ITEM METADATA RECORD
Title: Design, Fabrication and Interface Characterization of High Mobility InGaAs Channel MOS(FET) Devices (Ontwerp, fabricatie en interface karakterisatie van hoge mobiliteit InGaAs kanaal MOS transistoren)
Other Titles: Design, Fabrication and Interface Characterization of High Mobility InGaAs Channel MOS(FET) Devices
Authors: Alian, Alireza; S0193335
Issue Date: 15-Oct-2012
Abstract: The state-of-the-art knowledge about the passivation as well as the historical advancements for realizing III-V channel MOSFETs are discussed.Starting with MOSCAPs being the simplest electrically measurable gate stack devices, the influence of various chemical and thermal treatments on the electrical characteristics of the MOSCAPs were investigated. Besides slightly reduced accumulation frequency dispersion in the C-V response of the n-type MOSCAPs, sulfur passivation of the surface through aqueous (NH4)2S treatment (being a very well-known III-V surface passivation scheme) was found to even degrade the interface state density compared to a simple HCl surface treatment. NH4OH treatment of the surface, another reported passivating solution, did not improve the interface characteristics neither. Issues influencing the quantification of the interface state density in the InGaAs MOSCAP were discussed.A fabrication process flow was developed to make long channel MOSFET transistors in order to study the influence of the surface treatments on the transistor characteristics. A MOSFET platform also mitigates part of the quantification issues associated with the interface state density. In the effort to make such MOSFET devices, a digital etch process capable of controlled, precise and smooth etching of the InGaAs material was also developed and characterized. Except for the epitaxial growth of the layers, developments associated with the MOSFET processing flow as well as the complete fabrication of the MOSFET devices investigated in this thesis was performed as part of the PhD work.Electrical characterization of the fabricated InGaAs channel MOSFETs demonstrated the importance of hydrogen annealing in improving the interface properties by reducing the interface state density from above 4×1013 cm−2eV−1 to about 4×1012 cm−2eV−1 in the midgap of InGaAs. Further investigations on sulfur passivated MOSFETs revealed that sulfur treated devices exhibit 3 to 4 times higher drive current than the HCl treated devices. The degraded interface state density after sulfur passivation translated into degraded sub-threshold characteristics of the sulfur treated MOSFETs. The higher drive current of the sulfur treated MOSFET despite its higher interface state density compared to the HCl treated device confirms the donor-like nature of the midgap interface states making them neutral during the on-state operation of the device.A complete characterization of the bulk oxide defects applying the TSCIS oxide trap characterization method revealed that the defect density inside the gate oxide (here Al2O3) is significantly influenced by the type of the surface chemical treatment prior to the gate oxide deposition. Sulfur passivation of the surface was found to decrease the density of the oxide traps by a factor of 3 to 4 compared to that of the device with HCl surface treatment. The energetic position of these oxide traps was found to be at the conduction band edge of InGaAs, therefore these traps are accessible to the electrons of the conduction band of InGaAs. Consequently, the increase in the drive current of the device with sulfur passivation over the HCl treated one was attributed to the increased electron mobility inside the channel of the device through the reduction of Coulomb scattering of the current carriers due to the charged traps inside the oxide.TOFSIMS characterization of the bulk of the Al2O3 revealed significantly higher (about 4 times) indium concentration inside the oxide with HCl surface pretreatment compared to that of the oxide with (NH4)2S surface pretreatment. Based