Issue Browser
Volume 27, Issue 2, Feb 2006
Electronic Structure of Semiconductor Nanocrystals
Li Jingbo, Wang Linwang, Wei Suhuai
Chin. J. Semicond.  2006, 27(2): 191-196
Abstract PDF

This paper reviews our recent development of the use of the large-scale pseudopotential method to calculate the electronic structure of semiconductor nanocrystals,such as quantum dots and wires,which often contain tens of thousands of atoms.The calculated size-dependent exciton energies and absorption spectra of quantum dots and wires are in good agreement with experiments.We show that the electronic structure of a nanocrystal can be tuned not only by its size,but also by its shape.Finally,we show that defect properties in quantum dots can be significantly different from those in bulk semiconductors.

This paper reviews our recent development of the use of the large-scale pseudopotential method to calculate the electronic structure of semiconductor nanocrystals,such as quantum dots and wires,which often contain tens of thousands of atoms.The calculated size-dependent exciton energies and absorption spectra of quantum dots and wires are in good agreement with experiments.We show that the electronic structure of a nanocrystal can be tuned not only by its size,but also by its shape.Finally,we show that defect properties in quantum dots can be significantly different from those in bulk semiconductors.
InAs Wires on InP (001)
Wu Ju, Wang Zhanguo
Chin. J. Semicond.  2006, 27(2): 197-203
Abstract PDF

The heterostructure of InAs/In0.52Al0.48As/InP is unique in that InAs wires instead of dots self-assemble in molecular beam epitaxy.These InAs wires have some distinctive features in their growth and structure.This paper summarizes the investigations of the growth and structural properties of InAs wires that have been performed in our laboratory recently.

The heterostructure of InAs/In0.52Al0.48As/InP is unique in that InAs wires instead of dots self-assemble in molecular beam epitaxy.These InAs wires have some distinctive features in their growth and structure.This paper summarizes the investigations of the growth and structural properties of InAs wires that have been performed in our laboratory recently.
PAPERS
Quantum and Transport Mobilities of a Two-Dimensional Electron Gas in the Presence of the Rashba Spin-Orbit Interaction
Xu Wen
Chin. J. Semicond.  2006, 27(2): 204-217
Abstract PDF

A systematic theoretical approach is developed to study the electronic and transport properties of a two-dimensional electron gas (2DEG) in the presence of spin-orbit interactions induced by the Rashba effect.The standard random-phase approximation is employed to calculate the screening length caused by electron-electron interaction in different transition channels.The quantum and transport mobilities in different spin branches are evaluated using the momentum-balance equation derived from the Boltzmann equation,in which the electron interactions with both the remote and background impurities are taken into account in an InAlAs/InGaAs heterojunction at low-temperatures

A systematic theoretical approach is developed to study the electronic and transport properties of a two-dimensional electron gas (2DEG) in the presence of spin-orbit interactions induced by the Rashba effect.The standard random-phase approximation is employed to calculate the screening length caused by electron-electron interaction in different transition channels.The quantum and transport mobilities in different spin branches are evaluated using the momentum-balance equation derived from the Boltzmann equation,in which the electron interactions with both the remote and background impurities are taken into account in an InAlAs/InGaAs heterojunction at low-temperatures
Quantum Measurement of Single Electron State by a Mesoscopic Detector
Hu Xuening, Luo Junyan, Li Xinqi
Chin. J. Semicond.  2006, 27(2): 218-222
Abstract PDF

A realistic measurement setup for a system such as a charged two-state (qubit) or multi-state quantum system measured by a mesoscopic detector,is theoretically studied.To properly describe the measurement-induced back-action,a detailed-balance preserved quantum master equation treatment is developed.The established framework is applicable for arbitrary voltages and temperatures.

A realistic measurement setup for a system such as a charged two-state (qubit) or multi-state quantum system measured by a mesoscopic detector,is theoretically studied.To properly describe the measurement-induced back-action,a detailed-balance preserved quantum master equation treatment is developed.The established framework is applicable for arbitrary voltages and temperatures.
Study of NiSi/Si Interface by Cross-Section Transmission Electron Microscopy
Jiang Yulong, Ru Guoping, Qu Xinping, Li Bingzong
Chin. J. Semicond.  2006, 27(2): 223-228
Abstract PDF

Different silicidation processes are employed to form NiSi,and the NiSi/Si interface corresponding to each process is studied by cross-section transmission electron microscopy (XTEM).With the sputter deposition of a nickel thin film,nickel silicidation is realized on undoped and doped (As and B) Si(001) substrates by rapid thermal processing (RTP).The formation of NiSi is demonstrated by X-ray diffraction and Raman scattering spectroscopy.The influence of the substrate doping and annealing process (one-step RTP and two-step RTP) on the NiSi/Si interface is investigated.The results show that for one-step RTP the silicidation on As-doped and undoped Si substrates causes a rougher NiSi/Si interface,while the two-step RTP results in a much smoother NiSi/Si interface.High resolution XTEM study shows that axiotaxy along the Si〈111〉 direction forms in all samples,in which specific NiSi planes align with Si(111) planes in the substrate.Axiotaxy with spacing mismatch is also discussed.

Different silicidation processes are employed to form NiSi,and the NiSi/Si interface corresponding to each process is studied by cross-section transmission electron microscopy (XTEM).With the sputter deposition of a nickel thin film,nickel silicidation is realized on undoped and doped (As and B) Si(001) substrates by rapid thermal processing (RTP).The formation of NiSi is demonstrated by X-ray diffraction and Raman scattering spectroscopy.The influence of the substrate doping and annealing process (one-step RTP and two-step RTP) on the NiSi/Si interface is investigated.The results show that for one-step RTP the silicidation on As-doped and undoped Si substrates causes a rougher NiSi/Si interface,while the two-step RTP results in a much smoother NiSi/Si interface.High resolution XTEM study shows that axiotaxy along the Si〈111〉 direction forms in all samples,in which specific NiSi planes align with Si(111) planes in the substrate.Axiotaxy with spacing mismatch is also discussed.
Analysis of PTCDA/ITO Surface and Interface Using X-ray Photoelectron Spectroscopy and Atomic Force Microscopy
Ou Guping, Song Zhen, Gui Wenming, Zhang Fujia
Chin. J. Semicond.  2006, 27(2): 229-234
Abstract PDF

