Technology and characterization of GaN-based heterostructure field effect transistors (HFETs) [Elektronische Ressource] / Michael Fieger

Informations

• Publié par: rheinisch-westfalischen_technischen_hochschule_-rwth-_aachen
• Publié le: 01 janvier 2011
• Nombre de lectures: 25
• Langue: English
• Poids de l'ouvrage: 2 Mo

Extrait

Technology and Characterization
of GaN-based
Heterostructure Field Effect
Transistors
(HFETs)
Von der Fakultät für Elektrotechnik und Informationstechnik
der Rheinisch-Westfälischen Technischen Hochschule Aachen
zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften
genehmigte Dissertation
vorgelegt von
Diplom-Physiker
Michael Fieger
aus Aachen
Berichter: Prof. Dr.-Ing. Andrei Vescan
Prof. Dr. rer. nat. Hans Lüth
Tag der mündlichen Prüfung: 29. Juli 2010
Diese Dissertation ist auf den Internetseiten der Hochschulbibliothek online verfügbar.Contents
1 Introduction and Motivation 2
2 Theoretical Background 4
2.1 Material properties of the Group III nitrides . . . . . . . . . . . . . . . . . . . . 4
2.2 Crystal structure and polarization effects of the Group III nitrides . . . . . . . . . 6
2.3 Heterostructures and 2DEG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.4 Heterostructure field effect transistors (HFETs) . . . . . . . . . . . . . . . . . . 12
2.4.1 Metal-semiconductor contacts . . . . . . . . . . . . . . . . . . . . . . . 13
2.4.2 Principle of HFETs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4.3 Static DC I-V characteristics . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4.4 Pulsed DC I-V characteristics . . . . . . . . . . . . . . . . . . . . . . . 19
2.4.5 Small-signal characterization . . . . . . . . . . . . . . . . . . . . . . . . 20
2.4.6 Load-pull measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4.7 Electrical limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3 State-of-the-art GaN-based HFETs 29
4 Development of an HFET baseline process 32
4.1 Mesa isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.2 Ohmic contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.3 Schottky contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.4 Surface passivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.4.1 DC characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.4.2 Large-signal characterization . . . . . . . . . . . . . . . . . . . . . . . . 54
5 AlGaN/GaN HFETs 58
5.1 Variation of the aluminium concentration in the AlGaN barrier layer . . . . . . . 58
5.1.1 Transport properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.1.2 DC characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.1.3 Small-signal characterization . . . . . . . . . . . . . . . . . . . . . . . . 62
5.1.4 Pulsed DC characterization . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.2 Metal-Insulator HFETs (MISHFETs) . . . . . . . . . . . . . . . . . . . . . . . . 66
5.2.1 DC performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.2.2 RF performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.3 Early surface passivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2Contents
5.3.1 DC characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.3.2 Load-pull measurements . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6 AlInN/GaN HFETs 79
6.1 Variation of the aluminium concentration in the AlInN barrier layer . . . . . . . 80
6.1.1 Sample structure and processing . . . . . . . . . . . . . . . . . . . . . . 80
6.1.2 Transport properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.1.3 DC characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
6.1.4 Small-signal performance . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.1.5 Pulsed I-V characteristics . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.2 Lattice-matched HFETs with a thin (10 nm) barrier layer . . . . . . . . . . . . . 91
6.2.1 Transport properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
6.2.2 DC characteristics of unpassivated HFETs . . . . . . . . . . . . . . . . . 93
6.2.3 Small-signal measurements of unpassivated devices . . . . . . . . . . . . 96
6.2.4 Influence of aN -based SiN surface passivation on device performance . 982
6.3 Impact of post-gate annealing processes on the DC performance of HFETs . . . . 100
6.3.1 Schottky contact characteristics . . . . . . . . . . . . . . . . . . . . . . 101
6.3.2 Transistor performance . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
6.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
7 Summary and Outlook 106
Appendix 109
Bibliography 110
Acknowledgement 123
11 Introduction and Motivation
