Radiation-induced Interface Trap Research Articles

Correlation of Radiation-Induced Interface Traps With Band Edge Energy Through Band Structure-Based Analysis of Electrostatics of UTB SOI Devices

Correlation of Radiation-Induced Interface Traps With Band Edge Energy Through Band Structure-Based Analysis of Electrostatics of UTB SOI Devices

Total Ionizing Dose Effects on DRAM Data Retention Time

Dynamic random access memory (DRAM) data retention time degradation induced by radiation exposure is investigated in this paper. We present the experimental setup and the results of total ionizing dose (TID) test on a COTS SDRAM device. We observed a significant retention time reduction related to the absorbed dose and we assume radiation induced interface trap generation as the origin of the retention time reduction. By measuring individual cells retention time before and after radiation exposures, we found out that the reduction is not homogeneous among cells: the amount of leakage current increase depends on the position and the energy level of the generated trap, leading to a wide distribution of retention time reduction. Of particular interest is the fact that device was unbiased during irradiation and that no post-irradiation recovery was observed.

Mechanism of the Saturation of the Radiation Induced Interface Trap Buildup

Ionizing radiation impact leads to degradation of electrical parameters of microelectronic devices. It is necessary to take this fact to account when dealing with microcircuits for space applications and high energy physics. Main physical reason of radiation-induced failures of spaceship and front end electronic equipment is buidup of interface traps at Si-SiO2 interface in semiconductor transistor structures. The original mechanism of interface trap annealing based on radiation induced charge neutralization (RICN) effect is presented. It is supposed that the positive charge of trapped holes in oxide is transformed through electron capture into a new defect (the AD center). The AD centers act as interface traps. The appearance of the A–D+ state leads to the annihilation of the AD center or annealing of interface trap. The annihilation process can be stimulated by radiation induced or substrate electrons. The competitive between accumulation and annihilation processes leads to saturation of the interface trap buildup. The value of density of interface trap in saturation depends on product of interface trap accumulation rate (Kacc)it and constant KAD which is function of thermal velocity, capture cross-section of AD center, generation rate and electron yield of radiation induced electrons. The extraction of these parameters allows explaining a known experimental data. The alternative mechanism of the interface trap saturation connected with the exhaustion of initial interface trap precursors is considered.

The Effects of Proton-Defect Interactions on Radiation-Induced Interface-Trap Formation and Annealing

Interface-trap buildup and annealing are examined as a function of temperature, dose rate, and H2 concentration using a physics-based model. The roles of a number of common oxide defects in radiation-induced interface-trap buildup are evaluated under various conditions. Defects near the interface play a significant role in determining interface-trap buildup by trapping protons as proton concentration and temperature increase.

Charge Pumping Measurements of Radiation-Induced Interface-Trap Density in Floating-Body SOI FinFETs

We demonstrate that, by monitoring source and drain currents during alternating-current gate pulses, reliable estimates of radiation-induced interface-trap density can be obtained for conventional floating-body SOI FinFETs without body contacts. Estimates of effective interface-trap densities are shown for two development stage technologies, before irradiation, and for doses up to 1 Mrad(SiO2). Straightforward estimates of effective interface-trap density are obtained for these floating-body FinFETs, with either a high-K or oxynitride gate dielectric, when the fin width is less than ~70 nm, the channel length is less than ~100 nm, and there is full gate control of the relevant interfaces. This modified charge pumping technique provides estimates of interface-trap density without detailed analysis, adjustable fitting parameters, and/or device simulation, in contrast to DCIV and/or gated-diode techniques. Moreover, this technique enables more accurate estimates of radiation-induced interface-trap density in floating-body FinFETs than midgap charge separation in cases for which isolation leakage contributes significantly to subthreshold current-voltage stretchout.

Modeling the Non-Uniform Distribution of Radiation-Induced Interface Traps

The incorporation of total ionizing dose (TID) effects into surface-potential-based compact models requires calculating the dependence of surface potential $({\psi}_{s})$ on radiation-induced defect densities. This dependence is obtained through the introduction of the oxide trapped charge $(N_{ot})$ and the interface trap $(N_{it})$ areal densities into the surface potential equation (SPE). In this work we present an approach for introducing a non-uniform energy distribution of interface traps into calculations of ${\psi}_{s}$ . The approach is verified experimentally through comparisons with capacitance vs. voltage ( $C$ - $V$ ) characteristics of metal-oxide-semiconductor (MOS) capacitors exposed to ionizing radiation.