The electronic states of the surface and interface of 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA)/indium-tin-oxide (ITO) thin film are investigated using X-ray photoelectron spectroscopy (XPS).Atomic force microscopy (AFM) is also applied to investigate the pattern of PTCDA/ITO film.XPS results show that there are two main peaks,which are associated with C atoms in the perylene rings and acid anhydride groups,located at 284.6 and 288.7eV,respectively,in the C1s spectrum of the original surface.It can be deduced from the emergence of a small peak at 290.4eV in the C1s spectrum that some C atoms are oxidized by O atoms from ITO.The binding energies of O atoms in CO bonds and COC bonds are 531.5 and 533.4eV respectively.At the interface,the peak at the high binding energy in the C1s spectrum disappears,and the peak value shifts about 0.2eV to lower binding energy.There is a significant 1.5eV chemical shift to lower binding energy in the O1s spectrum.These observations indicate that perylene rings inside PTCDA molecules are combined with In vacancies in the ITO at the interface.The AFM results show that PTCDA molecules formed an island-like structure a height of about 14nm.The sizes of the crystal grains are about 100~300nm.The island-like pattern comes from the delocalized π bonds of adjacent molecules in PTCDA and the combination of vacancies in ITO with perylene rings at the PTCDA/ITO interface.

The electronic states of the surface and interface of 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA)/indium-tin-oxide (ITO) thin film are investigated using X-ray photoelectron spectroscopy (XPS).Atomic force microscopy (AFM) is also applied to investigate the pattern of PTCDA/ITO film.XPS results show that there are two main peaks,which are associated with C atoms in the perylene rings and acid anhydride groups,located at 284.6 and 288.7eV,respectively,in the C1s spectrum of the original surface.It can be deduced from the emergence of a small peak at 290.4eV in the C1s spectrum that some C atoms are oxidized by O atoms from ITO.The binding energies of O atoms in CO bonds and COC bonds are 531.5 and 533.4eV respectively.At the interface,the peak at the high binding energy in the C1s spectrum disappears,and the peak value shifts about 0.2eV to lower binding energy.There is a significant 1.5eV chemical shift to lower binding energy in the O1s spectrum.These observations indicate that perylene rings inside PTCDA molecules are combined with In vacancies in the ITO at the interface.The AFM results show that PTCDA molecules formed an island-like structure a height of about 14nm.The sizes of the crystal grains are about 100~300nm.The island-like pattern comes from the delocalized π bonds of adjacent molecules in PTCDA and the combination of vacancies in ITO with perylene rings at the PTCDA/ITO interface.
Analysis of PTCDA/ITO Surface and Interface Using X-ray Photoelectron Spectroscopy and Atomic Force Microscopy
Tang Ning, Shen Bo, Wang Maojun, Yang Zhijian, Xu Ke, Zhang Guoyi, Gui Yongsheng, Zhu Bo, Guo Shaoling, Chu Junhao
Chin. J. Semicond.  2006, 27(2): 235-238
Abstract PDF

Magnetotransport properties of two-dimensional electron gases (2DEG) in AlxGa1-xN/GaN heterostructures with different Al compositions are investigated by magnetotransport measurements at low temperatures and in high magnetic fields.It is found that heterostructures with a lower Al composition in the barrier have lower 2DEG concentration and higher 2DEG mobility.

Magnetotransport properties of two-dimensional electron gases (2DEG) in AlxGa1-xN/GaN heterostructures with different Al compositions are investigated by magnetotransport measurements at low temperatures and in high magnetic fields.It is found that heterostructures with a lower Al composition in the barrier have lower 2DEG concentration and higher 2DEG mobility.
Effect of Lattice Mismatch on Luminescence of ZnO/Si Hetero-Structure
Fu Zhuxi, Sun Xiankai, Zhu Junjie, Lin Bixia
Chin. J. Semicond.  2006, 27(2): 239-244
Abstract PDF

The photoluminescence (PL) and Raman spectra of undoped ZnO films deposited directly on Si substrate (sample A),on Si substrate through a SiC buffer layer (sample B),and on a ZnO crystal wafer (sample C) are investigated.There are emission peaks centered at 3.18eV (ultraviolet,UV) and 2.38eV (green) in these samples.Comparing the Raman spectra and the variation of the PL peak intensities with annealing atmosphere,we conclude that the luminescence of the samples is related to the tensile strain in the ZnO film due to the lattice mismatch between the film and the substrate.In particular,the tensile strain reduces the formation energy of OZn antisite oxygen defects,which generate the green emission center.After annealing in oxygen-rich atmosphere,many OZn defects are generated.Thus,the intensity of green emission in ZnO/Si hetero-structure materials increases due to tensile strain in ZnO films.

The photoluminescence (PL) and Raman spectra of undoped ZnO films deposited directly on Si substrate (sample A),on Si substrate through a SiC buffer layer (sample B),and on a ZnO crystal wafer (sample C) are investigated.There are emission peaks centered at 3.18eV (ultraviolet,UV) and 2.38eV (green) in these samples.Comparing the Raman spectra and the variation of the PL peak intensities with annealing atmosphere,we conclude that the luminescence of the samples is related to the tensile strain in the ZnO film due to the lattice mismatch between the film and the substrate.In particular,the tensile strain reduces the formation energy of OZn antisite oxygen defects,which generate the green emission center.After annealing in oxygen-rich atmosphere,many OZn defects are generated.Thus,the intensity of green emission in ZnO/Si hetero-structure materials increases due to tensile strain in ZnO films.
Decarbonization and Decolorization of Large Sapphire Crystals Grown by the Temperature Gradient Technique
Xu Jun, Zhou Guoqing, Deng Peizhen, Si Jiliang, Qian Xiaobo, Wang Yinzhen, Zhou Shengming, Zhou Yongzong, Zhu Renyuan
Chin. J. Semicond.  2006, 27(2): 245-248
Abstract PDF

A crystalline sapphire (Al2O3) boule (110×80mm3) grown by the temperature gradient technique (TGT) is a bit colored due to carbon volatilization from the graphite heater at high temperatures and the absorption of transitional metal inclusions in the raw material.The sapphire becomes colorless and transparent after decolorization and decarbonization in successive annealings in air and hydrogen at high temperatures.The quality,optical transmissivity,and homogeneity of the sapphire are remarkably improved.

