碲镉汞的离子注入

在3～5和8～12μm 波段,碲镉汞(MCT)已成为红外探测和热成象的最重要的半导体材料。目前,高质量光电二极管可利用包含离子注入形成 n-p 结的工艺来制作。但是,掺杂机理还说不清楚,因为这种化合物半导体的电学性质,不仅由杂质决定,而且还由“缺陷”及与化学配比的偏离所决定。离子注入 MCT 的主要特征是:在注入的材料中,退火前观察到的是 n~+电学掺杂。“缺陷”似乎是产生这个现象的原因,因为形成这一 n~+层不需要任何进一步退火,同时也不怎么依赖注入条件及注入离子的性质。本文描述了离子注入 MCT 的一般性质。我们将给出观测缺陷所引起的载流子浓度饱和、结迁移和退火材料性质等许多问题。这些问题的回答有助于了解离子注入 MCT 的掺杂机理。看来经过注入杂质,接着进行退火的掺杂也是一个更为复杂的问题,因为它主要与缺陷的退火和化学配比的控制有关。p 型掺杂杂质和 n 型掺杂缺陷之间的竞争,提出了通过离子注入形成 p 型层的可能性问题。     

退火过程中RP缺陷模型及数值模拟

根据一维动力学方程,提出了RP缺陷的演化模型,用于描述离子注入后硼杂质分布在退火过程中出现异常变化的物理现象.通过分析发现缺陷随退火时间呈指数变化,根据变化的时间常数与RP缺陷对间隙原子束缚能的大小有关的原理,提出RP缺陷对间隙原子的束缚能为2.41eV.将该模型模拟硼杂质随退火时间的分布时,得到缺陷的分布与硼原子的分布变化趋势一致,且变化的时间常数相近,这给出了退火中硼出现异常分布的一种新的解释.     

退火过程中RP缺陷模型及数值模拟

根据一维动力学方程，提出了RP缺陷的演化模型，用于描述离子注入后硼杂质分布在退火过程中出现异常变化的物理现象．通过分析发现缺陷随退火时间呈指数变化，根据变化的时间常数与RP缺陷对间隙原子束缚能的大小有关的原理，提出RP缺陷对间隙原子的束缚能为2.41eV．将该模型模拟硼杂质随退火时间的分布时，得到缺陷的分布与硼原子的分布变化趋势一致，且变化的时间常数相近，这给出了退火中硼出现异常分布的一种新的解释．     

Te~ 离子注入溶液生长的Pb_(1-x)Sn_xTe制备P-N结

采用Te~ 离子注入使N型Pb_(1-x)Sn_xTe转变成P型层。Te~ 离子浓度大于1×10~(15)/厘米~2,并经适当的后退火处理在N型Pb_(1-x)Sn_xTe表面上产生大约1微米厚P型层。近年来,曾采用离子注入在Pb_(1-x)Sn_xTe晶体中制备P-N结,晶体的类型转变是由于用离子注入(如同化学掺杂)引进外来的原子或因注入产生的晶格缺陷造成的。在本文中,我们报导通过基质原子Te的注入控制化学计算的偏差使N型Pb_(1-x)Sn_xTe晶体类型转变。为了减少因注入而产生的晶格缺陷(贡献出N型电导)还需要适当的后退火。     

Si/NiO异质pn结的光电性能研究

利用磁控溅射方法成功制备了Si/NiO异质pn结。实验表明,退火温度升到400℃时,Si/NiO异质结呈现一定的整流特性;600℃退火的Si/NiO异质结呈现很好的整流特性,正向开启电压达到3.5V,-7V时出现漏电流。这可能是因为600℃退火后,样品结晶转好,应力释放,缺陷减少,从而改善了样品的整流特性。这一结果得到了X射线衍射(XRD)、原子显微镜(AFM)和紫外(UV)结果的充分支持。     

中波碲镉汞p-on-n高温工作技术研究

针对碲镉汞中波p-on-n技术进行研究,采用二次离子质谱仪分析注入后及退火后As离子在碲镉汞材料中的浓度分布,使用透射电镜表征激活退火后离子注入损伤修复状态,通过半导体参数测试仪评价pn结的IV特性,将探测器芯片装在变温杜瓦中测试其不同温度下的焦平面技术指标。研究结果表明,As离子注入后在碲镉汞体内形成大量缺陷,经过富汞退火后缺陷得到修复,同时As离子进一步向内扩散,制备的pn结工作稳定表明As离子得到有效激活,制备的中波p-on-n探测器芯片在120 K温度下有效像元率可以达到99%以上。     

