Two new and very cool microscopy techniques have been announced recently. One, the optofluidic microscope, could put an entire microscope, including display, into a device the size of an iPod. The other, photoactivated localization microscopy (PALM), was invented by two unemployed scientists with $50,000 in two months. And both microscopes are sub-wavelength--they use light to gather images that can be more detailed than the so-called diffraction limit.
The optofluidic microscope uses microfluidics to move the sample past a single line of single-pixel light sensors. So instead of needing a two-dimensional CCD element with many thousands of pixels, it just needs a few dozen sensors. What's more, it doesn't even need lenses. By putting the holes for the sensors at a shallow angle to the specimen's travel, the holes can be spaced widely apart from each other, but adjacent holes will still cover closely-spaced strips of the specimen. And if the holes are sub-wavelength and the sample is very close to them, then the light that comes through the hole will have passed through only a tiny area of the sample. On the other side of the hole, anything that measures light will do. The small size and simple design may make it a good fit for on-the-spot medical diagnostics in developing nations. Here is a news story about the microscope, and here is a presentation (PDF) with more details and a clearer picture of the arrangement of holes.
The PALM microscope uses a very simple, yet very clever, technique to find the location of individual proteins in a cell with nanometer precision--even when they are close together.
The above image is the one the news stories show, but I found their captions confusing. The image shows five views. The cloud of points near the top represents what's really there. The top plane is what an ordinary microscope sees. The bottom plane is what an electron microscope sees. The plane just below the points is what PALM sees, and the next-to-bottom plane is a very informative superposition of PALM and electron microscope images.
It's been known for a little while that an ordinary optical microscope could be used to find a single isolated nanometer-scale spot of light with nanometer precision. And recently, fluorescent chemicals have been developed that can be turned on by hitting them with light. So all you have to do is to first, label all the proteins you want to see. Then hit the cell with a very weak burst of light--which will turn on only a few of the proteins. Determine their locations, since they will be far enough away from each other to be effectively isolated. Then, either turn them off or wait for them to photobleach (burn out), hit the cell with another weak pulse of light to turn on a few more, and repeat until you've found them all. Here is a news story about the invention and the inventors. Here's another news story about this technique, plus several other sub-wavelength microscopy techniques.
The articles don't say so, but I have to assume someone is working on combining the PALM technique with confocal microscopy to get 3D images.
Some of these microscopy techniques may be enabling technologies for people trying to characterize small molecular structures. As larger engineered molecular structures are designed and built more easily and rapidly, it's going to be more and more important to be able to tell whether you've built what you think you have. Once researchers can be confident in their structural engineering, they will start to add functions, and look for ways to combine the blocks into larger, perhaps heterogeneous aggregates.
Any day now, we'll have to switch from predicting the development of molecular manufacturing, to reporting on it...