A crystalline sapphire (Al2O3) boule (110×80mm3) grown by the temperature gradient technique (TGT) is a bit colored due to carbon volatilization from the graphite heater at high temperatures and the absorption of transitional metal inclusions in the raw material.The sapphire becomes colorless and transparent after decolorization and decarbonization in successive annealings in air and hydrogen at high temperatures.The quality,optical transmissivity,and homogeneity of the sapphire are remarkably improved.
Fabrication and Emission Properties of a n-ZnO/p-GaN Heterojunction Light-Emitting Diode
Zhou Xin, Gu Shulin, Zhu Shunming, Ye Jiandong, Liu Wei, Liu Songmin, Hu Liqun, Zheng Youdou, Zhang Rong, Shi Yi
Chin. J. Semicond.  2006, 27(2): 249-253
Abstract PDF

We report the fabrication and characterization of light-emitting diodes based on n-ZnO/p-GaN heterojunctions.The n-type ZnO epilayer is deposited by metalorganic chemical vapor deposition (MOCVD) on a MOCVD grown Mg-doped p-GaN layer to form a p-n heterojunction.During the etching process, the relation between the etching depth and the etching time is linear in a HF and NH4Cl solution of a certain ratio.The etching rates of the SiO2 and ZnO are well controlled,which are essential for device fabrication.The current-voltage relationship of this heterojunction shows a diode-like rectifying behavior.In contrast to previous reports,electroluminescence (EL) emissions are observed by the naked eye at room temperature from the n-ZnO/p-GaN heterojunction under forward- and reverse-bias.The origins of these EL emissions are discussed in comparison with the photoluminescence spectra.

We report the fabrication and characterization of light-emitting diodes based on n-ZnO/p-GaN heterojunctions.The n-type ZnO epilayer is deposited by metalorganic chemical vapor deposition (MOCVD) on a MOCVD grown Mg-doped p-GaN layer to form a p-n heterojunction.During the etching process, the relation between the etching depth and the etching time is linear in a HF and NH4Cl solution of a certain ratio.The etching rates of the SiO2 and ZnO are well controlled,which are essential for device fabrication.The current-voltage relationship of this heterojunction shows a diode-like rectifying behavior.In contrast to previous reports,electroluminescence (EL) emissions are observed by the naked eye at room temperature from the n-ZnO/p-GaN heterojunction under forward- and reverse-bias.The origins of these EL emissions are discussed in comparison with the photoluminescence spectra.
Design and Simulation of a Light-Activated Darlington Transistor Based on a SiCGe/3C-SiC Hetero-Structure
Chen Zhiming, Ren Ping, Pu Hongbin
Chin. J. Semicond.  2006, 27(2): 254-257
Abstract PDF

A light-activated Darlington heterojunction transistor based on a SiCGe/3C-SiC hetero-structure is proposed for anti-EMI(electromagnetic interference) applications.The performance of the novel power switch is simulated using ISE.In comparison with the switches based on other polytypes of SiC,the design benefits from having fewer lattice mismatches between the SiCGe and 3C-SiC.A maximum common emitter current gain of about 890 and superb light-activation characteristics may be achievable.The performance simulation demonstrates that the device has a good I-V characteristic with a turn-on voltage knee of about 4V.

A light-activated Darlington heterojunction transistor based on a SiCGe/3C-SiC hetero-structure is proposed for anti-EMI(electromagnetic interference) applications.The performance of the novel power switch is simulated using ISE.In comparison with the switches based on other polytypes of SiC,the design benefits from having fewer lattice mismatches between the SiCGe and 3C-SiC.A maximum common emitter current gain of about 890 and superb light-activation characteristics may be achievable.The performance simulation demonstrates that the device has a good I-V characteristic with a turn-on voltage knee of about 4V.
Fine-Grain Sleep Transistor Insertion for Leakage Reduction
Yang Huazhong, Wang Yu, Lin Hai, Luo Rong, Wang Hui
Chin. J. Semicond.  2006, 27(2): 258-265
Abstract PDF

A fine-grain sleep transistor insertion technique based on our simplified leakage current and delay models is proposed to reduce leakage current. The key idea is to model the leakage current reduction problem as a mixed-integer linear programming (MLP) problem in order to simultaneously place and size the sleep transistors optimally. Because of better circuit slack utilization, our experimental results show that the MLP model can save leakage by 79.75%, 93.56%, and 94.99% when the circuit slowdown is 0%, 3%, and 5%, respectively. The MLP model also achieves on average 74.79% less area penalty compared to the conventional fixed slowdown method when the circuit slowdown is 7%.

A fine-grain sleep transistor insertion technique based on our simplified leakage current and delay models is proposed to reduce leakage current. The key idea is to model the leakage current reduction problem as a mixed-integer linear programming (MLP) problem in order to simultaneously place and size the sleep transistors optimally. Because of better circuit slack utilization, our experimental results show that the MLP model can save leakage by 79.75%, 93.56%, and 94.99% when the circuit slowdown is 0%, 3%, and 5%, respectively. The MLP model also achieves on average 74.79% less area penalty compared to the conventional fixed slowdown method when the circuit slowdown is 7%.
S-Band 1mm SiC MESFET with 2W Output on Semi-Insulated SiC Substrate
Cai Shujun, Pan Hongshu, Chen Hao, Li Liang, Zhao Zhenping
Chin. J. Semicond.  2006, 27(2): 266-269
Abstract PDF

A SiC MESFET structure is successfully prepared on a semi-insulated 50mm SiC substrate using a hot-wall SiC reactor.The doping concentration for the channel layer is about 1.7×1017cm-3,and the thickness is about 0.35μm.An unintentionally n-doped buffer layer is employed between the substrate and the channel layer.A cap layer for Ohmic contact is doped to 1019cm-3.MESFET devices are fabricated using inductively coupled plasma etching and other conventional tools.Power devices with a 1mm gate width are measured and a 2W output at 2GHz is obtained.

