Roberto Etchenique and Oscar Filevich of Universidad de Buenos Aires in Argentina were attempting to carry out time-lapse measurements of cell cultures in long-term experiments. However, for such investigations, temperature stability is a necessity. The problem was that standard techniques based on thermometers, thermocouples and thermistors do not work well in small systems such as petri dishes, microfluidic chambers or flow injection analysis reactors.This montage of false-color images shows the heating of a petri dish as captured with a technique that uses the fluorescence of a ruthenium complex to visualize temperatures.Therefore, the researchers developed their own technique, one based on monitoring changes in fluorescence with temperature of a ruthenium complex, [Ru(bpy)3]2+. Etchenique was familiar with the complex from other research, and he knew that it exhibited a Stokes shift of ∼150 nm. The large shift meant that separating the excitation from emission light could be done with a simple filter — an important advantage.“When you have to deal with somewhat subtle variations of fluorescence, a simple way to filter is really important,” he said.In developing the method, the scientists first measured the temperature dependence of the fluorescence of [Ru(bpy)3]CI2 in 10–5 M solution using a spectrofluorometer from Ocean Optics Inc. of Dunedin, Fla., and a 470-nm LED made by Philips Lumileds of San Jose, Calif., as a light source.They heated the solution, taking its temperature with an immersed calibrated thermistor. Besides the spectrofluorometer measurements, they took images of the fluorescence on an 8-megapixel digital camera from Canon Inc., using a gelatin filter that prevented light below 550 nm from passing through, thereby removing the blue excitation light.From these measurements, they derived a series of intensity curves from 20 to 60 °C, enabling them to determine the quantum yield fluorescence of the solution versus temperature. When plotted against temperature, the quantum yield was a declining straight line, making temperature measurements convenient.Armed with this information, the investigators tried the technique in several situations. First, they ran the fluorescent solution through a heated flow injection analysis reactor constructed of coiled polyethylene tubing. They constructed a two-dimensional fluorescent screen for temperature imaging by encapsulating the solution between two cover glasses, and then placed the screen in a petri dish.For both the flowing solution and the screen, they captured the resulting fluorescence with a camera and converted the images into temperature after calibration at two points with another sensor. They achieved a resolution of 0.05 K, which they verified by comparing fluorescence results with those obtained with a calibrated thermistor. With software, they produced false-color images, creating an easy visual representation of the temperature at any point.Etchenique noted that the technique works with commercial cameras and standard LEDs. Results might be improved with the use of a CCD that has a higher dynamic range — 16 bits instead of the 12 provided by the camera that they used. However, he added, the resolution that they achieved was better than what they were looking for by a factor of five.He foresees several possible applications, including the measurement of temperature in small cell-growing chambers. He also would like to see the technique — or one similar — make its way into the starting point of their research.“I would like to have plastic petri dishes in which the bottom is covered with a temperature-sensitive layer,” he said.Analytical Chemistry, Nov. 1, 2006, pp. 7499-7503.