MINNEAPOLIS, MN -- It's not quite a magic crystal, but it assembles itself just as its "master" bids and even performs the trick of doubling the frequency of light. University of Minnesota researchers have created an organic crystal lattice that forms in exact predetermined architecture--a rare feat--and exhibits polarity, which is as difficult as coaxing magnets to line up with all the north poles facing the same way. The technology opens the door to new types of "polar" crystals, many of which can change red light to green or blue light and may be used to build shorter wavelength lasers. Other potential uses include optical switching and drug purification. The research will be published in the Nov. 30 issue of Science. On the molecular level, a crystal lattice is a series of chambers where the "floor," "ceiling" and "pillars" are all held in place by weak chemical attractions. Inside the chambers are "guest" molecules that perform some function. In order for the crystal to work, all the chambers must be the same and all the guest molecules must be aligned in the same direction. If some guests are standing heads up and others are upside down or sideways, the material as a whole will have no sense of direction--no polarity. "It's usually difficult to get molecules to go the way you want because the forces [pulling them into position] are so weak," said Michael Ward, professor of chemical engineering and materials science and first author of the Science paper. "What we're doing is crystal engineering, which means designing solid-state structures based on molecules by looking at the molecules themselves and asking how they'll guide themselves into a 3-D crystal lattice." The new crystals double the frequency of light, a property often associated with polar crystals. Although the intensity of light that exits the crystal is lessened, the phenomenon means that, for example, green or blue lasers may be fashioned using red light. Blue lasers have been difficult to fabricate, said Ward, but some made from inorganic materials do exist. They are desirable for telecommunications because information can be transmitted faster at higher frequencies and shorter wavelengths. Related crystal lattices can also be used to separate molecules with the same composition but different structure. In particular, they may have the potential for separating drugs that exist in two forms, like right- and left-handed gloves. The need for such separation and purity of a certain handedness is exemplified by thalidomide; one form alleviates morning sickness, the other causes birth defects. In optical switching, the crystals would work as transistors to pass information. A crystal lattice is like a hotel for its guest molecules. Unlike hotels for people, however, the guests take their places at the same time the lattice assembles itself. The floors and ceilings are identical sheets consisting of a patchwork of two different types of molecules, with "pillars"--a third type of molecule--stretching between. Such lattices are symmetrical, and a guest molecule would have no idea which way to orient itself within the framework of sheets and pillars. In a polar crystal, the guest molecules are polar, which means that one end is relatively negative in charge and the other end relatively positive. A row of polar guests would tend to line up in alternating orientation, so that positive ends are bracketed by negative ends and vice versa. Thus, their charges would cancel each other out and the crystal would be nonpolar. But Ward and his colleagues used pillars shaped like bananas. These molecules can only fit between the sheets if they line up in one direction, i.e. ))) or (((, but not both. This in turn forces all the guests inside the lattice to line up in the same direction, and the material exhibits polarity. "Now that we know how to do this, we can try other 'pillar' molecules and other 'guests' with the same basic framework," said Ward. "What makes this unique is, it's so amenable to modifications." Working with Ward on the project were K. Travis Holman and Adam Pivovar. The work was supported by the National Science Foundation, and Holman was partially supported by the National Science and Engineering Research Council of Canada.