A SiC MESFET structure is successfully prepared on a semi-insulated 50mm SiC substrate using a hot-wall SiC reactor.The doping concentration for the channel layer is about 1.7×1017cm-3,and the thickness is about 0.35μm.An unintentionally n-doped buffer layer is employed between the substrate and the channel layer.A cap layer for Ohmic contact is doped to 1019cm-3.MESFET devices are fabricated using inductively coupled plasma etching and other conventional tools.Power devices with a 1mm gate width are measured and a 2W output at 2GHz is obtained.
High Power SiGe X-Band (8~10GHz) Heterojunction Bipolar Transistors and Amplifiers
Ma Zhenqiang, Wang Guogong, Jiang Ningyue, Ponchak G E, Alterovitz S A
Chin. J. Semicond.  2006, 27(2): 270-275
Abstract PDF

Limited by increased parasitics and thermal effects as device size increases,current commercial SiGe power HBTs are difficult to operate at X-band (8~12GHz) frequencies with adequate power added efficiencies at high power levels.We find that,by changing the heterostructure and doping profile of SiGe HBTs,their power gain can be significantly improved without resorting to substantial lateral scaling.Furthermore,employing a common-base configuration with a proper doping profile instead of a common-emitter configuration improves the power gain characteristics of SiGe HBTs,thus permitting these devices to be efficiently operated at X-band frequencies.In this paper,we report the results of SiGe power HBTs and MMIC power amplifiers operating at 8~10GHz.At 10GHz,a 22.5dBm (178mW) RF output power with a concurrent gain of 7.32dB is measured at the peak power-added efficiency of 20.0%,and a maximum RF output power of 24.0dBm (250mW) is achieved from a 20 emitter finger SiGe power HBT.The demonstration of a single-stage X-band medium-power linear MMIC power amplifier is also realized at 8GHz.Employing a 10-emitter finger SiGe HBT and on-chip input and output matching passive components,a linear gain of 9.7dB,a maximum output power of 23.4dBm,and peak power added efficiency of 16% are achieved from the power amplifier.The MMIC exhibits very low distortion with 3rd order intermodulation (IM) suppression C/I of -13dBc at an output power of 21.2dBm and over 20dBm 3rd order output intercept point (OIP3).

Limited by increased parasitics and thermal effects as device size increases,current commercial SiGe power HBTs are difficult to operate at X-band (8~12GHz) frequencies with adequate power added efficiencies at high power levels.We find that,by changing the heterostructure and doping profile of SiGe HBTs,their power gain can be significantly improved without resorting to substantial lateral scaling.Furthermore,employing a common-base configuration with a proper doping profile instead of a common-emitter configuration improves the power gain characteristics of SiGe HBTs,thus permitting these devices to be efficiently operated at X-band frequencies.In this paper,we report the results of SiGe power HBTs and MMIC power amplifiers operating at 8~10GHz.At 10GHz,a 22.5dBm (178mW) RF output power with a concurrent gain of 7.32dB is measured at the peak power-added efficiency of 20.0%,and a maximum RF output power of 24.0dBm (250mW) is achieved from a 20 emitter finger SiGe power HBT.The demonstration of a single-stage X-band medium-power linear MMIC power amplifier is also realized at 8GHz.Employing a 10-emitter finger SiGe HBT and on-chip input and output matching passive components,a linear gain of 9.7dB,a maximum output power of 23.4dBm,and peak power added efficiency of 16% are achieved from the power amplifier.The MMIC exhibits very low distortion with 3rd order intermodulation (IM) suppression C/I of -13dBc at an output power of 21.2dBm and over 20dBm 3rd order output intercept point (OIP3).
Numerical Explanation of Slow Transients in an AlGaN/GaN HEMT
Zhang Jinfeng, Hao Yue
Chin. J. Semicond.  2006, 27(2): 276-282
Abstract PDF

A series of slow drain current recovery transients at different gate biases after a short-term stress are observed in an AlGaN/GaN HEMT.As the variation of the time constants of the transients is small,the working trap is determined to be electronic.A numerical simulation verifies this conclusion and reproduces the measured transients.The electron traps at different spatial positions in the device-on the ungated surface of the AlGaN layer,in the AlGaN barrier,and in the GaN layer are considered;corresponding behaviors in the stress and the transients are discussed;and for the simulated transients,agreement with and deviation from the measured transients are explained.Based on this discussion,we suggest that the measured transients are caused by the combined effects of a deep surface trap and a bulk trap in the GaN layer.

A series of slow drain current recovery transients at different gate biases after a short-term stress are observed in an AlGaN/GaN HEMT.As the variation of the time constants of the transients is small,the working trap is determined to be electronic.A numerical simulation verifies this conclusion and reproduces the measured transients.The electron traps at different spatial positions in the device-on the ungated surface of the AlGaN layer,in the AlGaN barrier,and in the GaN layer are considered;corresponding behaviors in the stress and the transients are discussed;and for the simulated transients,agreement with and deviation from the measured transients are explained.Based on this discussion,we suggest that the measured transients are caused by the combined effects of a deep surface trap and a bulk trap in the GaN layer.
Study of Electron Mobility in 4H-SiC Buried-Channel MOSFETs
Gao Jinxia, Zhang Yimen, Zhang Yuming
Chin. J. Semicond.  2006, 27(2): 283-289
Abstract PDF

The effects of several factors on mobility in 4H-SiC buried-channel (BC) MOSFETs are studied.A simple model that gives a quantitative analysis of series resistance effects on the effective mobility and field-effect mobility is proposed.A series resistance not only decreases field-effect mobility but also reduces the gate voltage corresponding to the peak field-effect mobility.The dependence of the peak field-effect mobility on series resistance follows a simple quadratic polynomial.The effects of uniform and exponential interface state distributions in the forbidden band on field-effect mobility are analyzed with an analytical model.The effects of non-uniform interface states can be ignored at lower gate voltages but become more obvious as the gate bias increases.