C离子注入Si中Si-C合金的形成及其特征

利用离子注入和高温退火的方法在 Si中生长了 C含量为 0 .6 %— 1.0 %的 Si1 - x Cx 合金 ,研究了注入过程中产生的损伤缺陷、注入 C离子的剂量及退火工艺对合金形成的影响 ,探讨了合金的形成机理及合金产生的应变分布的起因 .如果注入的 C离子剂量小于引起 Si非晶化的剂量 ,退火过程中注入产生的损伤缺陷容易与 C原子结合形成缺陷团簇 ,难于形成 Si1 - x Cx 合金 ,而预先利用 Si离子注入引进损伤有利于 Si1 - x Cx 合金的形成 ;但如果注入的C离子可以引起 Si的非晶化 ,预先注入产生的损伤缺陷不利于 Si1 - x Cx 合金的形成 .与慢速退火工艺相比 ,快速     

砷注入硅快速热退火缺陷的HVEM观察

随着集成度的提高,要求芯片的线条进一步缩小,但根据按比例缩小原则,纵向尺寸也必须减小。因此VLSI要求p—n结的结深越来越浅。离子注S和快速热退火为制备这种高浓度浅p—n结提供了一种要求的手段。当然离子注入产生的辐照缺陷及其在退火过程中的行为就自然受到人们的注意。由于砷在硅中固溶度较高、扩散系数小,离子注入的平均投影射程小等一系列特点,在制备高浓度、浅N+P结方面引起了相当的重视。本文研究了砷注入硅快速热退火后缺陷的特性。即砷注入硅后经不同温度不同时间退火后的缺陷。所用的试样为6—8ΩcmP—型<100>硅片,经室温120kev砷离子注入,剂量为1×10~(16)cm~(-2),分别在900℃、1000℃、1080℃和1150℃不同时间红外退火。並利用霍尔效应和剥层技术测定载流子浓度分布,     

C离子注入Si中Si-C合金的形成及其特征

利用离子注入和高温退火的方法在Si中生长了C含量为0.6%—1.0%的Si1-xCx合金,研究了注入过程中产生的损伤缺陷、注入C离子的剂量及退火工艺对合金形成的影响,探讨了合金的形成机理及合金产生的应变分布的起因．如果注入的C离子剂量小于引起Si非晶化的剂量,退火过程中注入产生的损伤缺陷容易与C原子结合形成缺陷团簇,难于形成Si1-xCx合金,而预先利用Si离子注入引进损伤有利于Si1-xCx合金的形成;但如果注入的C离子可以引起Si的非晶化,预先注入产生的损伤缺陷不利于Si1-xCx合金的形成．与慢速退火工艺相比,快速热退火工艺有利于Si1-xCx合金的形成．离子注入的C原子在空间分布不均匀,退火过程中将形成应变不同的Si1-x-Cx合金区域．     

Si/NiO异质pn结的光电性能研究

利用磁控溅射方法成功制备了Si/NiO异质pn结。实 验表明,退火温度升到400℃时,Si/NiO异 质结呈现一定的整流特性;600℃退火的Si/NiO异质结呈现很好的整 流特性,正向开启 电压达到3.5VV时出现漏电流。这可 能是因为600℃退火后,样品结晶转好,应力 释放,缺陷减少,从而改善了样品的整流特性。这一结果得到了X射线衍射(XRD)、原子显微 镜(AFM)和紫外(UV)结果的充分支持。     

P型Hg0.76Cd0.24Te的子能带结构

制备了Ｈｇ１－ｘＣｄｘＴｅ（ｘ＝０．２４）ＭＩＳ器件，用自制的高精度差分电容谱仪测量了器件的Ｃ－Ｖ谱。根据褚君浩等提出的实验模型拟合测得的电容谱，获得了反型层电子子能带结构、，与本文的修正变分自洽方法计算结果基本一致。并获得了有关界面态与绝缘层固定电荷的结果。     

The growth and characterization of (211) and (133) Oriented (Hg,Cd)Te epilayers (211)B GaAs by organometallic vapor phase epitaxy

The growth of CdTe buffer layers on (211)B GaAs substrates by organometallic vapor phase epitaxy (OMVPE) was studied, and it was found that, depending on the growth conditions, either the (211) or (133) epitaxial orientation could be formed. In some cases, an epilayer showing a mixed (211) and (133) orientation was also observed. The influence of several growth parameters on the orientation of the CdTe layer was investigated, and it was found that the Te/Cd ratio, together with the growth temperature, have the most significant effect in determining the epilayer orientation. From these results, it was then possible to select nominally optimized growth conditions for CdTe buffer layers of both orientations. (Hg,Cd)Te layers of the same orientations could then be grown and characterized. Although double crystal x-ray diffraction measurements indicated a somewhat better crystalline perfection in the (133) (Hg,Cd)Te layers, these layers showed a poor surface morphology compared to the (211) orientation. Measurement of etch pit densities also indicated defect densities to be typically half an order of magnitude higher in the (133) orientation. Diodes were formed by ion implantation in both orientations and significantly better results were obtained on the (211) (Hg,Cd)Te layers.     