Gate length dependence of SOI NMOS device response to total dose irradiation

The gate length dependence of PD SOI NMOS device on total dose irradiation is investigated, which is exposed to 60Co gamma ray at a dose rate of 50 rad(Si)/s. The result shows that the transistor with shorter gate length shows larger radiation-induced interface trap density, which leads to the maximum transconductance degradation. The local floating body effect induces the output characteristic variation of irradiated MOSFET with gate length. After irradiation, the breakdown voltage of short channel SOI device decreases. Due to the buried oxide, the radiation-induced degradation of short channel SOI device is much serious compared with that of long channel SOI device.

The Effects of Aging and Hydrogen on the Radiation Response of Gated Lateral PNP Bipolar Transistors

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> Complex interplay between hydrogen-related defect formation and passivation is observed in irradiated bipolar transistors. Hydrogen soaking experiments are performed to evaluate the dependence of defect buildup and annealing in gated lateral bipolar transistors on hydrogen exposure. Comparisons of the radiation responses of transistors tested in 2009 to identical devices from the same wafer tested in 2003 show that aging has reduced the amount of radiation-induced interface trap and oxide trapped charge formation in most cases. These results demonstrate that the way in which the radiation response of a hydrogen-sensitive device evolves with age depends on whether hydrogen is diffusing into or out of the device, and whether the initial defect concentration favors passivation or depassivation reactions. These results strongly suggest that hydrogen exposure cannot replace low-dose-rate irradiation in ELDRS tests for bipolar devices and ICs without extensive characterization testing. </para>

Mechanisms of Enhanced Radiation-Induced Degradation Due to Excess Molecular Hydrogen in Bipolar Oxides

Bipolar junction test structures packaged in hermetically sealed packages with excess molecular hydrogen (H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ) showed enhanced degradation after radiation exposure. Using chemical kinetics, we propose a model that quantitatively establishes the relationship between excess H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> and radiation-induced interface trap formation. Using environments with different molecular hydrogen concentrations, radiation experiments were performed and the experimental data showed excellent agreement with the proposed model. The results, both experimentally and theoretically, showed increased radiation induced degradation with H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> concentration, and device degradation saturate at both high and low ends of H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> concentrations.

A 1/f noise based research of radiation induced interface trap buildup process

Based on the two-step interface trap buildup model and the statistical thermodynamics mechanism of point defects in solids, a relation between the radiation induced increase of oxide hole-traps and the buildup of interface traps in MOSFET is proposed. Then, based on the correlation between pre-irradiation 1/f noise power spectral density and post-irradiation oxide-trap charge, a sim-empirical expression of pre-irradiation 1/f noise and post-irradiation interface traps buildup is established which agrees well with the experimental results. This model shows that the process of interface trap buildup was influenced by the increasing of oxide hole-traps that could dissociated hydrogen into proton under irradiation. So, the magnitude of pre-irradiation 1/f noise is directly proportional to post-irradiation interface-trap charge and sub-threshold slope. Our results not only proves that the 1/f noise can play a important role in MOSFET irradiation effects, but also proves a new nondestructive method to estimate MOSFET radiation response.

Dose-rate dependence of radiation-induced interface trap density in silicon bipolar transistors

The key effects of ionizing radiation on silicon dioxide are described and computed. Inclusion of bimolecular electron–hole recombination is shown to produce a saturation of fixed charge at high total dose and to a reduction in interface trap density at high dose-rates. These results can explain the enhanced low dose rate sensitivity (ELDRS) phenomenon. These results also predict a new dose-rate dependence at very high dose rates.

Estimation and verification of radiation induced N/sub ot/ and N/sub it/ energy distribution using combined bipolar and MOS characterization methods in gated bipolar devices

Complementary bipolar and MOS characterization techniques, specifically the gate sweep (GS) and sub-threshold sweep (SS), are used to estimate the radiation induced oxide charge (N/sub ot/) and interface trap (N/sub it/) buildup in gated bipolar test devices. The gate sweep and sub-threshold sweep data from recent TID testing of gated lateral PNP devices suggests an asymmetric energy distribution of interface traps after ionizing radiation exposure. Charge pumping (CP) experiments were done on the test devices to estimate the energy distribution of interface traps induced by radiation. The CP results are used in this paper to confirm the analytical findings from the GS and SS techniques and solidify the use of the complementary method as a simple way of determining radiation induced interface trap distribution in gated bipolar devices.

