FACTORS CONTROLLING THE RESISTANCE OF OHMIC CONTACTS TO GERMANIUM TELLURIDE

Open Access
Author:
Aldosari, Haila Mohammed
Graduate Program:
Materials Science and Engineering
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
June 13, 2016
Committee Members:
  • Suzanne E Mohney, Dissertation Advisor
  • Suzanne E Mohney, Committee Chair
  • Joan Marie Redwing, Committee Member
  • Patrick M Lenahan, Committee Member
  • S Ashok, Outside Member
Keywords:
  • phase change material (PCM)
  • GeTe
  • Mo
  • Ni
  • Au
  • contact
  • X-ray photoelectron spectroscopy (XPS)
  • Transmission electron microscopy (TEM)
Abstract:
Ohmic contacts with extremely low resistance and controlled morphologies are required to improve the performance of GeTe-based radio frequency (RF) switches. These contacts also require good thermal stability, since GeTe is a phase change material that switches back and forth between the amorphous and crystalline states through controlled Joule or laser heating. In this thesis, the role of metal selection, pre-metallization surface treatments, and post-metallization annealing on the resistance of Ohmic contacts to GeTe is examined. X-ray photoelectron spectroscopy (XPS) was used to characterize surfaces of polycrystalline α-GeTe films after different surface treatments, including UV-O3, in-situ Ar+ ion etching, deionized water (DI H2O), ammonium sulfide (NH4)2S, and hydrochloric acid (HCl), in an effort to understand the effect of pre-metallization surface treatments on the resistance of different metal/GeTe contacts. The combination of UV-O3 and subsequent oxide removal procedures affect the GeTe surface stoichiometry, as does Ar+ plasma without UV-O3 treatment. Ar+ plasma, DI H2O, and (NH4)2S preparations leave the GeTe surface richer in Ge compared to the HCl treatment, which leaves the surface Te-rich. The thesis focuses on contacts based on three different metals: Ni, Au, and Mo. The trend in the as-deposited specific contact resistance (ρc) as a function of surface treatments demonstrated a strong dependence on the choice of the metal used. For Ni-based contacts, different surface treatments resulted in ρc ≈ 4 × 10-8 Ω.cm2, except for the HCl treatment, which resulted in contacts with much higher resistance, due to formation of a 5nm thick layer of Ni1.29Te at the interface. A constant ρc ≈ 1.2 × 10-8 Ω.cm2 was found for Au (100 nm) contacts, regardless of the surface treatment used. For Mo-based contacts, in-situ Ar+ plasma treatment provided the lowest ρc ≈ 7 ± 1 × 10-9 Ω.cm2. These measurements were made using a refined transfer length method (RTLM) test structure; we found that our circular transfer length test structure (CTLM) provided inaccurate results due to the effect of the metal sheet resistance on the extraction of such very low values of specific contact resistance. We discuss how phase diagrams can be used to understand interfacial reactions and predict which metallizations could provide long-term thermal stability. As expected from our calculated phase diagram, Auger electron spectroscopy (AES) and cross-sectional transmission electron microscopy (TEM) revealed that reaction occurs between Ni and GeTe, even at room temperature. After annealing, there is an increase in contact resistance along with serious morphological changes in the metallization. Au contacts are unreactive on GeTe, even after annealing at 350 °C for 30 min. However, volatilization of tellurium still occurs through openings that appear in the gold film. Mo-based contacts may be reactive, but the kinetics of the reaction are much slower than the Ni case. An increase in ρc values was observed after annealing, regardless of the metallization. The similarity of as-deposited contact resistances obtained with very different metal work functions points to Fermi level pinning at the surface of GeTe. Moreover, the low values of the specific contact resistances correspond to low Schottky barrier heights. Finally, this thesis points to possible routes to maintain the lowest achievable ρc without loss of thermal stability.