on the results, sulfur treatment reduces the indium diffusion/segregation from the InGaAs into the Al2O3 layer. Correlating the TOFSIMS data and the TSCIS results suggests indium as a possible origin of traps inside the oxide.A new type of sulfur treatment of the InGaAs surface using the vapor from the aqueous (NH4)2S solution was investigated and MOSFET characteristics identical to the aqueous (NH4)2S treated MOSFETs were measured. Finally, the influence of the surface potential fluctuations on the simulated C-V response of the p-type InGaAs MOSCAP devices was explored. It was found that the observed stretch out in the experimental C-V characteristic of the p-type InGaAs MOSCAPs as well as their large accumulation capacitance frequency dispersion can be originating from the surface potential fluctuations caused by the charged interface traps. It was also found that the surface potential fluctuations can have a significant impact on the conductance characteristic of the devices. The surface potential fluctuations can limit the measurable Dit energy range within the bandgap of the semiconductor and neglecting the surface potential fluctuations results in underestimation of the interface state density.
Table of Contents: Introduction
1.1 Goal and approach of this study
2 Literature review
2.1 Interface states and surface passivation
2.1.1 Interface state models
2.1.2 In situ oxide deposition
2.1.3 Atomic Layer Deposition (ALD) self-cleaning effect
2.1.4 Density functional theory (DFT) simulations
2.1.5 Sulfur passivation
2.1.6 Oxide traps
2.2 GaAs
2.3 InGaAs
2.4 State-of-the-art
3 Experimental techniques
3.1 Device fabrication flows
3.1.1 MOSCAP fabrication flow
3.1.1.1 Surface treatment
3.1.2 MOSFET fabrication flow
3.1.3 Source/drain isolation etch (gate recess)
3.1.3.1 Using an etch-stop layer
3.1.3.2 Timed etching of n+InGaAs
3.1.3.3 Digital etching
3.1.4 Deposition tools
3.1.4.1 Atomic Layer Deposition (ALD)
3.1.4.2 Molecular Beam Epitaxy (MBE)
3.1.4.3 Metal Organic Chemical Vapor Deposition (MOCVD)
3.2 Electrical characterization methods
3.2.1 The conductance method and Dit extraction
3.2.1.1 Difficulties in Dit quantification
3.2.2 Trap Spectroscopy by Charge Injection and Sensing (TSCIS)
3.3 Physical characterization tools
3.3.1 X-ray Photoelectron Spectroscopy (XPS)
3.3.2 Time Of Flight Secondary Ion Mass Spectroscopy (TOFSIMS)
4 InGaAs MOSCAP characterization
4.1 Electrical characterization
4.1.1 HCl versus (NH4)2S treatment
4.1.2 NH4OH treatment
4.1.3 Impact of FGA
4.2 Physical characterization
4.2.2 The native oxide growth rate
4.3 Conclusions
5 InGaAs MOSFET characterization
5.1 Impact of FGA on MOSFET characteristics
5.1.1 C-V characterization
5.1.2 I-V characterization
5.2 Impact of the surface pretreatment
5.2.1 C-V characterization
5.2.2 I-V characterization
5.3 Conclusions
6 Ammonium sulfide vapor (ASV) surface treatment
6.1 Device description
6.2 Electrical characterization
6.2.1 C-V characterization
6.2.2 I-V characterization
6.2.2.1 Off-state characteristics
6.2.2.2 On-state characteristics
6.3 XPS characterization
6.4 Conclusions
7 Oxide traps
7.1 Device Characterization
7.2 Conclusions
8 Understanding the C-V response of the p-type InGaAs MOSCAP
8.1 Influence of the surface potential fluctuations on the C-V characteristic
8.2 Influence of the surface potential fluctuations on the conductance characteristic
8.2.1 Dit extraction in presence of surface potential fluctuations
8.3 Conclusions
9 Conclusions
9.1 Future work
Bibliography
List of publications
Curriculum Vitae
ISBN: 978-94-6018-577-9
Publication status: published
KU Leuven publication type: TH
Appears in Collections:Surface and Interface Engineered Materials
Associated Section of ESAT - INSYS, Integrated Systems

Files in This Item:
File Status SizeFormat
Thesis_alirezaAlian_Final_BW.pdf Published 11698KbAdobe PDFView/Open

These files are only available to some KU Leuven staff members

 


All items in Lirias are protected by copyright, with all rights reserved.