FluorEVER^TM MICROSCOPE

With the intention of developing nanocrystals (refer to separate section on Quantum Dots) that will work well with LED fluorescence, I have been experimenting with some of the newer 1-watt and 5-watt LEDs currently available. Comparing the ultrabright LEDs used in my initial experiments with the new breed of LEDs is a little like comparing the glow of a malnourished, fatigued, insecticide-sprayed, swatted, over-the-hill firefly to the light of the midday sun! But even with these dazzling devices, too much light is lost in installing them along the usual illumination paths in regular microscopes. I tried inserting a 5-watt LED, collimator, and heat sink into the lamphouse of my Leitz microscope. The visual results were a little disappointing. It was becoming obvious that the only way to squeeze every available photon from the LEDs was to design and build a microscope from the ground up.

The result of this optimized design is a rather curious-looking reflection microscope I have dubbed the FluorEVER Microscope. Why? The manufacturer of the LEDs I use rates the lifetime of the devices at 50,000 - 100,000 hours. However, over that period of time the light intensity can be expected to decrease at least by half. So 10,000 hours seems like a good, conservative useful lifetime for the LEDs. This works out to approximately 3-4 years of use as compared to about a month or two for a standard mercury lamp, assuming equal daily usage!

I developed my design around infinity-corrected objectives and, primarily, 1-watt and 5-watt LEDs from Luxeon. (Not all the colors available in the 1-watt LEDs are available in the 5-watt units.) A holographic diffuser was incorporated to keep light loss by inefficient scattering at a minimum. Because of the efficiency required, no prisms or extraneous mirrors were used, and so the microscope is monocular. This has proven to be satisfactory, especially since the primary use of this microscope is for CCD recording. Interference-type exciter filters are generally used unless the dye-emission wavelengths are so far removed from the LED-emission wavelengths that such filters are unnecessary. Usually, inexpensive long-pass filters are used as barrier filters to block reflected LED light. Sometimes, however, interference filters are needed to restrict the emission range. As indicated, the microscope was designed with light efficiency and throughput as the chief goal. There was, however, one point in the microscope where I compromised light efficiency. The reason for that will be discussed below.

FluorEVER Microscope

The microscope is seen here set up for visual and b&w CCD use. (Pass the mouse pointer anywhere over the photo to see the component names.) The eyepiece is fixed at 90° to the microscope axis using a first surface mirror in a flip-mirror housing. Below the diagonal housing are the barrier filter wheels with selectable long-pass and band-pass filters. Further down the tube is the tube lens required for infinity-focus microscopes. The diagonal holder below that houses a beamsplitter that reflects light from the vertical illuminator (seen projecting to the left) down through the microscope objective to the specimen. The reflected fluorescence passes through the beamsplitter, through the barrier filter(s), and is viewed by eye or passed to the camera.

Ideally, the beamsplitter should be a dichromatic design that would allow up to 81% efficiency in illumination and transmission. However, each different LED would require its own very expensive dichromat coated for reflection of its specific wavelength as well as for transmission suited to specific dyes. For the sake of overall cost and setup time, a 50/50 half-silvered beamsplitter was used. This brought the beamsplitter efficiency down to 25%, at best. For early exploratory work evaluating specimens, dyes, and filters, this compromise was necessary.

The auxiliary lamp is used to supply fill-in light while viewing. Also, LEDs operating at other wavelengths than the standard 1- and 5- watt Luxeons can be placed in the lamp for special lighting demands. The lamp is operated off a separate, variable power supply.

The vertical illuminator contains the super-ultra-mega-bright LED and has its own power supply. Collecting and focusing lenses, an excitation filter, and appropriate optical diffusers complete the unit.

Finally, the microscope is focused using the vertical adjustment micrometer to raise and lower the stage.

The CCD camera

As seen in the above configuration, the recording end of the microscope is a relatively inexpensive but highly sensitive integrating b&w video camera (rated ~0.0001 lux @ f/1.2). Alternatively, a highly sensitive (0.0003 lux @ f/1.2) but non-integrating b&w camera is used for moderately low-level light videography (e.g., recording stained and brightly fluorescing bacterial flagella in motion). An integrating color camera is employed when color recording is desirable. The resolution of the color camera is less than that of the b&w cameras. An intensified CCD (ICCD) camera is used for very low-level light recording, as in bioluminescence or very faint fluorescence. I typically use a Dage-MTI GenIISys unit. Finally, when feasible, the cold camera referred to earlier can also be used for higher quality, low-level illumination work.

For tricolor work with any of the b&w cameras, three exposures are made through three different emission filters. In the dual-filter wheel setup shown, the lower wheel holds the long-pass, LED-light blocking filters. The second wheel placed closer to the camera or eyepiece holds band-pass filters for red, green, and blue. This arrangement allows the broadband light passing through the long-pass filters to be further separated into additional channels. The three resultant images are combined in software to yield a tricolor photo. In some instances, the assignment of red, green, or blue channels to specific images may be arbitrary, depending upon the desired appearance of the final photo. This is necessarily true if blue light has been totally blocked by the long-pass filter. An extra benefit of combining three images to produce one is the averaging out of random noise present in any individual picture. The downside is loss of any hope to capture rapid motion. Both the LED and the excitation filter may be changed between exposures too.

For some work, I make use of my total internal reflection technique described earlier that uses a prism coupled to the glass microslide. In that case, the auxiliary lamp comes into play. This microscope is definitely a work-in-progress. Over time, improvements will be added to make the microscope even more flexible for fluorescence studies.

But before continuing with in-depth testing of the new microscope...

In my original paper dealing with LED Fluorescence (see References section), I suggested that medical researchers try using the technique for screening patients' sputum for tuberculosis. It is well-known that mycobacteria (such as M. tuberculosis) can be stained with fluorescent dyes like Auramine O. Once stained, the bacteria are not easily decolorized. That is the basis for the fluorescent screening method.

Not having access to the tuberculosis bacteria, M. phlei and M. smegmatis were selected. They have waxy walls containing mycolic acid just like their deadly relative and should react similarly. I found it quite challenging to stain those bacteria with Auramine O and Rhodamine B but finally succeeded by using a very aggressive, elevated-temperature staining method. The pair of fluorescent dyes combined to lend a golden-orange hue to the bacilli. Employing the original ultrabright LEDs described earlier, the LED fluorescence technique proved marginal, though this may be due to inexperience with the staining method. However, with the FluorEVER Microscope the mycobacteria stood out beautifully. Through the eyepiece, the bacteria glimmered like a sea of golden threads against a dark background.

The medium-power images seen here were recorded in black & white and pseudocolored. A 40X Pl Fluorite objective was used.

The following pages present additional photographic results of experiments intended to test the new microscope. Coming up, a look at fungi and algae, including two bioluminescent organisms.