Thermomigration of Te precipitates and improvement of (Cd,Zn)Te substrate characteristics for the fabrication of LWIR (Hg,Cd)Te photodiodes

(Cd,Zn)Te wafers containing Te precipitates have been annealed under well defined thermodynamic conditions at temperatures below and above the melting of Te. Results of the examination of the wafers with infrared microscopy before and after the anneals indicate a substantial reduction of the Te precipitates in wafers annealed at temperatures in excess of the melting point of Te compared with those annealed at temperatures below the melting point of Te. These results confirm the thermomigration of liquid Te precipitates to be the principally operative mechanism during annealing in the elimination of these precipitates in (Cd,Zn)Te wafers. The occurrence of Te precipitates in (Hg,Cd)Te epitaxial layers grown on (Cd,Zn)Te substrates containing Te precipitates is also explained on the basis of thermomigration of these precipitates during LPE growth from the substrates to the epilayers. Absence of occurrence of Te precipitates in (Hg,Cd)Te epilayers grown on annealed (Cd,Zn)Te substrates with negligible Te precipitates is also confirmed. Usefulness of annealing (Cd,Zn)Te substrates—to eliminate Te precipitates—prior to epilayer growth is confirmed via demonstration of improved long wavelength infrared (Hg,Cd)Te device array performance uniformity in epitaxial layers grown on (Cd,Zn)Te substrates with negligible Te precipitates after annealing.     

The oscillations in ESR spectra of Hg0.76Cd0.24Te implanted by Ag+ at the X and Q-bands

The objects of the investigation were uniformly Ag+ doped Hg0.76Cd0.24Te mercury chalcogenide monocrystals obtained by ion implantation with subsequent thermal annealing over 20 days.After implantation and annealing the conductivity was inverted from n-type with carrier concentration of 1016 cm-3 to p-type with carrier concentration of ≈ 3.9 × 1015 cm-3.The investigations of microwave absorption derivative (Dp/Dh) showed the existence of strong oscillations in the magnetic field for Ag:Hg0.76Cd0.24Te in the temperature range 4.2-12 K.The concentration and effective mass of charge carrier were determined from oscillation period and temperature dependency of oscillation amplitude.We suppose that this phenomenon is similar to the de Haas-van Alphen effect in weakly correlated electron system with imperfect nesting vector.     

Annealing experiments in heavily arsenic-doped (Hg,Cd)Te

Arsenic doped molecular beam epitaxy (MBE) (Hg,Cd)Te films were grown on (Cd,Zn)Te substrates. The concentration of arsenic was varied from 5 x 10¹⁸ cm^-3 to 1 x 10²⁰ cm^-3. After the growth, the epitaxial layers were annealed at various partial pressures of Hg within the existence region of (Hg,Cd)Te at temperatures ranging from 400 to 500°C. Hall effect and resistivity measurements were carried out subsequent to the anneals. 77K hole concentration measurements indicate that for concentrations of arsenic <10¹⁹ cm⁻³, most of the arsenic is electrically active acting as acceptors interstitially and/or occupying Te lattice sites at the highest Hg pressures. At lower Hg pressures, particularly at annealing temperatures of 450°C and higher, compensation by arsenic centers acting as donors appears to set in and the hole concentration decreases with decrease in Hg pressure. These results indicate the amphoteric behavior of arsenic and its similarity to the behavior of phosphorus in (Hg,Cd)Te previously inferred by us. A qualitative model which requires the presence of arsenic occupying both interstitial and Te lattice sites along with formation of pairs of arsenic centers is conjectured.     

1-2 micron (Hg, Cd)Te photodetectors

The performance of mercury cadmium telluride detectors in the 1-2 micron spectral region has been predicted from basic material parameters. Photovoltaic devices should be characterized by specific responsivities of 1 A/W for a 1000 ohm load when transit time limited to less than 20 ns. Photoconductive detectors made from n-type material should have radiative lifetimes of 1 ms. The feasibility of high performance 1-2 micron (Hg, Cd)Te detectors has been demonstrated experimentally. Deep junction devices operating at room temperature without bias have been fabricated by impurity indiffusion. Detectivities at 1.75 microns approached 10¹⁰cm.Hz^1/2/W with open-circuit responsivities of approximately 500 V/W. In addition, 1.5 micron detectors have been fabricated from p-type, 25 Ω.cm material. With no bias at room temperature, these detectors showed D* λ>10¹⁰cm.Hz/12/W, open-circuit responsivities in excess of 10³V/W, and response times on the order of microseconds. These preliminary results indicate that detectors fabricated from the pseudobinary alloy of (Hg, Cd)Te are well suited for high speed, near infrared photodetection in which room temperature operation is required.     