Common origin for enhanced low-dose-rate sensitivity and bias temperature instability under negative bias

Degradation due to irradiation is known to be associated with the presence of hydrogen in the bulk of the gate oxide, in bulk Si, and at the Si/SiO/sub 2/ interface. Previous studies have shown that the migration of protons in the oxide for positive applied gate bias and their reactions at the interface can account for the time and dose-rate dependence of the degradation. Recent experiments, however, have shown that interfacial degradation can occur even in the presence of strong negative gate bias that prevents the arrival of protons at the interface from the oxide side. This result suggests that mechanisms in addition to proton drift in SiO/sub 2/ can lead to radiation-induced interface-trap formation. Since previous work on modeling the enhanced low-dose-rate sensitivity (ELDRS) of irradiated bipolar devices was based on formation of an electrostatic barrier that hinders proton transport to the interface at high dose rates, this effect also must be examined in more detail. In this work we use results from first-principles calculations to demonstrate that hydrogen can also be released easily in bulk Si, and especially in the near interfacial area. This hydrogen moves readily to the interface under negative bias. Typical hydrogen precursors in Si are identified as H-dopant complexes. ELDRS shares thus a common origin with another critical reliability phenomenon, the negative bias-temperature instability.

Radiation hardness comparison of MOS capacitors using tungsten polycide and cobalt polycideas gate electrode materials

We investigate the radiation hardness of metal-oxide-semiconductor (MOS) capacitors with tungsten polycide (WSi/sub x/) and those with cobalt polycide (CoSi <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> ) as gate electrode materials. CoSi <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> has been considered as a gate/contact material for MOS devices in 0.18 μm integrated circuit fabrication due to its low resistivity and good thermal stability. However, we found that MOS capacitors with a CoSi <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> gate electrode exhibited an increase in radiation-induced interface trap density shift of more than one order of magnitude, and more than eighteen times larger in radiation-induced flatband voltage shifts compared with those with the WSi/sub x/ gate electrode, after 1 Mrad Co/sup 60/ /spl gamma/-ray irradiation under no applied bias.

Influence of total-dose radiation on the electrical characteristics of SOI MOSFETs

Silicon-on-insulator (SOI) technologies have been developed for radiation-hardened applications for many years and are rapidly becoming a main-stream commercial technology. The total dose response of SOI devices is more complex than for bulk-silicon devices due to the buried oxide. Radiation-induced trapped charge in the buried oxide can increase the leakage current of partially-depleted transistors and decrease the threshold voltage and increase the leakage current of fully-depleted transistors. We review process and design techniques that have been developed to reduce the net amount of radiation-induced positive charge trapped in the buried oxide or to mitigate the effects of trapped charge in the buried oxide, respectively. We also compare the total dose response of devices operated in single-gate (SG) and double-gate (DG) mode as a function of silicon film thickness. DG-MOSFETs are systematically less sensitive to radiation-induced trapped charge and interface trap.

Ionizing radiation and hot carrier effects in SiC MOS devices

We performed comparative studies of radiation-induced interface trap generation at n-6H-SiC/SiO2 and Si/SiO2 interfaces for high doses of low energy x-rays in the range 0 to 200 Mrad (Si). We found that radiation-induced interface trap generation at the n-6H-SiC/SiO2 interface is much lower than at the Si/SiO2 interface. The experiments provide evidence of interface trap generation near the SiC conduction band associated with oxide traps.

Effects of hydrogen transport and reactions on microelectronics radiation response and reliability

The transport and reactions of hydrogen-related species play critical roles in determining the ionizing radiation response and long-term reliability of Si-based metal-oxide-semiconductor (MOS) and bipolar microelectronics technologies. The role of hydrogen is reviewed in radiation-induced interface-trap formation in metal and polycrystalline-Si-gate devices, with an emphasis on understanding the transport and reaction mechanisms responsible for defect formation. Effects of hydrogen on radiation-induced oxide and border-trap charge densities are discussed. Enhanced low-dose-rate effects in bipolar devices and integrated circuits are sensitive to hydrogen transport and reactions in the oxides that overlie the emitter–base junctions. These enhanced low-dose-rate effects present great challenges for defining practical, cost-effective acceptance tests of bipolar devices for space radiation environments. In addition, a brief discussion is provided of the effects of hydrogen-related species on MOS long-term reliability, including high-field stress, dielectric leakage (e.g., stress-induced leakage current), oxide breakdown, hot-carrier effects, dopant passivation, and low-frequency (1/ f) noise. The potential impacts on latent interface-trap formation are discussed for retarded hydrogen transport, trapping, and/or dopant passivation. Further, differences in proton transport and reactions are noted for (a) conventional Si/SiO 2/Si structures and (b) those with high oxygen vacancy densities that are exposed to high-temperature hydrogen annealing ambients. In the former, hydrogen is consumed irreversibly by processes such as interface-trap formation; in the latter, hydrogen often can be cycled from the gate to the Si interface reversibly without losses for up to 10 6 cycles or more. Moreover, the inferred transport rates are much higher for post-forming-gas anneal transport than for radiation-induced interface-trap buildup. Finally, the motion of hydrogen during post-process, but pre-irradiation, thermal treatments (e.g., burn-in, or temperature cycles during the packaging process) has been associated with significant changes in the radiation response of both MOS and bipolar transistors. Typical mitigation techniques are discussed to reduce the effects of hydrogen on MOS radiation response and long-term reliability. Several unresolved issues and opportunities for future work are identified. The present knowledge of hydrogen effects on the long-term reliability and radiation response of Si based microelectronics will be broadened significantly as alternative dielectrics to SiO 2 are increasingly introduced into the manufacturing process. A brief example of the significant effects of hydrogen on dielectric layers other than SiO 2 is provided for deposited diamond thin films.

