Chemical & Engineering News,
March 18, 1996

Copyright © 1996 by the American Chemical Society.

Array detectors are transforming optical spectroscopy

Stu Borman,

C&EN Washington

Enthusiastic predictions in the 1980s that charge-coupled device (CCD) and charge-injection device (CID) array detectors would revolutionize optical spectroscopy are turning out to be correct, says chemistry professor M. Bonner Denton of the University of Arizona, Tucson.

Denton - who arranged a Pittcon symposium on "Present and Future Impact of Array Detectors on Spectroscopy" - was an early enthusiast for analytical applications of array detectors. Now, finally, "array detector technology has really taken off," he says.

Array detectors are area light-detection devices. They are the electronic equivalents of color film - in fact, CCDs are used as detectors in video cameras. Light causes the devices to generate charge that is read out quickly and interpreted to construct images.

"Array detectors have become widely accepted in state-of-the-art Raman, fluorescence, and phosphorescence spectroscopies, with virtually all major manufacturers providing systems that include this technology," says Denton. "More recently, a revolution has taken place in atomic spectroscopy and mass spectrometry, with the introduction of both CCD- and CID-based systems. And announcements of devices suitable for X-ray, UV-Vis, and near- and medium-IR spectroscopy by various array detector manufacturers promise to have a major impact on the future of spectroscopic measurements in these wavelength regions."

The dynamic range of some CCD and CID scientific camera systems is so good that they can see stars during the day. "Human eyes have only about an eight-bit [28] dynamic range in gray scale - one part in 256," explains Denton. "CCDs and CIDs have gray scales of one part in 64,000, 128,000, or even 256,000. What this allows you to do is ... to go outside [at noon] and look at the blue sky with one of these cameras, focus it to infinity, and see stars."

This 512 x 512-pixel CCD, measuring about 1.2 cm on each side, is reflecting an image of the sun. The complex electronic circuitry on its surface acts as a diffraction grating, generating multiple solar images.

And array detector technology continues to develop. For example, Denton says a device was recently announced with over 81 million pixels and very high resolution. Prices of some CCD and CID systems are coming down as well.

One analytical application of array detectors discussed at the symposium was fluorescence imaging of biological cells. Chemistry professor Edward S. Yeung and coworkers at Iowa State University, Ames, use the natural fluorescence of the neurotransmitter serotonin to image neuron-type cells called astrocytes, using a UV laser for fluorescence excitation and a CCD camera to record the images.

"We take 'movies' of the astrocytes as the cells take up and release serotonin under external stimulus," says Yeung. Using the CCD to image this process in real time makes it possible to determine how fast serotonin is released and to visualize sites on the cell where release occurs.

"The interesting result is we find that these cells actually take up serotonin and store it in different locations in the cell," says Yeung. "Also, serotonin is released at different rates from different parts of the cell."

Microscope system can be used for CCD-based UV-excited fluorescence imaging of neurotransmitters. In the photos above, a normal optical microscope image of astrocyte is shown at top. In UV fluorescence image (305-nm excitation) of the astrocyte (lower photo), bright areas are sub-cellular regions with higher content of serotonin. The UV-CCD system makes it possible to monitor serotonin by its native fluorescence-that is, without the chemical derivatization required by normal microscopy.

Yeung and coworkers are also using array detectors to simultaneously see single molecules fluorescing on a glass or quartz surface. "We actually have some data that indicate we are seeing the diffusion of these single molecules from one location to another," says Yeung.

The fluorescence is generated by evanescent wave excitation, a phenomenon in which the electric field of a light wave "leaks" out a little into solution at a solid-liquid interface. The volume of the observation area is defined by the resolution element (pixel size) of the microscope and the depth of penetration of the evanescent wave into the solution. In the experiment, this turned out to be only about 7 attoliters (10 - 18L). In a nanomolar dye solution, only one of every several hundred of these extremely small volume units will contain a molecule. The field of view they use enables them to view about 30 molecules at a time.

"This is an extremely simple approach to look at single molecules that everyone can do in their own lab as long as they have a normal optical microscope and the right kind of CCD," says Yeung. What's required is an intensified CCD, one with a light-amplifying device placed in front of it. The technique can potentially be used, he says, to visualize binding phenomena and chemical reactions of single molecules at surfaces.

The cost effectiveness of some CCD-based instruments was emphasized by Gary R. Sims, president of Spectral Instruments, Tucson, Ariz. "A new class of spectroscopic instruments has emerged in the last two or three years, made by one of several companies - ourselves, Ocean Optics, Control Development, and Analytical Spectral Devices," says Sims.

The instruments combine a low-cost array detector with fiber optics. The optical fiber delivers light to a sample and brings transmitted or reflected light back to a spectrometer.