The effects of several factors on mobility in 4H-SiC buried-channel (BC) MOSFETs are studied.A simple model that gives a quantitative analysis of series resistance effects on the effective mobility and field-effect mobility is proposed.A series resistance not only decreases field-effect mobility but also reduces the gate voltage corresponding to the peak field-effect mobility.The dependence of the peak field-effect mobility on series resistance follows a simple quadratic polynomial.The effects of uniform and exponential interface state distributions in the forbidden band on field-effect mobility are analyzed with an analytical model.The effects of non-uniform interface states can be ignored at lower gate voltages but become more obvious as the gate bias increases.
Breakdown Voltage and Charge to Breakdown Investigation of Gate Oxide of 0.18μm Dual Gate CMOS Process with Different Measurement Methods
Zhao Yi, Wan Xinggong, Xu Xiangming, Cao Gang, Bu Jiao
Chin. J. Semicond.  2006, 27(2): 290-293
Abstract PDF

Breakdown voltage (Vbd) and charge to breakdown (Qbd) are two parameters often used to evaluate gate oxide reliability.In this paper,we investigate the effects of measurement methods on Vbd and Qbd of the gate oxide of a 0.18μm dual gate CMOS process.Voltage ramps (V-ramp) and current ramps (J-ramp) are used to evaluate gate oxide reliability.The thin and thick gate oxides are all evaluated in the accumulation condition.Our experimental results show that the measurement methods affect Vbd only slightly but affect Qbd seriously,as do the measurement conditions.This affects the I-t curves obtained with the J-ramp and V-ramp methods.From the I-t curve,it can be seen that Qbd obtained using a J-ramp is much bigger than that with a V-ramp.At the same time,the Weibull slopes of Qbd are definitely smaller than those of Vbd.This means that Vbd is more reliable than Qbd.Thus we should be careful to use Qbd to evaluate the reliability of 0.18μm or beyond CMOS process gate oxide.

Breakdown voltage (Vbd) and charge to breakdown (Qbd) are two parameters often used to evaluate gate oxide reliability.In this paper,we investigate the effects of measurement methods on Vbd and Qbd of the gate oxide of a 0.18μm dual gate CMOS process.Voltage ramps (V-ramp) and current ramps (J-ramp) are used to evaluate gate oxide reliability.The thin and thick gate oxides are all evaluated in the accumulation condition.Our experimental results show that the measurement methods affect Vbd only slightly but affect Qbd seriously,as do the measurement conditions.This affects the I-t curves obtained with the J-ramp and V-ramp methods.From the I-t curve,it can be seen that Qbd obtained using a J-ramp is much bigger than that with a V-ramp.At the same time,the Weibull slopes of Qbd are definitely smaller than those of Vbd.This means that Vbd is more reliable than Qbd.Thus we should be careful to use Qbd to evaluate the reliability of 0.18μm or beyond CMOS process gate oxide.
Axial Local Lifetime Control in High-Voltage Diodes Based on Proximity Gettering of Platinum by Proton-Implantation Damages
Jia Yunpeng, Zhang Bin, Sun Yuechen, Kang Baowei
Chin. J. Semicond.  2006, 27(2): 294-297
Abstract PDF

A new lifetime control technique-localized platinum lifetime control (LPLC) is introduced.Silicon samples are implanted with 550keV protons at dosages from 1×1013 to 5×1014cm-2.Subsequently,platinum diffusion in silicon is performed at 700 or 750℃ for 15 or 30min,respectively.Then the in-diffused platinum into damaged regions of the proton-implanted silicon is investigated by use of deep-level transient spectroscopy (DLTS).Finally,for all of the LPLC samples,the distribution of the in-diffused substitutional platinum agrees well with the damage distribution resulting from the low-dosage proton implantation.Also,the diodes show a very low leakage current even at elevated temperatures while keeping the major advantages of ion irradiation devices,including low turn-off loss and soft recovery.

A new lifetime control technique-localized platinum lifetime control (LPLC) is introduced.Silicon samples are implanted with 550keV protons at dosages from 1×1013 to 5×1014cm-2.Subsequently,platinum diffusion in silicon is performed at 700 or 750℃ for 15 or 30min,respectively.Then the in-diffused platinum into damaged regions of the proton-implanted silicon is investigated by use of deep-level transient spectroscopy (DLTS).Finally,for all of the LPLC samples,the distribution of the in-diffused substitutional platinum agrees well with the damage distribution resulting from the low-dosage proton implantation.Also,the diodes show a very low leakage current even at elevated temperatures while keeping the major advantages of ion irradiation devices,including low turn-off loss and soft recovery.
Two-Dimensional Static Numerical Modeling and Simulation of AlGaN/GaN HEMT
Xue Lijun, Xia Yang, Liu Ming, Wang Yan, Shao Xue, Lu Jing, Ma Jie, Xie Changqing, Yu Zhiping
Chin. J. Semicond.  2006, 27(2): 298-303
Abstract PDF

AlGaN/GaN HEMTs are investigated by numerical simulation from the self-consistent solution of Schrdinger-Poisson-hydrodynamic (HD) systems.The influences of polarization charge and quantum effects are considered in this model.Then the two-dimensional conduction band and electron distribution,electron temperature characteristics,Id versus Vd and Id versus Vg,transfer characteristics and transconductance curves are obtained.Corresponding analysis and discussion based on the simulation results are subsequently given.

AlGaN/GaN HEMTs are investigated by numerical simulation from the self-consistent solution of Schrdinger-Poisson-hydrodynamic (HD) systems.The influences of polarization charge and quantum effects are considered in this model.Then the two-dimensional conduction band and electron distribution,electron temperature characteristics,Id versus Vd and Id versus Vg,transfer characteristics and transconductance curves are obtained.Corresponding analysis and discussion based on the simulation results are subsequently given.
Monte Carlo Simulation of Impurity Scattering Effect in Resonant-Phonon-Assisted Terahertz Quantum-Cascade Lasers
Cao Juncheng, Lü Jingtao
Chin. J. Semicond.  2006, 27(2): 304-308
Abstract PDF

We study the influence of ionized impurity scattering on the electron transport in resonant-phonon-assisted terahertz (THz) quantum-cascade lasers (QCLs).We treat the ionized impurity scattering rates within the single subband static screening approximation.We find that the ionized impurity scattering supplies an additional current channel across the device,and affects the electrondistribution in different subbands.We conclude that the ionized impurity scattering should be taken into account in the study of the transport properties of resonant-phonon-assisted THz QCLs.

