No abstract

No abstract

No abstract

Passivant-Hg 1-x Cd x Te interface has been studied for the CdTe and anodic oxide (AO) passivants. The former passivation process yields five times lower surface recombination velocity than the latter process. Temperature dependence of surface recombination velocity of the CdTe/n-HgCdTe and AO/nHgCdTe interface is analyzed. Activation energy of the surface traps for CdTe and AO-passivated wafers are estimated to be in the range of 7-10 meV. These levels are understood to be arising from Hg vacancies at the HgCdTe surface. Fixed charge density for CdTe/n-HgCdTe interface measured by CV technique is 5 ϫ 10 10 cm Ϫ2 , which is comparable to the epitaxially grown CdTe films. An order of magnitude improvement in responsivity and a factor of 4 increase in specific detectivity (D*) is achieved by CdTe passivation over AO passivation. This study has been conducted on photoconductive detectors to qualify the CdTe passivation process, with an ultimate aim to use it for the passivation of p-on-n and n-on-p HgCdTe photodiodes.

The effect of fixed surface charge density (Q f ) on the performance of HgCdTe (MCT) photoconductive detectors is presented in this study. The figure of merit such as the specific detectivity (D * ) of Hg 0.78 Cd 0.22 Te photoconductive detectors operating at 77 K is calculated with 300 K background and 2π field of view (FOV). The equations leading to the above calculations are developed for the case of a one-dimensional model. Our calculations show that: (i) D * shows maxima with varying Q f ; (ii) improvement in D * due to accumulation is observed for values of Q f up to 5 × 10 10 cm −2 for materials with low as well as high bulk minority carrier lifetimes (τ b ); and (iii) D * is limited by shunt resistance due to passivants having Q f > 5 × 10 10 cm −2 for low-τ b materials. Furthermore, D * is found to degrade by an order of magnitude from its peak value with increase in Q f from 5 × 10 10 cm −2 to 1 × 10 13 cm −2 in a material having τ b = 10 ns, whereas no such degradation in D * is seen for the same range of Q f in a material having τ b ≥ 1 µs. In addition, our calculations predict that the accumulation layer improves D * by a factor of 1.5 for surfaces having an initial surface recombination velocity as high as 10 5 cm s −1 .

Mesa structures were etched in HgCdTe using different Br 2 /HBr/Ethylene glycol (EG) formulations. Etch rate and degree of anisotropy (A) were studied in detail for all of the combinations. Addition of EG to the conventional etchant gave A Ͼ 0.5, with controllable etch rates. Optimum etchant composition was determined to be 2% Br 2 in a 3:1 mixture of EG:HBr. This composition resulted in a good anisotropy factor of ϳ0.6 and a reasonably optimum etch rate of ϳ2.5 µm/min, with rms surface roughness of ϳ2 nm. Kinetics of the etching reaction have also been studied for the optimum etchant concentration and an etching mechanism has been proposed.

Generation-recombination (g-r) processes in the passivant/HgCdTe interface region are shown to complicate the transient photoconductive (PC) decays. Anomalous PC decays showing delayed peaks are observed and modeled for the first time. These peaks are correlated with the electron and hole traps in the interface region. Activation energy and density of the electron traps in the anodic oxide/n-Hg 0.78 Cd 0.22 Te interface region are estimated to be 12 meV and 1.1 × 10 10 cm -2 , respectively. Density of hole traps is estimated to be 2.4 × 10 10 cm -2 .

No abstract