a demonstration, the scientists immersed the coverslip-bound microspheres in a solution containing the fluorescent molecule Alexa 647 at various concentrations. They illuminated the solution with a HeNe laser at 633 nm and collected the resulting 670-nm fluorescence with 0.25- to 0.4-NA lenses priced from $4.20 to $350 and made by Thorlabs of Newton, N.J., and Zeiss.For detection they used an avalanche photodiode from PerkinElmer of Fremont, Calif. A dichroic mirror and filters from Omega Optical of Brattleboro, Vt., kept stray light from the laser out of the detector.Making 40 equal 12,000From computer simulations, the researchers estimated that the photonic nanojet for the microspheres was about 390 nm wide and 1.65 μm long, close to the values for the observation volume typically reached with a high-NA objective. Tests with concentrations of Alexa 647 ranging from 1 to 1000 nM showed that the setup produced viable fluorescence correlation spectroscopy results, with the capability of detecting single molecules with all but the very cheapest lens.Fluorescence intensity correlation functions obtained with a 0.25-NA doublet and a 3-μm microsphere are shown (top). The bottom image shows a reference correlation function with a standard water-immersion 1.2-NA objective and no microsphere. The dye was Alexa Fluor 647, diluted in pure water to a 40-nM concentration. The performance of the system, the researchers noted, wasn’t as good as that of a confocal system with a $12,000 1.2-NA Zeiss objective. However, useful results could be achieved with a lens costing less than $40.As for the future, Wenger said that some problems have to be overcome before this approach can be used in a fully integrated on-chip fluorescence diagnostic system. Chief among them is the fact that the low-NA lens and microsphere have to be properly aligned and the distance between them carefully controlled.Although it would be possible to design a microsystem in which these two elements are fixed, he noted that this would be technically challenging and quite expensive. So the group is exploring other solutions. “Another strategy we’re currently working on is to replace the low-numerical-aperture lens by a simpler fixed optical system.”The group is also engaged in other work with microspheres. One avenue of investigation involves the performance of a microsphere and a high-NA lens. Initial results in this area are quite promising, Wenger said.