We study the influence of ionized impurity scattering on the electron transport in resonant-phonon-assisted terahertz (THz) quantum-cascade lasers (QCLs).We treat the ionized impurity scattering rates within the single subband static screening approximation.We find that the ionized impurity scattering supplies an additional current channel across the device,and affects the electrondistribution in different subbands.We conclude that the ionized impurity scattering should be taken into account in the study of the transport properties of resonant-phonon-assisted THz QCLs.
Fabrication and Numerical Simulation of a Micromachined Contact Cantilever RF-MEMS Switch
Sun Jianhai, Cui Dafu, Xiao Jiang
Chin. J. Semicond.  2006, 27(2): 309-312
Abstract PDF

This paper reports a contact cantilever MEMS switch.The designed switch has a metal cantilever that acts as an electrostatically activated switch with processing options to achieve dielectric isolation of the control-voltage path from the signal path.To obtain good material properties,an ANSYS FEA tool is used to optimize the structure.The RF MEMS switch is fabricated via a surface micromachining process.The switch has an actuation voltage of 12V,which is close to the simulated value of 11V.The measured and the HFSS simulated isolations are both higher than -20dB from 0.05 to 10GHz.The measured insertion loss is less than -0.9dB,relatively larger than the simulated loss of less than -0.2dB from 0.05 to 10GHz.This is because a contact resistance introduced due to poor physical contact between the bottom lines and the metal cantilever.

This paper reports a contact cantilever MEMS switch.The designed switch has a metal cantilever that acts as an electrostatically activated switch with processing options to achieve dielectric isolation of the control-voltage path from the signal path.To obtain good material properties,an ANSYS FEA tool is used to optimize the structure.The RF MEMS switch is fabricated via a surface micromachining process.The switch has an actuation voltage of 12V,which is close to the simulated value of 11V.The measured and the HFSS simulated isolations are both higher than -20dB from 0.05 to 10GHz.The measured insertion loss is less than -0.9dB,relatively larger than the simulated loss of less than -0.2dB from 0.05 to 10GHz.This is because a contact resistance introduced due to poor physical contact between the bottom lines and the metal cantilever.
A New CMOS Image Sensor with a High Fill Factor and the Dynamic Digital Double Sampling Technique
Liu Yu, Wang Guoyu
Chin. J. Semicond.  2006, 27(2): 313-317
Abstract PDF

A single CMOS image sensor based on a 0.35μm process along with its design and implementation is introduced.The architecture of an active pixel sensor is used in the chip.The fill factor of a pixel cell can reach 43%,higher than the traditional factor of 30%.Moreover,compared with the conventional method whose fixed pattern noise (FPN) is around 0.5%, a dynamic digital double sampling technique is developed,which possesses simpler circuit architecture and a better FPN suppression outcome.The CMOS image sensor chip is implemented in the 0.35μm mixed signal process of a Chartered by MPW.The experimental results show that the chip operates well,with an FPN of about 0.17%.

A single CMOS image sensor based on a 0.35μm process along with its design and implementation is introduced.The architecture of an active pixel sensor is used in the chip.The fill factor of a pixel cell can reach 43%,higher than the traditional factor of 30%.Moreover,compared with the conventional method whose fixed pattern noise (FPN) is around 0.5%, a dynamic digital double sampling technique is developed,which possesses simpler circuit architecture and a better FPN suppression outcome.The CMOS image sensor chip is implemented in the 0.35μm mixed signal process of a Chartered by MPW.The experimental results show that the chip operates well,with an FPN of about 0.17%.
A Low Power SRAM/SOI Memory Cell Design
Yu Yang, Zhao Qian, Shao Zhibiao
Chin. J. Semicond.  2006, 27(2): 318-322
Abstract PDF

A modified four transistor (4T) self-body-bias structured SRAM/SOI memory cell is proposed.The structure is designed and its parameters are obtained by performance simulation and analysis with TSUPREM4 and MEDICI.The structure saves area and its process is simplified by using the body resistor with buried p+ channel beneath the nMOS gate instead of the pMOS of 6T CMOS SRAM.Furthermore,this structure can operate safely with a 0.5V supply voltage,which may be prevalent in the near future.Finally,compared to conventional 6T CMOS SRAM,this structure’s transient responses are normal and its power dissipation is 10 times smaller.

A modified four transistor (4T) self-body-bias structured SRAM/SOI memory cell is proposed.The structure is designed and its parameters are obtained by performance simulation and analysis with TSUPREM4 and MEDICI.The structure saves area and its process is simplified by using the body resistor with buried p+ channel beneath the nMOS gate instead of the pMOS of 6T CMOS SRAM.Furthermore,this structure can operate safely with a 0.5V supply voltage,which may be prevalent in the near future.Finally,compared to conventional 6T CMOS SRAM,this structure’s transient responses are normal and its power dissipation is 10 times smaller.
Monolithically Integrated Optoelectronic Receivers Implemented in 0.25μm MS/RF CMOS
Chen Hongda, Gao Peng, Mao Luhong, Huang Jiale
Chin. J. Semicond.  2006, 27(2): 323-327
Abstract PDF

A monolithically integrated optoelectronic receiver is presented.A silicon-based photo-diode and receiver circuits are integrated on identical substrates in order to eliminate the parasitics induced by hybrid packaging.Implemented in the present deep sub-micron MS/RF (mixed signal,radio frequency) CMOS,this monolithically OEIC takes advantage of several new features to improve the performance of the photo-diode and eventually the whole OEIC.

