Total ionizing dose effects in shallow trench isolation oxides

doi:10.1016/j.microrel.2008.04.004

Microelectronics Reliability

Volume 48, Issue 7, July 2008, Pages 1000-1007

https://doi.org/10.1016/j.microrel.2008.04.004 Get rights and content

Abstract

The peaked evolution of leakage current with total ionizing dose observed in transistors in 130 nm generation technologies is studied with field oxide field effect transistors (FOXFETs) that use the shallow trench isolation as gate oxide. The overall radiation response of these structures is determined by the balance between positive charge trapped in the bulk of the oxide and negative charge in defect centers at its interface with the silicon substrate. That these are mostly interface traps and not border traps is demonstrated through dynamic transconductance and variable-frequency charge-pumping measurements. These interface traps, whose formation is only marginally sensitive to the bias polarity across the oxide, have been observed to anneal at temperatures as low as 80 °C. At moderate or low dose rate, the buildup of interface traps more than offsets the increase in field oxide leakage due to oxide-trap charge. Consequences of these observations for circuit reliability are discussed.

Introduction

Radiation tests of CMOS transistors in the 130 nm node from several sources show that 2.2 nm thick gate oxides are virtually unaffected by total ionizing dose (TID) irradiation up to multi-Mrd(SiO₂) levels [1], [2]. This property has been confirmed by noise measurements, which have not evidenced any significant buildup of defects close to the oxide–channel interface (1/f noise is particularly sensitive to the presence of such defects) [3].

The shallow trench isolation (STI) oxide that is used to isolate transistors in this technology node still has a thickness of several 100 nm. Therefore, this oxide is expected to be very sensitive to radiation. Documented consequences of radiation-induced charge trapping in the STI seriously affecting the functionality of transistors are source–drain leakage currents in NMOS transistors and width-dependent threshold voltage shifts of both NMOS and PMOS transistors (an effect that has been called radiation-induced narrow channel effect, RINCE) [1].

All these effects have been shown to increase with TID during a continuous irradiation at laboratory dose rates (about 25 krd(SiO₂)/min) and room temperature up to about 3–6 Mrd(SiO₂). When irradiation is continued beyond that point under the same conditions, leakage currents and threshold voltage shifts of NMOS transistors decrease towards pre-irradiation values, while threshold voltage shifts of PMOS transistors continue to increase [1]. This is shown in Fig. 1.

To understand the radiation effects on the survival or failure of integrated circuits (ICs) in a radiation environment, it is important to consider that significant threshold voltage shifts (of magnitude ∼50–80 mV) in PMOS FETs have been observed exclusively on transistors with the smallest gate width, ΔV_th becoming quickly negligible with increased W (RINCE). Therefore, the main parameters that influence the survival of the IC – the most relevant being the leakage current – are those affecting NMOS transistors and exhibiting an evolution “peaking” at a given TID level.

It has been demonstrated for MOS technologies that the value of the peak degradation typically depends on the dose rate, temperature, and applied bias voltage [4]. Therefore the radiation response of ICs in the real application can only be predicted once the main properties of the trapping centers in the STI oxide are known.

Section snippets

Test structures

The study was carried out on custom-designed samples from a commercial 130 nm CMOS technology available via a multi-project-wafer service. It concentrated on field oxide field effect transistors (FOXFETs), devices using the STI oxide as the gate dielectric, as shown in Fig. 2. The standard measurement techniques used to study radiation effects on gate oxides of MOS transistors in this case allow the study of defect buildup in the STI oxide and at its bottom interface with the silicon substrate.

Radiation-induced defects in the STI oxide

Measurements aimed at studying the characteristics of the defects induced by radiation in the STI have been performed on the FOXFETs with polysilicon gate and n-well source and drain with different techniques.

Subthreshold swing technique

The evolution of the drain current vs. gate voltage of FOXFETs irradiated with X-rays is shown in Fig. 5.

The threshold voltage V_th, extracted with a linear fit on the square root of the I_d(V_gs) curve in saturation, shifts to smaller values at the beginning of the irradiation, and it later

Discussion

Our results indicate that the peaked evolution of both ΔV_th and leakage current in NMOS transistors shown in Fig. 1 is determined by the balance between positive charge trapped in the bulk of the STI oxide and negative charge trapped at its interface with the silicon substrate. In the case of PMOS FETs, positive charge is trapped in both types of defects and contributes to the steady increase of the V_th in Fig. 1.

From dynamic transconductance and charge-pumping measurements, it appears that the

Summary and conclusion

A comparison of the evolution of NMOS transistor leakage and threshold voltage of FOXFETs provides a good correlation between the TID response of sidewall and bottom of the STI in a 130 nm generation technology. This suggests uniform trends in defect buildup near all STI interfaces, and indicates that FOXFETs can be effectively used to study the properties of the trapping centers responsible for NMOS source–drain leakage.

The peaked evolution of leakage current observed in NMOS transistors at

Acknowledgements

The authors are grateful to Ali Mohammadzadeh and Bob Nickson at EAS-ESTEC for their help in accessing the ⁶⁰Co irradiation source for the low dose rate experiments, and for their assistance in setting up and running the irradiation test.

