One of the most important technical advances in fiber optic telecommunication systems has been the optical fiber amplifier (OFA). The use of an all-optical fiber optic amplifier instead of a semiconductor optical amplifier has many inherent advantages. For example, the most popular erbium-doped fiber amplifiers (EDFAs) have a broad range of optical amplification in the 1550-nm window; these erbiumdoped fibers are easy to incorporate directly into an allâfiber optic telecommunication system. Today the erbium-doped silica fibers are the most common rare-earth (RE)-doped fiber amplifiers, as they have high gain and low noise. EDFAs are quite versatile in that they may be used as preamplifiers, in-line and booster amplifiers, as well as sources. They can provide more than 20 dB of gain across a broad wavelength band, which enables them to be extremely useful as components in dense wavelength-division multiplexed (DWDM) systems. Additionally, they may be pumped with 800-, 980-, and 1480-nm sources and their quantum efficiency can be quite high.
There are many excellent references to OFAs, including several books on EDFAs. In addition there has been a plethora of RE ions doped into silica glass, and their emission spectra and use as OFAs have been thoroughly studied. Partly as a consequence of the success of rare-earth active oxide glass fibers, it was natural to study the introduction of some of the same RE ions used so successfully in oxide glasses to dope the newer fluoride glasses. The motivation behind this concept was to take advantage of the lower phonon frequencies of the fluorides compared to those of oxide glasses. That is, an IR transparent glass like the HMFGs will have fewer phonons to contribute to nonradiative processes. Nonradiative transitions take energy away from the desired radiative or laser transitions and, therefore, lead to lower quantum efficiencies. It would seem, then, that merely choosing a host glass with a lower phonon frequency would lead to a much more efficient amplifier. Unfortunately, there are other considerations that have limited the use of the IR fiber hosts as a replacement for the oxide glass EDFAs. In fact, today the best application of the fluoride or other IR fiber hosts is for amplification in the 1300-nm telecom band.