passed through an absorbing sample, a modified interferogram can be recorded. Fourier transformation
of this second interferogram produces a spectrum of the source modified by the characteristic
absorptions of the sample. Subtraction of the first spectrum from the second then yields the spectrum of
the sample which is the equivalent of a spectrum recorded by a double beam dispersive
spectrophotometer.
A full-range IR spectrum can be recorded in about 1 second and sensitivity can be enhanced by
accumulating multiple scans which are then computer-processed to increase the signal-to-noise ratio.
An important use for FT-IR spectrometers is in coupling them to gas chromatographs in what is called a
'coupled' or 'hyphenated' technique, i.e. GC-IR (p. 117). This enables the spectra of eluting peaks to be
recorded without the need to trap the component or stop the gas flow. Computer enhancement and
manipulations of the recorded spectra are additional features of both stand-alone FT-IR spectrometers
and GC-IR systems.
Detectors
The detector in a spectrometer must produce a signal related to the intensity of the radiation falling on
it. For instruments operating in the visible region a photovoltaic or barrier-layer cell is the simplest and
cheapest available. Current produced when radiation falls on a layer of a semiconductor material, e.g.
selenium, sandwiched between two metallic electrodes, is proportional to the power of the incident
radiation and can be monitored by a galvanometer. Barrier layer cells are robust and are often used in
portable instruments but they are not very sensitive and tend to be unstable during extended use.
A more sensitive device is the photoelectric cell or vacuum phototube. Electrons dislodged when
radiation strikes the surface of a sensitive photocathode are collected by an anode and constitute a
current which is proportional to the radiation intensity. After amplification, the signal can be fed to a
galvanometer, digital voltmeter or potentiometric chart recorder or VDU. Much higher sensitivity is
obtained from a photomultiplier tube which contains up to sixteen electrodes (dynodes) successive ones
being stabilized at increasingly positive potentials. Each primary electron ejected from the photocathode
by incident radiation and collected by the first dynode causes two or more secondary electrons to be
ejected by the latter. These are collected by the second dynode where the same process occurs. An
electron cascade develops which results in an amplification factor of at least 10^6 by the time the
electrons reach the anode. Photomultiplier tubes are very widely used but are more expensive than
phototubes due to the need for stabilizing circuitry for each dynode to ensure long-term operational
stability of the detector.
Diode array detectors consist of silicon integrated circuit (IC) chips incorporating up to one or two
hundred pairs of photodiodes and capacitors. Each photodiode measures about 0.05 × 0.5 mm and is
sensitive to