Properties of ion-implanted junctions in mercury—cadmium—telluride

The formation of n-p junctions by ion-implantation in Hg0.71Cd0.29Te is shown to be a result of implantation damage. n-p photodiodes have been made by implantation of Ar, B, Al, and P in a p-type substrate with acceptor concentration of 4 × 10¹⁶cm^-3. The implanted n-type layer is characterized by sheet electron concentration of 10¹⁴to 10¹⁵cm^-2and electron mobility higher than 10³cm². V^-1. s^-1, for ion doses in the range 10¹³-5 × 10¹⁴cm^-2. The photodiodes have a spectral cutoff of 5.2 µm, quantum efficiency higher than 80 percent, and differential resistance by area product above 2000 Ω . cm²at 77 K. The temperature dependence of the differential resistance is discussed. The junction capacitance dependence on reverse voltage fits a linearly graded junction model. Reverse current characteristics at 77 K have been investigated using gate-controlled diodes. The results suggest that reverse breakdown is dominated by interband tunneling in field-induced junctions at the surface, for both polarities of surface potential.     

BF_2~+,F~+ +B~+和Ar~++B~+注入硅快速退火剩余损伤及对浅P~+N结的影响(英文)

The residual damage is analysed by transmission electron microscopy (TEM) for BF2+, F+ + B+ and Ar+ + B+ implanted silicon after rapid thermal annealing(RTA). And the reverse leakage current of the implanted diodes is measured using a FJ-356 electrometer. The results show that 1 ) The residual damage due to BF2+ implantation is less than that of F+ + B + and Ar++ B+ implantation. 2) The reverse leakage current of BF2+ implanted diodes is less than that of F+ + B+ and Ar++ B + implanted diodes. 3) The reverse leakage current of F++B+ and Ar++ B+ implanted diodes increases with the increase of F+ and Ar+ energies, respectively. Therefore the physical behaviour of the interaction between molecular ion and silicon is different from that of the interaction between individual atom ion and silicon.     

Dependency of p-n junction depth on ion species implanted in HgCdTe

We assessed p-n junction depth and stress in HgCdTe implanted with various ion species. After post-implantation annealing, junction depth was measured using a differential Hall measurement. The stress induced by implantation was determined by measuring wafer curvature or using Raman spectroscopy. Samples with B or Mg implantation had p-n junctions that were three to five times deeper than samples with Ca, Cd, or Hg implantation. The implantation of elements with a small ionic radius, such as B or Mg, leads to compressive stress in implantation layers. In contrast, the implantation of Ca, Cd, or Hg, which are ions with a radius similar to that of Hg cations, leads to a comparatively high tensile stress in the layers. The results indicate that the stress is caused by a change in volume when the implanted ions substitute the Hg sublattice and by an increase in volume caused by the generation of vacancies and interstistials. These results suggest that implantation-induced stress is an important factor influencing the depth of p-n junctions.     

Interdiffusion behavior of Hg in HgTe/CdTe superlattices grown on Cd0.96Zn0.04 Te (211)B substrates by molecular beam epitaxy

Double-crystal x-ray rocking curve (DCRC) and secondary-ion mass-spectroscopy (SIMS) measurements have been performed to investigate the effect of rapid thermal annealing on the interdiffusion behavior of Hg in HgTe/CdTe superlattices grown on Cd_0.96Zn_0.04Te (211)B substrates by molecular beam epitaxy. The sharp satellite peaks of the DCRC measurements on a 100-period HgTe/CdTe (100Å/100Å) superlattice show a periodic arrangement of the superlattice with high-quality interfaces. The negative direction of the entropy change obtained from the diffusion coefficients as a function of the reciprocal of the temperature after RTA indicates that the Hg diffusion for the annealed HgTe/CdTe superlattice is caused by an interstitial mechanism. The Cd and the Hg concentration profiles near the annealed HgTe/CdTe superlattice interfaces, as measured by SIMS, show a nonlinear behavior for Hg, originating from the interstitial diffusion mechanism of the Hg composition. These results indicate that a nonlinear interdiffusion behavior is dominant for HgTe/CdTe superlattices annealed at 190°C and that the rectangular shape of HgTe/CdTe superlattices may change to a parabolic shape because of the intermixing of Hg and Cd due to the thermal treatment.