The determination of Si-SiO/sub 2/ interface trap density in irradiated four-terminal VDMOSFETs using charge pumping

The utility of charge pumping to measure Si-SiO/sub 2/ interface trap density in irradiated four-terminal VDMOSFETs is demonstrated. A modification of the conventional charge pumping approach is employed, where recombination of charge through interface traps in the neck region is measured in the drain. Three components of drain current resulting from the charge pumping measurement are identified. When the device is properly biased, charge pumping current can be separated from the other components of drain current and modeled over a wide range of interface trap densities using standard charge pumping theory. When sources of error are accounted for, radiation-induced interface trap densities measured with charge pumping are in good quantitative agreement with those estimated with the midgap charge separation and subthreshold hump techniques.

Oxide, interface, and border traps in thermal, N2O, and N2O-nitrided oxides

We have combined thermally stimulated-current (TSC) and capacitance–voltage (C–V) measurements to estimate oxide, interface, and effective border trap densities in 6–23 nm thermal, N2O, and N2O-nitrided oxides exposed to ionizing radiation or high-field electron injection. Defect densities depend strongly on oxide processing, but radiation exposure and moderate high-field stress lead to similar trapped hole peak thermal energy distributions (between ∼1.7 and ∼2.0 eV) for all processes. This suggests that similar defects dominate the oxide charge trapping properties in these devices. Radiation-induced hole and interface trap generation efficiencies (0.1%–1%) in the best N2O and N2O-nitrided oxides are comparable to the best radiation hardened oxides in the literature. After ∼10 Mrad(SiO2) x-ray irradiation or ∼10 mC/cm2 constant current Fowler–Nordheim injection, effective border trap densities as high as ∼5×1011 cm−2 are inferred from C–V hysteresis. These measurements suggest irradiation and high-field stress cause similar border trap energy distributions. In each case, even higher densities of compensating trapped electrons in the oxides (up to 2×1012 cm−2) are inferred from combined TSC and C–V measurements. These trapped electrons prevent conventional C–V methods from providing accurate estimates of the total oxide trap charge density in many irradiation or high-field stress studies. Fewer compensating electrons per trapped hole (∼26%±5%) are found for irradiation of N2O and N2O-nitrided oxides than for thermal oxides (∼46%±7%). More compensating electrons are also found for high-field electron injection than radiation exposure, emphasizing the significance of border traps to metal-oxide-semiconductor long term reliability. The primary effect of nitrogen on charge trapping in these oxides appears to be improvement of the near interfacial oxide in which border traps are found.

Effects of switched gate bias on radiation-induced interface trap formation (MOS transistors)

Switched gate bias experiments were used to test the hydrogen model for the time-dependent buildup of interface traps n/sub it/ after a radiation pulse. In the hydrogen model, slow transport of radiation-induced H/sup +/ ions through the oxide to the Si/SiO/sub 2/ interface is the rate-limiting step in the n/sub it/ formation process. A model based on dispersive transport theory was used to numerically simulate the time-dependent H/sup +/ drift through the oxide toward the gate under negative bias, then back toward the Si/SiO/sub 2/ interface under positive bias. Good agreement between the simulations and experiment is obtained for the final magnitude of Delta N/sub it/, both for positive and negative bias during irradiation. The procedure used for the numerical simulations provides only a lower bound on the transport time when the direction of the gate bias is switched during the N/sub it/ buildup. This behavior is due to a memory effect in the dispersive H/sup +/ transport. >