"The lowest cost instruments use linear CCDs - not the scientific CCDs that people have discussed for many years, but rather devices that were built to be used in fax machines and photocopiers. So the price of the arrays is very low - about $30." In contrast, traditional spectrophotometer detectors are photodiode arrays, which typically cost thousands of dollars, he says.

"Linear CCDs have no moving parts, and the use of fiber optics is ideal for these detectors because their physical size is very small and on the same order as fiber-optic diameters. Putting those two technologies together makes a reasonably good instrument, and performancewise it can be as good as the traditional state-of-the-art laboratory spectrophotometer."

Sims described how an Analytical Spectral Devices instrument is being used for research on coral in the Caribbean. "A scuba diver takes his optical fiber, dives down 20 feet underwater, and measures fluorescence from the coral polyps in place, while someone up on the boat has a spectrophotometer and a laptop computer acquiring the spectra."

Application areas for the instruments, according to Sims, include laboratory research where cost is a key consideration, educational demonstrations of instrument capabilities, process analysis, and analysis of highly toxic or radioactive samples - where the remoteness of fiber-optic sampling can provide an important safety margin.

Although linear CCDs are generally not sensitive enough for Raman spectroscopy, high-performance scientific array detectors certainly are. "Until very recently, Raman spectroscopy was not considered a very viable technique for routine chemical analysis because the sensitivity wasn't there," says Denton. "The best number in the literature for a fiber-optic type of probe and cell configuration - not surface-enhanced or resonance Raman - was 0.05 molar. We've now extended that down to 10 - 7 molar," using a fiber-optic diode-laser Raman spectrometer with array detection.

His group is currently using the instrument to monitor organic plumes moving through aquifers at different levels and to identify unknown pollutants at illegal dump sites. Array-detector-based Raman spectroscopy, says Denton, "is a wonderful monitoring tool for various chemical processes, biotechnology processes, quality control, and pharmaceutical laboratories."

Denton and coworkers are also investigating the use of array detectors for thin-layer chromatography (TLC), generally considered to be only semiquantitative and not very sensitive. "But we've developed a new sample application system to use in conjunction with array detectors that allows us to put 50 or 100 samples on a single plate," he says. "You develop those samples into maybe 30 components each and look at them with one of the CCD or CID camera systems, which can see differences that the human eye can't see."

The researchers use software developed originally for astronomical imaging to accurately integrate the volume of the TLC spots. "We're able to achieve reproducibilities and accuracies on the 2 to 5% scale," says Denton, "and our detection limits on many significant compounds, like aflatoxins, are 2 to 3 picograms. We can actually see aflatoxins in the brand of peanut butter that I normally buy off the local store shelf," he says. The system has a throughput of 50 samples a minute.

Denton also spoke on the use of array detectors for doing Laue X-ray diffractometry (C&EN, March 11, page 30) and mass spectroscopy. "JEOL U.S.A. is showing a new CCD-detector mass spectrometer system that demonstrates two orders of magnitude better sensitivity than conventional single-channel ion-multiplier detector technology," he says.

Thomas M. Jovin, chairman of the department of molecular biology at the Max Planck Institute for Biophysical Chemistry, Göttingen, Germany, and coworkers are using array detectors in conjunction with fluorescence and phosphorescence microscopy to study biological processes in cells involving hormones and growth factors. One such study focuses on the interaction of cholera toxin with cell surfaces and its mechanism of action inside cells. The group is also developing a technique for diagnosis of lung cancer, involving fiber-optic imaging through a bronchoscope. "It's imaging in situ, so to speak," says Jovin.

James R. Janesick, former head of a CCD development group at California Institute of Technology's Jet Propulsion Laboratory (JPL), and now executive vice president and chief scientific officer of Pixel Vision, in Huntington Beach, Calif., calls CCDs "near-perfect devices" in terms of some key performance parameters. But he and his coworkers are trying to develop even better ones.

At JPL, Janesick's group was responsible for development of CCDs for several National Aeronautics & Space Administration missions, including the Hubble Space Telescope. CCDs for these missions are "slow-scan" devices with readout rates of 100,000 pixels per second or less. But some scientific applications - such as monitoring chemical reactions with picosecond time resolution - require much higher speeds. Today's challenge, says Janesick, is to develop high-speed scientific CCDs for those applications - a task that Pixel Vision and other companies are working on.

Fast-scan cyclic voltammograms of dopamine release upon electrical stimulation (vertical arrows) of mouse brain tissue, obtained by Wightman, Caron, and coworkers, show high levels of dopamine in wild-type mice (with normal dopamine transporters) and lower levels of dopamine release in heterozygous mice (with lower transporter activity) and homozygous mice (lacking transporters altogether). Lowering of dopamine release in transporter-deficient mice apparently reflects an attempt by the animals' systems to down-regulate dopamine to compensate for decreased uptake by the transporter.

Return to Article Index

[ACS Home Page] [ACS Publications Division Page]