A monolithically integrated optoelectronic receiver is presented.A silicon-based photo-diode and receiver circuits are integrated on identical substrates in order to eliminate the parasitics induced by hybrid packaging.Implemented in the present deep sub-micron MS/RF (mixed signal,radio frequency) CMOS,this monolithically OEIC takes advantage of several new features to improve the performance of the photo-diode and eventually the whole OEIC.
Tapered Multimode Interference Combiners for Coherent Receivers
Wu Zhigang, Zhang Weigang, Wang Zhi, Kai Guiyun, Yuan Shuzhong, Dong Xiaoyi, Utaka Katsuyuki, Wada Yasuo
Chin. J. Semicond.  2006, 27(2): 328-335
Abstract PDF

A new tapered multimode interference (MMI)-based coherent lightwave combiner is reported.A comprehensive theoretical analysis of mode behaviors in the tapered MMI waveguide is presented,and the output characteristics of the tapered MMI combiners with various structures are demonstrated.The combiner is fabricated on a silicon-on-insulator (SOI) substrate.Due to its advantages of having no end-facet reflection,easy extension to a multi-port configuration,high tolerance for fabrication errors,and compact size,the tapered MMI is a good candidate for a coherent lightwave combiner to be used in large-scale photonic integrated circuits

A new tapered multimode interference (MMI)-based coherent lightwave combiner is reported.A comprehensive theoretical analysis of mode behaviors in the tapered MMI waveguide is presented,and the output characteristics of the tapered MMI combiners with various structures are demonstrated.The combiner is fabricated on a silicon-on-insulator (SOI) substrate.Due to its advantages of having no end-facet reflection,easy extension to a multi-port configuration,high tolerance for fabrication errors,and compact size,the tapered MMI is a good candidate for a coherent lightwave combiner to be used in large-scale photonic integrated circuits
Growth of ZnO Single Crystal by Chemical Vapor Transport Method
Zhao Youwen, Dong Zhiyuan, Wei Xuecheng, Duan Manlong, Li Jinmin
Chin. J. Semicond.  2006, 27(2): 336-339
Abstract PDF

A ZnO single crystal is grown by a chemical vapor transport method.The size of the self-nucleated grown ZnO single crystal is 5mm×8mm through the control of temperature gradient of source and growth zones and the assistance of mass transport enhancement effect of carbon in the growth process.By using a GaN buffered sapphire substrate,a ZnO single crystal of 32mm diameter,4mm thick is obtained.PL spectroscopy and X-ray diffraction technique are used to characterize the properties and the quality of the ZnO single crystal.Thermodynamic phenomena of the single crystal growth process are also discussed.

A ZnO single crystal is grown by a chemical vapor transport method.The size of the self-nucleated grown ZnO single crystal is 5mm×8mm through the control of temperature gradient of source and growth zones and the assistance of mass transport enhancement effect of carbon in the growth process.By using a GaN buffered sapphire substrate,a ZnO single crystal of 32mm diameter,4mm thick is obtained.PL spectroscopy and X-ray diffraction technique are used to characterize the properties and the quality of the ZnO single crystal.Thermodynamic phenomena of the single crystal growth process are also discussed.
Characteristics of Annealing of InN Films
Xie Zili, Zhang Rong, Xiu Xiangqian, Bi Zhaoxia, Liu Bin, Pu Lin, Chen Dunjun, Han Ping, Gu Shulin, Jiang Ruoliang, Zhu Shunming, , Zhao Hong, Shi Yi
Chin. J. Semicond.  2006, 27(2): 340-344
Abstract PDF

The characteristics of annealing of InN films with NH3 atmosphere are investigated.The XRD,SEM,and XPS are used to analyze the samples.The experiments indicate that the crystalline quality and morphology of InN do not evolve monotonously with annealing temperatures.During the annealing,In atoms resulting from the volatilization of N atoms aggregate on the surface,and thus In grains form.When the annealing temperature is higher than 425℃,the density of In grains decreases due to the desorption of In atoms.The results of XRD and SEM indicate that the sample with most dense In grains has a low crystalline quality.Thus,we consider that In grains prevent the improvement of InN quality by the thermal annealing maybe due to the separation of subjacent InN from the ammonia environment and the structural mismatch between In and InN grains.

The characteristics of annealing of InN films with NH3 atmosphere are investigated.The XRD,SEM,and XPS are used to analyze the samples.The experiments indicate that the crystalline quality and morphology of InN do not evolve monotonously with annealing temperatures.During the annealing,In atoms resulting from the volatilization of N atoms aggregate on the surface,and thus In grains form.When the annealing temperature is higher than 425℃,the density of In grains decreases due to the desorption of In atoms.The results of XRD and SEM indicate that the sample with most dense In grains has a low crystalline quality.Thus,we consider that In grains prevent the improvement of InN quality by the thermal annealing maybe due to the separation of subjacent InN from the ammonia environment and the structural mismatch between In and InN grains.
Photoluminescence and Application of Nonlinear Optical Property of nc-Si-SiO2 Films
Guo Hengqun, Lin Shangxin, Wang Qiming
Chin. J. Semicond.  2006, 27(2): 345-349
Abstract PDF

nc-Si-SiO2 films are prepared by RF magnetron sputtering technique and thermal annealing.The photoluminescence (PL) spectra at room temperature have 3 luminescent band peaks at 360,430,and 835nm,respectively.The PL mechanism is discussed in combination with absorption measurement,PL excitation,and X-ray diffraction.Passive Q-switched operation of Nd∶YAG lasers is demonstrated with a nc-Si-SiO2 film as a saturable absorber.

nc-Si-SiO2 films are prepared by RF magnetron sputtering technique and thermal annealing.The photoluminescence (PL) spectra at room temperature have 3 luminescent band peaks at 360,430,and 835nm,respectively.The PL mechanism is discussed in combination with absorption measurement,PL excitation,and X-ray diffraction.Passive Q-switched operation of Nd∶YAG lasers is demonstrated with a nc-Si-SiO2 film as a saturable absorber.
Study of Thermal Characteristics of Semiconductor Light-Emitting Devices
Zhang Yuezong, Feng Shiwei, Xie Xuesong, Li Ying, Yang Ji, Sun Jingying, Lü Changzhi
Chin. J. Semicond.  2006, 27(2): 350-353
Abstract PDF

The heating response curves of temperature rise and thermal resistance of semiconductor power emitting-lighting diodes are obtained according to electrical methods.The curve shows one or more sidesteps to reflect device’s inside thermal resistance constitution and physical structure.The temperature rise and thermal resistance are amended with a covering method.A transient heating response theory is also used to inspect the package structure of the devices.

