Growth Of Semi-insulating InP Research Articles

Growth of semi-insulating InP with uniform axial Fe doping by a double-crucible LEC technique

The level of an Fe-doped InP melt in a growth crucible is kept constant throughout a Czochralski growth experiment by replenishing it with undoped melt from a second crucible. Since the distribution coefficient of iron is very small (about 0.001), the Fe concentration in the growth crucible is virtually unchanged during the pulling. As a result the ingot obtained exhibits axial Fe concentration much more uniform than in standard LEC crystals. This growth method can increase the yield of semi-insulating InP wafers as it prevents the accumulation of iron and the formation of precipitates in the last to freeze part of the crystal.

Ti doping of InP and GaInAs using TiCl 4

TiCl 4 is a very efficient source of Ti doping for InP grown by low pressure organometallic vapor phase epitaxy (OMVPE), and we report some characteristics of the doping behavior. Precipitates are visible in the material (TEM observation) for Ti concentrations > 10 18 cm -3, although morphological problems only appear for concentrations above 10 19 cm -3. As reported by other authors, a delay in incorporation is seen, and attributed to the reaction of TiCl 4 with oxygen or oxygen-containing compounds adsorbed on the reactor walls. Also as reported before, we find that the level of electrical activity of Ti in InP stays low, although it is sufficient to allow the growth of semi-insulating InP:Zn + Ti or InP:Fe + Ti. TiCl 4 does not introduce Ti in AlInAs, but Ti doping of GaInAs is obtained, and can render the material highly resistive. The incorporation of Zn is seen to be influenced (slightly increased) by the introduction of Ti.

Growth of semi-insulating InP: Fe in the low pressure hydride VPE system

Growth of InP:Fe has been performed in a low pressure hydride VPE system at a total pressure of 2000 Pa. As a dopant, iron chloride obtained from an elemental iron source reacting with gaseous HCl has been used. Both the source reaction and the growth have been performed in a nitrogen ambient. Fe doping levels between 10 16 and approximately 10 20 cm -3 could be adjusted by varying the HCl flow over the iron source. The growth temperature has only a slight effect on the Fe incorporation mechanism. Additionally, the Fe incorporation is found to depend on the InP growth rate. I–V measurements reveal semi-insulating behaviour for samples with iron concentrations above 10 17 cm -3, leading to specific resistivities in the 10 8 Ω cm region.

Growth of semi-insulating InP by GSMBE

Fe doped semi-insulating InP layers have been grown by gas source MBE with a solid iron source. Structure as n-i-n, p-i-n and p-n-i-n were characterized by I( V) measurements and secondary ion mass spectroscopy profiling (SIMS). As shown by SIMS, uniform Fe doping and abrupt transitions are achieved for the different structures studied. Resistivities as high as 1.5×10 9 Ω cm are determined from I( V) curves for Fe concentrations in the 10 17 cm -3 range. Lasers with semi-insulating layers have been realized for the first time by gas source MBE. Preliminary results show power emission of 43 mW, without antireflecting coating, comparable to state-of-the-art characteristics.

Method for reliable liquid phase epitaxial growth of semi-insulating InP doped with Ti and Zn

Routine growth of semi-insulating (Ti:Zn)InP with resistivity values between 4×104 and 105 ω cm has been achieved using liquid phase epitaxy. A consistent bakeout procedure for In followed by p-doping using p+-InP source material results in p- epitaxial layers with Na - Nd adjustable up to 2.0×1016 cm-3. Addition of Ti and Zn, followed by crystal growth in the absence of O2, gives the desired semi-insulating InP epitaxial layers at a growth temperature of only 635°C. This growth temperature is 220°C less than that required for InP:Fe, making (Ti:Zn)InP a potentially suitable current confining layer for buried heterostructure lasers.