Advances in Imaging and Electron Physics (eBook)

eBook Download: EPUB

1999 | 1. Auflage
453 Seiten
Elsevier Science (Verlag)
978-0-08-057775-3 (ISBN)

Infos

Advances in Imaging & Electron Physics merges two long-running serials--Advances in Electronics & Electron Physics and Advances in Optical & Electron Microscopy. The series features extended articles on the physics of electron devices (especially semiconductor devices), particle optics at high and low energies, microlithography, image science and digital image processing, electromagnetic wave propagation, electron microscopy, and the computing methods used in all these domains.

Inhalt

Front Cover 1
Advances in Imaging and Electron Physics, Volume 109 4
Copyright Page 5
Contents 6
Contributors 10
Preface 12
Forthcoming Contributors 14
Chapter 1. Development and Applications of a New Deep Level Transient Spectroscopy Method and New Averaging Techniques 20
I. Introduction 21
II. Review of the Deep-Level Transient Spectroscopy Method 25
III. Averaging and Recording of Digital DLTS Transient Signals 58
IV. Feedback Circuits and Experimental Setup for CC-DLTS and CR-DLTS 80
V. Constant-Resistance DLTS in Enhancement Mode MOSFETs 91
VI. Constant-Resistance DLTS in Depletion Mode MOSFETs 102
VII. Constant-Resistance DLTS in Junction Field-Effect Transistors 115
VIII. Conclusions and Areas for Future Research 138
References 141
Appendix A: Magnitude Errors 154
Appendix B: Time Constant Errors 159
Appendix C: Noise Sources and Signal-to-Noise Ratio (SNR) in the DLTS Transients 160
Appendix D: Electrical Circuit of the Pseudo-Logarithmic Generator 164
Appendix E: Electrical Circuits of the Feedback Circuit 165
Appendix F: Listing of a Template for a DLTS Measurement Program 167
Appendix G: Listing of a Template for a DLTS Analysis Program 171
Appendix H: Radiation-Induced Defects in Silicon 174
List of Acronyms 175
List of Symbols 177
Chapter 2. Complex Dyadic Multiresolution Analyses 182
I. Introduction 182
II. The Spline Example 183
III. Multiresolution and Wavelet 186
IV. Daubechies’ Wavelets 189
V. Symmetric Daubechies Wavelets 194
VI. The Phase of SDW Scaling Function 195
VII. The Mallat Algorithm with Complex Filters 197
VIII. Restoration from the Phase 199
IX. lmage Enhancement 202
X. Complex Shrinkage 206
XI. Conclusion 212
References 215
Reading List 216
Chapter 3. Lattice Vector Quantization for Wavelet-Based Image Coding 218
I. Introduction 219
II. Quantization of Wavelet Coefficients 220
III. Lattice Quantization Fundamentals 232
IV. Lattices 237
V. Quantization Algorithms for Selected Lattices 248
VI. Counting the Lattice Points 252
VII. Scaling Algorithm 261
VIII. Selecting a Lattice for Quantization 263
IX. Entropy Coding of Lattice Vectors 269
X. Experimental Results 273
XI. Conclusions 277
Appendix A: Cartan Matrices of Some Root Systems 279
References 280
Chapter 4. Fuzzy Cellular Neural Networks and Their Applications to Image Processing 284
I. Introduction 285
II. Fuzzy Cellular Neural Networks 294
III. Theory of Fuzzy Cellular Neural Networks 309
IV. FCNN as Computational Arrays 346
V. Embed Linguistic Statements into FCNN 379
VI. Learning Algorithms of FCNN 394
VII. Generic Algorithm for FCNN 420
VIII. Applications of Discrete-Time FCNN 426
IX. Conclusions and Future Work 439
References 446
Index 466

Textauszug

Development and Applications of a New Deep Level Transient Spectroscopy Method and New Averaging Techniques