The years 1992 and 1993 played, without doubt, an essential role in the development of Group
III nitride-based optoelectronic and electronic devices. On the one hand, in 1992, Nakamura [90]
was the first to find a way of realizing p-type doping for GaN. This invention can be seen as a key
date for applications in optoelectronic devices. Since then, LEDs and LDs have become possi-
ble in a spectral range which could not be employed using conventional semiconductors like Si,
(Al)GaAs or AlInGaP. Especially, the blue and white LED have continued their incomparable
triumph ever since. On the other hand, in 1993, Khan [59] demonstrated a heterostructure field
effect transistor based on an AlGaN/GaN layer structure for the first time. A two-dimensional
electron gas (2DEG) in combination with superior material properties in terms of large bandgap,
high breakdown voltage and high saturation carrier velocity was an ideal precondition for appli-
cation in electronic devices.
In the field of these devices, e.g. transistors needed for high-voltage and high-frequency appli-
cations, GaN-based HFETs have become increasingly important. Because the Group III nitrides
are strongly polar materials, the generation of a 2DEG is possible even without any doping in
the barrier layer. An additional strain resulting from growing lattice-mismatched AlGaN on GaN
induces a piezoelectric charge which supplies further electrons to the HFET channel. This total
13 2channel charge can top 1×10 electrons/cm - roughly four to five times higher than for Al-
GaAs/GaAs HFETs. From the demonstration of the very first AlGaN/GaN HFET until today’s
state-of-the-art devices, tremendous progress has been made. However, until today the great
potential of the GaN-based HFET technology has not yet been fully exploited and the commer-
cialization remained rather marginal. The reason is that there are many fields which still need
further improvement:
• Due to the lack of large-size area native GaN substrates, heteroepitaxy still depends on
substrates like sapphire, SiC or silicon. Related to these non-native substrates, issues like
lattice mismatch and different thermal expansion coefficients are still existing.
• DC-to-RF dispersion effects and current collapse represent a major challenge to the com-
mercialization of the technology. These effects are mainly attributed to traps located at the
2semiconductor surface. Therefore, a passivation of the surface is inevitable. SiN passiva-
tion films have been found to mitigate those effects.
• Schottky contacts on GaN-based heterostructures suffer from detrimental leakage current
problems. A thin dielectric layer sandwiched between the gate metal and the semiconduc-
tor surface seems to be a promising approach in order to alleviate the leakage problem.
Whereas the first issue is closely related to epitaxial design and substrate engineering, the other
issues are associated with the device processing technology.
This thesis was targeted to develop and establish a baseline process for realizing GaN-based
HFETs. Each single step of this process was optimized in order to improve device performance.
AlGaN/GaN as well as AlInN/GaN transistors were investigated in terms of their DC and RF
characteristics. Critical issues, as mentionend above, were discussed in detail. As a result,
a deeper physical understanding of the non-ideal behavior of GaN-based transistors could be
obtained.
This thesis is organized in the following way: Chapter 2 describes the theoretical framework
of the GaN material system including crystal structure and polarization effects. Based on the
description of the 2DEG the key topics of HFETs were discussed.
Chapter 3 gives an overview of the current status of the market situation of commercially
available GaN-based HFETs.
Chapter 4 focuses on the development and optimization of the process technology suitable for
realizing GaN-based transistors.
Chapter 5 and chapter 6 discuss in detail the characterization of AlGaN/GaN and AlInN/GaN
HFETs.
Finally, chapter 7 provides a retrospection of the conclusion which can be drawn from the
results of this thesis.
32 Theoretical Background
2.1 Material properties of the Group III nitrides
Wide bandgap materials, particularly the Group III nitrides like GaN, AlGaN, InGaN etc., have
attracted much attention due to various potential advantages compared to the conventional semi-
conductors. These advantages arise from the basic physical properties of the material associated
with the crystal structure. The direct bandgap of these materials and their alloys provides the pos-
sibilty to be used not only for electronic applications but also for optical applications