Hall-effect studies of modification of HgCdTe surface properties with ion implantation and thermal annealing

https://doi.org/10.1016/j.surfcoat.2020.125721Get rights and content

Highlights

  • Modification of the surface of HgCdTe films with arsenic implantation and annealing was studied.

  • Annealing was shown to annihilate structural defects induced by implantation.

  • High degree of activation of arsenic as a result of post-implantation annealing was achieved.

  • Activation of residual acceptor impurities by post-implantation annealing was discovered.

Abstract

Results of the Hall-effect studies of surface properties of n–type HgCdTe films modified with arsenic ion implantation and thermal annealing are reported on. A complete annihilation of implantation-induced extended defects (dislocation loops), quasi-point defects and related donor centers was observed as a result of a two-stage arsenic activation annealing. A high degree of activation of implanted arsenic was achieved with the annealing. In some cases, the annealing was found to lead to the modification of the properties of the ‘base’ layers not affected by implantation due to activation of uncontrolled acceptor defects and resulting changes in the degree of electrical compensation.

Introduction

HgCdTe solid solutions have been used in photodetectors operating in the long-wavelength infrared (LWIR) range for over 50 years [1]. The two most common technologies of HgCdTe-based photodiodes are n-on-p and p-on-n planar architectures. For these architectures, in the ‘base’ layer (p- and n-type, respectively), a highly doped area (n+- and p+-type, respectively) is fabricated with ion implantation. The n-on-p planar technology is the oldest one; these photodiodes are formed in p-type ‘base’ where mercury vacancies (VHg) serve as acceptors, and the n+-type doping is achieved with ion implantation of boron [2]. The use of VHg as the p-type intrinsic dopant, however, is known to degrade the electron lifetime due to the introduction of the centers of Shockley-Read recombination, so the resulting detectors exhibit high dark current. For this reason, p-on-n architectures have been gaining more and more interest in the last 20 years. Compared with the n-on-p technology, the n-type ‘base’ (active layer) is doped extrinsically, and the minority-carrier lifetime is governed by band-to-band CHCC recombination, providing the highest lifetimes possible. Additionally, the p-on-n technology decreases the serial resistance by means of employing high-mobility majority carriers (electrons) [3]. As a result, the p-on-n architecture allows for a decreased dark current, and thus, for improved operability at high temperature or for higher detector cut-off wavelength.

Typically, p+-on-n junctions are fabricated in HgCdTe with the use of arsenic ion implantation followed by thermal activation of implanted species [[4], [5], [6]]. Fabrication of such р+-on-n diodes appears to be more challenging than traditional boron-implantation n+-р technology. Ion implantation produces in the surface layer of HgCdTe both extended and quasi-point radiation defects (see, e.g., Refs. [7, 8]). Interstitial mercury atoms HgI released as a result of implantation damage interact with these defects and form two types of electrically-active donor centers [9,10], so irrespective of the dopant used, the as-implanted layer always exhibits n+-type conductivity. Thus, fabrication of р+-n structure requires an annealing, which should activate implanted arsenic and annihilate radiation defects while eventually maintaining original electrical properties of the n-‘base’. For an effective annealing, it is necessary to know the exact nature of the donor defects, their (and that of the dopant) location in the implanted material, and how the annealing affects their pattern. The most part of the earlier works was focused on structural studies of the implantation-damaged area and on its transformation under arsenic activation annealing [[3], [4], [5], [6], [7], [8]]. Detailed electrical studies, in their turn, were performed only on arsenic-implanted n+–p junctions formed in a p–type ‘base’ [9,10]. The purpose of this work was to study the effect of various types of thermal treatment on the properties of arsenic-implanted HgCdTe films and on the behavior of the implantation-induced donor defects in the n–type ‘base’.

Section snippets

Experimental details

The films were grown by molecular-beam epitaxy (MBE) at A.V. Rzhanov Institute of Semiconductor Physics (Novosibirsk, Russia) on (013) CdTe/ZnTe/Si substrates with the growth cycle controlled in situ by means of an ellipsometer [11]. The total thickness of the films was 7–9 μm. Their ‘absorber’ layers with CdTe molar fraction (composition) х ≈ 0.22 were in situ covered with graded-gap protective layers with composition ~0.46 at the surface. The thickness of the graded-gap layers was ~0.4 μm.

Optical reflectance

The structural perfection of the surface of the films and its modification by the implantation and annealings were studied with optical reflectance. In HgCdTe reflectance spectra in the visible range, two distinct peaks are typically observed, E1 and E1 + Δ1, which originate in transitions Λ4,5 → Λ6 and Λ6 → Λ6 [13,14]. The energy position and the shape of the peaks give information on the value of the energy gap (and, correspondingly, chemical composition) and the structural perfection of the

Conclusion

In conclusion, the results of the Hall-effect studies of modification of HgCdTe surface properties with arsenic ion implantation and thermal annealing showed that the formation of implantation-induced extended defects proceeds similarly in samples with p- and n-type conductivity of the photodiode ‘base’. As a result of activation annealing, a high degree of arsenic activation was achieved as well as complete annihilation of the extended defects (dislocation loops), quasi-point defects, and

CRediT authorship contribution statement

A.G. Korotaev: Conceptualization, Methodology, Writing - review & editing, Validation. I.I. Izhnin: Conceptualization, Methodology, Data curation, Writing - review & editing, Validation. K.D. Mynbaev: Conceptualization, Methodology, Writing - original draft, Validation. A.V. Voitsekhovskii: Conceptualization, Methodology, Writing - original draft, Validation. S.N. Nesmelov: Formal analysis, Writing - review & editing, Validation. S.M. Dzyadukh: Formal analysis, Writing - review & editing,

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (20)

  • I.I. Izhnin et al.

    Arsenic-ion implantation-induced defects in HgCdTe films studied with hall-effect measurements and mobility spectrum analysis

    Infr. Phys. Technol.

    (2019)
  • C. Lobre et al.

    Quantitative damage depth profiles in arsenic implanted HgCdTe

    Nucl. Instrum. Meth. B.

    (2013)
  • W. Lei et al.

    Progress, challenges, and opportunities for HgCdTe infrared materials and detectors

    Appl. Phys. Rev.

    (2015)
  • L.O. Bubulac et al.

    Boron and indium ion implanted junctions in HgCdTe grown on CdTe and CdTe/Al2O3

    J. Vac. Sci. Technol. A

    (1986)
  • L. Mollard et al.

    Status of p-on-n arsenic-implanted HgCdTe technologies

    J. Electron. Mater.

    (2011)
  • L. Mollard et al.

    Planar p-on-n HgCdTe FPAs by arsenic ion implantation

    J. Electron. Mater.

    (2009)
  • J.H. Park et al.

    Development of high-performance eSWIR HgCdTe-based focal-plane arrays on silicon substrates

    J. Electron. Mater.

    (2016)
  • C.Z. Shi et al.

    Barrier layer induced channeling effect of As ion implantation in HgCdTe and its influences on electrical properties of pn junctions

    Appl. Opt.

    (2016)
  • C. Lobre et al.

    Characterization of the microstructure of HgCdTe with p-type doping

    J. Electron. Mater.

    (2014)
  • O.Yu. Bonchyk et al.

    Nano-size defects in arsenic-implanted HgCdTe films: a HRTEM study

    Appl. Nanosci.

    (2019)
There are more references available in the full text version of this article.

Cited by (0)

View full text