The heating response curves of temperature rise and thermal resistance of semiconductor power emitting-lighting diodes are obtained according to electrical methods.The curve shows one or more sidesteps to reflect device’s inside thermal resistance constitution and physical structure.The temperature rise and thermal resistance are amended with a covering method.A transient heating response theory is also used to inspect the package structure of the devices.
An ADI Method for the Breakdown Voltage Analysis of Thin-Film SOI RESURF Structure with the High-Order Compact Finite Difference
Yu Zongguang, Liu Zhan, Wang Guozhang, Xu Ziming
Chin. J. Semicond.  2006, 27(2): 354-357
Abstract PDF

The application of ADI and high-order compact finite difference method to the breakdown voltage analysis of thin film SOI RESURF structure.Numerical results present that this method can decrease the number of iterative by 40% and reduce the computation time greatly.

The application of ADI and high-order compact finite difference method to the breakdown voltage analysis of thin film SOI RESURF structure.Numerical results present that this method can decrease the number of iterative by 40% and reduce the computation time greatly.
Research on CMOS Compatible Integrated Humidity Sensors
Peng Shaohua, Huang Qing’an, Qin Ming, Zhang Zhongping
Chin. J. Semicond.  2006, 27(2): 358-362
Abstract PDF

A CMOS compatible integrated humidity sensor is designed and fabricated.The humidity sensor and CMOS circuits are fabricated at the same chip.The humidity sensor employs an interdigital finger structure and a polyimide sensing material.A humidity sensing model is given.Because of minor capacitance variation during the full range,switched-capacitor techniques are employed in measurement circuits.The principle of the circuits is discussed,and simulation results are presented accordingly.The design,implemented in a 3μm polysilicon-gate CMOS,results in a good DC output at 5~35℃ and a long time stability.

A CMOS compatible integrated humidity sensor is designed and fabricated.The humidity sensor and CMOS circuits are fabricated at the same chip.The humidity sensor employs an interdigital finger structure and a polyimide sensing material.A humidity sensing model is given.Because of minor capacitance variation during the full range,switched-capacitor techniques are employed in measurement circuits.The principle of the circuits is discussed,and simulation results are presented accordingly.The design,implemented in a 3μm polysilicon-gate CMOS,results in a good DC output at 5~35℃ and a long time stability.
Room-Temperature Operated and Fully Gate-Bias Controlled Memory Devices Based on Self-Assembled InAs/GaAs Quantum Dots
Du Jun, Wang Qingpu, Balocco C, Song A M
Chin. J. Semicond.  2006, 27(2): 363-367
Abstract PDF

Memory devices fabricated in high-electron-mobility transistors with embedded InAs quantum dots (QDs) can be fully controlled by gate bias at room temperature.The memory effect is due to the deep levels induced by the QD layer,and rather than the charging and discharging of intrinsic energy levels in QDs,which is demonstrated by the hysteresis,real-time and bias-cooling C-V measurements

Memory devices fabricated in high-electron-mobility transistors with embedded InAs quantum dots (QDs) can be fully controlled by gate bias at room temperature.The memory effect is due to the deep levels induced by the QD layer,and rather than the charging and discharging of intrinsic energy levels in QDs,which is demonstrated by the hysteresis,real-time and bias-cooling C-V measurements
A Novel Triplexer Design Based on Arrayed Waveguide Grating
Lang Tingting, He Jianjun, He Sailing
Chin. J. Semicond.  2006, 27(2): 368-372
Abstract PDF

An arrayed waveguide grating (AWG) based triplexer has many advantages including small size and low cost.The commonly used wavelengths in passive optical network (PON) are 1310nm,1490nm and 1550nm,according to ITU G.983 standard.Because of the wide spectral range and very unequal channel spacings,a conventional AWG design cannot produce a satisfactory result.In this paper,a novel triplexer design based on AWG is presented.The spectral periodicity of the grating is utilized in the design so that the first wavelength works at a different diffraction order,and it is mapped to a wavelength that is approximately in the middle of the second and third wavelength channels.This results in a reduced free spectral range (FSR) requirement and almost equal distances between the input/output waveguides.Consequently,the AWG can operate at a higher diffraction order with a smaller device size,compared to the one designed by conventional methods.

An arrayed waveguide grating (AWG) based triplexer has many advantages including small size and low cost.The commonly used wavelengths in passive optical network (PON) are 1310nm,1490nm and 1550nm,according to ITU G.983 standard.Because of the wide spectral range and very unequal channel spacings,a conventional AWG design cannot produce a satisfactory result.In this paper,a novel triplexer design based on AWG is presented.The spectral periodicity of the grating is utilized in the design so that the first wavelength works at a different diffraction order,and it is mapped to a wavelength that is approximately in the middle of the second and third wavelength channels.This results in a reduced free spectral range (FSR) requirement and almost equal distances between the input/output waveguides.Consequently,the AWG can operate at a higher diffraction order with a smaller device size,compared to the one designed by conventional methods.
TECHNICAL PROGRESS
Characterization and Modeling for 0.13μm RF MOSFETs
Chi Yusong, Huang Fengyi, Wu Zhongjie, Zhang Shaoyong, Kong Xiaoming, Wang Zhigong
Chin. J. Semicond.  2006, 27(2): 373-376
Abstract PDF

RF nMOS transistors with optimized layouts are successfully fabricated in a 0.13μm CMOS RF/MS technology.DC characteristic and S-parameter measurements are performed.The measured results show that the cut-off frequency and the maximum oscillation frequency of the RF nMOS reach beyond 93 and 90GHz,respectively.Small-signal modeling is carried out for these RF MOSFETs using a small-signal equivalent circuit model.The simulation results fit well with the measurement results in frequency range from 100MHz to 30GHz.

RF nMOS transistors with optimized layouts are successfully fabricated in a 0.13μm CMOS RF/MS technology.DC characteristic and S-parameter measurements are performed.The measured results show that the cut-off frequency and the maximum oscillation frequency of the RF nMOS reach beyond 93 and 90GHz,respectively.Small-signal modeling is carried out for these RF MOSFETs using a small-signal equivalent circuit model.The simulation results fit well with the measurement results in frequency range from 100MHz to 30GHz.