Quantum dots are fluorescent nanoparticles of semiconducting material. The color of light that they emit varies with the size of the dot, shifting toward the blue end of the spectrum as they get smaller. Proposed applications for dots include lasers, color displays and bioimaging. But realizing their commercial potential has been slow, because the dots are expensive; they go for at least $2,000 a gram, largely because of pricey solvents used in their manufacture. Now a team at Rice University has demonstrated that much more economical solvents work just as well, slashing the outlay on materials. They have also determined how properties of the solvent affect dot formation, which may help in developing large-scale systems of production.

Indeed, the desire to scale up existing synthesis methods led chemist Michael S. Wong and his colleagues at Rice to the price-saving technique. In particular, they were interested in cadmium selenide dots, one of the most studied varieties. The commonly used recipes for cooking up cadmium selenide dots were 10 years old and involved heating precursor chemicals to temperatures of around 250 degrees Celsius in special solvents.

The researchers quickly realized that the cost of the standard recipes was prohibitive, and 90 percent of the material's expense came from the solvent octadecene. They then studied the literature to try to determine what made octadecene so special. For example, it is a long-chain hydrocarbon, which means it is compatible with oleic acid, another component of the recipe. But after an extensive review, they decided this molecular structure might not be so important after all.

What was important? Clearly, the solvent had to have a high boiling point and resist decomposing--and be cheaper than octadecene. The chemists were familiar with so-called heat-transfer fluids that met all three of those requirements. Heat-transfer fluids, however, are not used as solvents, being neither as clean nor as pure as typical solvents. But fortunately, that is not an issue for quantum dot manufacture.

So the team tried making quantum dots in two fluids, Dowtherm A and Therminol 66. Sure enough, the dots formed, much like in the conventional expensive octadecene. "We were very excited," Wong recalls. Using the cheaper of the two, Dowtherm A, cut the costs by 80 percent. The group is now also making dots of different composition than cadmium selenide using the same fluids.

The dots produced were somewhat smaller than those made using octadecene. After some study, the researchers came up with a mathematical model to account for those differences. The model predicts the rate of dot growth according to three quantities: the solvent's viscosity, the solubility of cadmium selenide in the solvent and the surface free energy, which relates to the stability of the quantum dot's surface in the presence of the solvent.

The number of dots made is still small--the team created only about 20 milligrams per batch. Batch processing may be suitable for applications that need only grams of dots, Wong says, but "the next step on the horizon is to develop a continuous-flow system for making quantum dots" in kilogram amounts. Because the quantum dot chemistry is very touchy, "we need to learn how to control it well in a reactor, so our mathematical model is a great starting point."