Plamen V . Kolev; M. Jamal Deen    School of Engineering Science, Simon Fraser University, Vancouver, British Columbia, Canada V5A 1S6
Department of Electrical and Computer Engineering, McMaster University, Hamilton, Ontario, Canada L8S 4K1

I INTRODUCTION

A Importance of Impurity Characterization

The rapid advances in semiconductor technology during the last few decades are closely related to the success in achieving significant increase in semiconductor material purity (Sze, 1983)1. The ability to detect and measure the properties of a very small amount of impurity atoms of structural defects in the semiconductor material and in the active regions of semiconductor devices is of fundamental importance for this progress. The electrical properties of these impurities or defects are of particular interest for both the performance (Kwan and Deen, 1998; Raychaudhuri et al., 1996a, b; Kwan et al., 1997) and reliability (Kwan et al., 1996; Raychaudhuri et al., 1996b; Deen and Raychaudhuri, 1994; Deen and Quon, 1991) of semiconductor devices (Graff, 1995; Pantelides, 1992; Deen and Zhu, 1993e; Sze, 1983, Zhu et a., 1992).

Shallow impurities in semiconductors generally contribute extra free carriers, electrons, or holes. By intentionally incorporating shallow impurities in the semiconductor material, the type and magnitude of the conductivity of the material is controlled. The properties of the shallow impurities related to the host semiconductor material are considered to be well understood. Still, even for the best-known semiconductor, silicon, some details of the interaction of the shallow impurities with the host atoms were only recently found (Karasyuk et al., 1994).

Other imperfections in the crystal structure, such as other impurity atoms, lattice point defects and impurity-defect complexes, are referred to as deep centers. Their role is primarily to control the generation, recombination, and lifetime of the current carriers. Despite the significant progress in the last two decades, deep centers have proven to be far more difficult to investigate than shallow impurities. In many cases, the physical nature of the center causing the appearance of a deep level is poorly understood or unknown (Pantelides, 1992).

B Deep-Level Transient Spectroscopy

Deep-level transient spectroscopy (DLTS) (Lang, 1974a) is a well-established research technique for characterization of electrically active centers deep in the semiconductor bandgap (Blood and Orton, 1992; Schroder, 1990). It is known for its high sensitivity and direct relation to the measured properties of the defects. In the last two decades, many variations of the method were developed and adapted for studying the defect properties of a variety of materials and devices. Still, new modifications and further improvements of the already existing DLTS techniques continue to be reported (Anand et al., 1992; Bosetti et al., 1995; Chretien et al., 1995; Doolittle and Rohatgi, 1992; Hacke and Okushi, 1997; Istratov and Vyvenko, 1995; Lossen et al., 1996; Martin, 1995; Ozder et al., 1996; Rancour, 1995; Shaban, 1996).

Despite the large variety of modifications, the method is not yet accepted as a standard characterization technique in the semiconductor industry. There are many reasons for this limited acceptance. First, in order to extract the properties of the traps, there is need for a wide variation in sample temperature. Second, a standard describing the settings and parameters of the measurement instrumentation has not been established. Third, the wide variety of test structures and the dependence of the signal magnitude on the size of the test device prevents the establishment of a standardized approach that is convenient for industrial applications. With the continued scaling-down in device dimensions in semiconductor integrated circuits and the emergence of new technologies, such as SiGe heterojunction bipolar technology, silicon-on-insulator (SOI), porous silicon, thin-film transistors, or copper metallization for VLSI and UL- SI, the importance of accurate measurement and control of the defects that introduce deep levels will progressively increase. Therefore, steps toward further improvement and refinement of the DLTS method and instrumentation have important practical applications.

