Also found in: Dictionary, Medical.
Semiconductor devices wherein minority charge is stored in a spatially defined depletion region (potential well) at the surface of a semiconductor, and is moved about the surface by transferring this charge to similar adjacent wells. The formation of the potential well is controlled by the manipulation of voltage applied to surface electrodes. Since a potential well represents a nonequilibrium state, it will fill with minority charge from normal thermal generation. Thus a charge-coupled device (CCD) must be continuously clocked or refreshed to maintain its usefulness. In general, the potential wells are strung together as shift registers. Charge is injected or generated at various input ports and then transferred to an output detector. By appropriate design to minimize the dispersive effects that are associated with the charge-transfer process, well-defined charge packets can be moved over relatively long distances through thousands of transfers.
There are several methods of controlling the charge motion, all of which rely upon providing a lower potential for the charge in the desired direction. When an electrode is placed in proximity to a semiconductor surface, the electrode's potential can control the near-surface potential within the semiconductor. The basis for this control is the same as for metal oxide semiconductor (MOS) transistor action. If closely spaced electrodes are at different voltages, they will form potential wells of different depths. Free charge will move from the region of higher potential to the one of lower potential.
An important property of a charge-coupled device is its ability to transfer almost all of the charge from one well to the next. Without this feature, charge packets would be quickly distorted and lose their identity. This ability to transfer charge is measured as transfer efficiency, which must be very good for the structure to be useful in long registers. Values greater than 99.9% per transfer are not uncommon. This means that only 10% of the original charge is lost after 100 transfers.
A second important property of a charge-coupled-device register is its lifetime. When the surface electrode is clocked high, the potential within the semiconductor also increases. Majority charge is swept away, leaving behind a depletion layer. If the potential is taken sufficiently high, the surface goes into deep depletion until an inversion layer is formed and adequate minority charge collected to satisfy the field requirements. The time it takes for minority charge to fill the well is the measure of well lifetime. The major sources of unwanted charge are: thermal diffusion of substrate minority charge to the edge of the depletion region, where it is collected in the well; electron-hole pair generation within the depletion region; and the emission of minority charge by traps. Surface-channel charge-coupled devices usually have a better lifetime, since surface-state trap emission is suppressed and the depletion regions are usually smaller.
The most significant current application of the charge-coupled-device concept is as an imaging device. Charge-coupled-device image sensors utilize the fact that silicon is sensitive to light. In fact, silicon is sensitive to wavelengths from about 400 to 1100 nanometers (from ultraviolet to near-infrared). When light photons penetrate the silicon surface, hole-electron pairs are created in the silicon. The number of hole-electron pairs created is a function of wavelength (photon energy), intensity (number of photons), and duration (length of time exposed to light).
In a charge-coupled-device image sensor, the light is focused upon an array of picture elements (pixels). These pixels collect the electrons as they are created. The number of electrons collected in each pixel is representative of the light intensity projected onto the sensor at that point. Periodically, the charges from all of the pixels are read out, and the image can then be reconstructed from the intensity and pixel location data.
There are two primary categories of image sensors. Linear image sensors have the pixels aligned along a central axis. Area image sensors have the pixels arranged in a rectangular (rows × columns) array pattern. Linear image sensors require relative motion between the sensor and the object being scanned. The relative motion is precisely known so that, as the object is scanned one line at a time, it can then be reconstructed one line at a time. Area image sensors do not require this motion.
The resolution of area image sensors has become equivalent to photographic film, enabling the development of digital photography. Cameras with very large, very high resolution area image sensors provide professional photographers better final pictures than are obtainable with conventional film, while lower-resolution, lower-cost, digital charge-coupled-device cameras are available to consumers.
A miniaturized charge-coupled-device camera allows a dentist to see inside a patient's mouth or a physician to see inside a patient's body. Charge-coupled-device area imagers are also used in intraoral dental x-ray systems. Charge-coupled-device-based systems with very large area image sensors have been introduced in mammography, to image x-rays of the human breast.
Astronomers have long used charge-coupled device area image sensor cameras mounted on very high power telescopes. By synchronizing the motion of the telescope with the Earth's rotation, the camera can “stare” at one spot in space for hours at a time. The long integration times allow distant objects to be imaged that are otherwise invisible. To keep the sensor from being saturated with thermally generated charge, these cameras typically cool the charge-coupled-device chip down to -50° to -100°C (-58° to -148°F). See Integrated circuits, Semiconductor