High speed infrared detector

#This invention relates to a low carrier concentration, high speed infrared detector and to a method for its manufacture. More specifically, this invention concerns itself with the development of a high speed, low carrier concentration PbSnTe photodiode that possesses a high quantum efficiency as a 10.6 .mu.m infrared detector.

#Infrared photodiodes have been found to be particularly useful as receivers for 10.6.mu.m carbon dioxide laser communication systems. As a result, a considerable research effort has evolved in an attempt to develop a high quantum efficiency infrared detector with a frequency response beyond one gigahertz and an operating temperature of 77.degree. K. Infrared photodiodes with a frequency response beyond 1 GHz must be of PIN structure in order to reduce the RC time to an acceptable value. However, high efficiency PbSnTe photodiode devices require a device design in which light is brought into the p-n junction without attenuation.

#In order to accomplish such a design, it has been found that a heterojunction device in which the n-type material is transparent to 8 - 14.mu.m radiation eliminates the effect of properties of the n-type layer on efficiency. For the high speed device of this invention, the n-layer serves as both a window to the incident radiation and as the wide depletion width required to achieve gigahertz bandwidths.

#In the drawings:

#FIG. 1a represents a graphical illustration of a phase diagram for the metal (Pb or Sn)--Te; while

#FIG. 1b discloses the region of stoichiometry of FIG. 1a in expanded form; and

#Pursuant to the above-defined objects, the present invention provides a heterojunction photodiode which is especially useful as a receiver in 10.6.mu.m laser communication systems. In order to operate effectively, and with high efficiency, and frequency response beyond 1 gigahertz, infrared photodiode detectors must be of the PIN structure. The fabrication of such a device, therefore, requires the construction of a heterojunction in which the n-type material is transparent to the 10.6.mu.m radiation in order to eliminate the effect of the properties of the n-layer on efficiency.

#To further understand the details of the present inventon, reference is made to the phase diagram of FIGS. 1a and 1b which illustrates the equilibrium relationship between the Pb--Sn--Te liquidus and solidus. For growth from a stoichiometric melt, a Te-rich solid results and is heavily p-type with two holes for every excess Te atom (.about.10.sup.19 cm.sup.-3 typical carrier concentration). With growth from a metal-rich melt, an appropriate reduction in a number of Te vacancies occurs in the solid and for the proper metal-Te ratio a stoichiometric solid results. With growth from a melt with an increasing metal-Te ratio, a solid will result that has a deficit of Te, i.e., n-type material.

#Low carrier concentration liquid phase epitaxial layers can be grown by a technique which can be understood by referring to FIG. 1b. Growth of low carrier concentration material can be obtained by growing straight from a to d where a to b results in p-type material and c to d results in n-type material. An intrinsic layer should result between p and n type regions during the growth from b to c. The temperature at which this region occurs is known as the "crossover" temperature. All of this material would be of the same energy gap and thus a homojunction would result. Alternatively, an n-type PbTe layer (grown from c' and d') could be grown over a p-type layer (grown from a to b) to provide a heterostructure. Similarly, a different Pb.sub.1-y Sn.sub.y Te layer could be grown c" to d" over a Pb.sub.1-x Sn.sub.x Te layer grown from a' to b' to achieve a heterostructure. Carrier concentrations as low as 1.times.10.sup.14 cm.sup.-3 have been obtained using these techniques. A problem exists, however, in achieving a low carrier concentration and a high Sn content simultaneously in a single layer. Low carrier concentrations are of course necessary if we are to obtain low capacitance devices. The high Sn content is necessary if we are to keep the 10.6.mu.m absorption coefficient near 10.sup.4 cm.sup.-1, which will ensure that the slow process of carrier diffusion is eliminated. The requirement that both of these conditions be met simultaneously present considerable difficulty in terms of growing near the crossover temperature. As one adds increasing amounts of Sn to Pb.sub.1-x Sn.sub.x Te, the crossover temperature is reduced such that it becomes increasingly difficult to grow any sizeable volume of low carrier concentration material at that low temperature. For example the x value to obtain a peak spectral response at 10.6.mu.m is near 20% and the crossover temperature for this composition is too low to provide thick layers of low carrier concentration material.