C Goals of This Chapter

This chapter first describes new and improved digital techniques for transient data processing that offer better sensitivity and more effective data storage. This approach is applicable for all variations of DLTS and can be easily adapted to other experiments involving recording and analysis of transient signals. Also, it opens up opportunities for further development of the group of isothermal DLTS techniques that rely on the analysis of the transient decay for extraction of the characteristic time constant and not on the thermal scan. Second, a new feedback circuit is described that allows for fast and sensitive operation in one very attractive and technically challenging variation of the method — constant capacitance (CC-)DLTS. It is important to note that this variation produces a signal with a magnitude that is independent of the area of the device and, therefore, the method is better suited for routine parametric control in the semiconductor industry. Third, a new technique, termed constant-resistance (CR-)DLTS is presented. This new technique is well-suited for measurements of field-effect transistors (FETs) regardless of their size and without compromise to the sensitivity of the measurement. Unlike CC-DLTS, because the sensitivity is dependent of the gain of the transistor (thus, of the channel width-to-length ratio and not of the active area), it allows for sensitive measurement of very small, deep-submicron devices, which are the basic transistors in the advanced microelectronics circuits and systems.

For corroboration of this technique, the results are compared to those obtained from conventional DLTS and CC-DLTS measurements. While the method has been applied to three different types of field-effect transistors, it can be easily used for a wider range of FETs. Illustrations are made with measurement of proton- and neutron-induced damages in metal-oxide- semiconductor (MOS) FETs and silicon and germanium junction FETs. The possibilities for measurement of interface trap density in the active interface of regular MOSFETs and for defect profiling using the new technique are also demonstrated.

In this chapter the emphasis is on the development of semiconductor metrology and instrumentation using mixed analog, digital and software engineering. The experimental results are used mainly for illustration of the new techniques and are not self-contained and complete studies.

D Organization of the Chapter

In Sect. II, the basics of the DLTS method are introduced and various techniques and instrumentation are described. The existing large variety of DLTS makes any attempt at classification a very complex task. Nevertheless, an attempt to define the criteria that can be used for classification of the DLTS techniques is made. Following these criteria, a classification scheme is demonstrated with examples using well-known and less popular DLTS techniques. One potential benefit of this classification is the identification of techniques and conditions that allows the reader to quickly tailor the setup to fit the specific properties of the device or material under investigation.

In Sect. III, two complementary digital signal processing techniques that are well suited for DLTS applications are presented. These techniques are new in the processing of digitized DLTS transients and can be used in virtually any DLTS experiment. Furthermore, their application can be easily extended to the data processing in other physical experiments that produce a transient signal. A mathematical model is presented and error analysis is made of pseudo- logarithmic averaging, a technique which is less well-known in digital signal processing. Because of the substantial increase of signal-to-noise ratio (SNR) and efficient data reduction, these techniques may increase interest in the isothermal DLTS techniques (Akita et al., 1993; Kim et al., 1993; Kiyota et al., 1992; Okushi and Tokumaru, 1980, 1981; Yoshida et al., 1993).

The latest developments in the feedback circuit used for CC-DLTS are presented in Sect. IV. This improved feedback circuit is essential for the successful implementation of the new CR-DLTS technique. The speed of the feedback is demonstrated by comparison of recorded traces from standard capacitance-transient DLTS, CC- and CR-DLTS. Guidelines are given for design of the feedback circuit and its setup during CC- and CR-DLTS measurements. The sensitivity achieved using the feedback circuit is demonstrated with measurements of interface-trap density of submicron MOSFETs.

A description of the new CR-DLTS technique is given in Sect. V. This new technique is similar to the conductance DLTS, but it is more sensitive and does not require simultaneous measurement of the transconductance...

Produkt-Details

• Erscheint lt. Verlag: 28.9.1999
• Mitarbeit: Herausgeber (Serie): Benjamin Kazan, Tom Mulvey; Chef-Herausgeber: Peter W. Hawkes
• Sprache: englisch
• Themenwelt: Sachbuch/Ratgeber
• Themenwelt: Mathematik / Informatik ► Informatik
• Themenwelt: Naturwissenschaften
• Themenwelt: Technik ► Elektrotechnik / Energietechnik
• ISBN-10: 0-08-057775-X / 008057775X
• ISBN-13: 978-0-08-057775-3 / 9780080577753