References (30)

L. Gonella
Total ionizing dose effects in 130-nm commercial CMOS technologies for HEP experiments
Nucl Instrum Meth Phys Res Section A
(2007)
D.M. Fleetwood et al.
Border traps: issues for MOS radiation response and long-term reliability
Microelectron Reliab
(1995)
F. Faccio et al.
Radiation-induced edge effects in deep submicron CMOS transistors
IEEE Trans Nucl Sci
(2005)
V. Re et al.
Total ionizing dose effects on the noise performances of a 0.13 μm CMOS technology
IEEE Trans Nucl Sci
(2006)
P.S. Winokur et al.
Total-dose radiation and annealing studies: implications for hardness assurance testing
IEEE Trans Nucl Sci
(1986)
L. Dusseault
Prediction of low dose-rate effects in power MOSFETs based on isochronal annealing measurements
J Appl Phys
(1997)
F. Saigné
Experimental validation of an accelerated method of oxide-trap-level characterization for predicting long term thermal effects in metal oxide semiconductor devices
IEEE Trans Nucl Sci
(1997)
T. Want et al.
Characterization of various stress-induced oxide traps in MOSFET’s by using a subthreshold transient current technique
IEEE Trans Electr Dev
(1998)
D.K. Schroder
Semiconductor material and device characterization
(1998)
G. Groeseneken et al.
A reliable approach to charge-pumping measurements in MOS transistors
IEEE Trans Electr Dev
(1984)

X.J. Chen et al.

Estimation and verification of radiation induced N_ot and N_it energy distribution using combined bipolar and MOS characterization methods in gated bipolar devices

IEEE Trans Nucl Sci

(2005)

X.J. Chen et al.

Nature of interface defect buildup in gated bipolar devices under low dose rate irradiation

IEEE Trans Nucl Sci

(2006)

R.E. Paulsen et al.

Observation of near-interface oxide traps with the charge pumping technique

IEEE Electr Dev Lett

(1992)

P.J. McWhorter et al.

Simple technique for separating the effects of interface traps and trapped-oxide charge in metal-oxide-semiconductor field-effect transistors

Appl Phys Lett

(1986)

M.R. Shaneyfelt et al.

Wafer-level radiation testing for hardness assurance

IEEE Trans Nucl Sci

(1991)

Cited by (87)

ASIC survival in the radiation environment of the LHC experiments: 30 years of struggle and still tantalizing
2023, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
The radiation environment at the CERN Large Hadron Collider (LHC) poses an unprecedented challenge to the reliable functionality of the electronics components composing the detector systems. Starting from the beginning of the 90s, a large effort has been dedicated at CERN and in the collaborating institutes to develop, produce and deploy Application-Specific Integrated Circuits (ASICs) tolerant to all types of radiation effects: Total Ionizing Dose (TID), Single Event Effects (SEE) and Displacement Damage (DD). This article briefly describes the approaches and methodologies that the High Energy Physics community developed or adopted for that purpose over the last 30 years.
Development and testing of a radiation-hard large-electrode DMAPS design in a 150 nm CMOS process
2022, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
The LF-Monopix chips are depleted monolithic active pixel sensors that follow the large-electrode design approach and implement a fast synchronous read-out architecture. They are designed in a 150 nm CMOS process and make use of large voltages ( $> 250 V$ ) and highly resistive substrates ( $> 2 kΩ \cdot cm$ ) to collect charge through drift and enhance their radiation hardness.
Samples of the first prototype (“LF-Monopix1”) with a thickness of $100 μ m$ were irradiated to assess the tolerance of the chip’s substrate and front-end circuitry to the surface and bulk damage doses expected at modern collider experiments. The device remained fully operational, with only a very small gain degradation and an increase in noise by less than $25 %$ after a total ionizing dose of $100 Mrad$ . Efficiency measurements in a sample exposed to a neutron fluence of $1 \times 1 0^{15} n_{eq} / {cm}^{2}$ showed that at least $96 %$ of all minimum ionizing particles going through a fully depleted detector are recorded in less than $25 ns$ . In the latest design (“LF-Monopix2”) the column length was tripled and the pixel pitch reduced by $40 %$ with respect to its predecessor. The chip was successfully thinned down while keeping its breakdown voltage above $400 V$ and achieving a front-end threshold dispersion of $\sim 100 e^{-}$ after tuning.
Role of the oxide trapped charges in charge coupled device ionizing radiation-induced dark signal
2021, Radiation Physics and Chemistry
Two types of charge-coupled devices (CCDs) have been irradiated by ⁶⁰Co γ-ray and dark signal is characterized before and after irradiation. After exposure to γ-ray, the evolution of mean dark signal exhibits an inflection point specific to a certain total ionizing dose (TID) value, beyond which dark signal increase rate rapidly reaches a level higher than the ones measured at other TIDs. The cause of inflection point occurrence is analyzed through the use of annealing tool and technology computer-aided design (TCAD) simulation in the custom CCDs. The experimental results demonstrate that when the cumulative oxide charge density achieves a specific value which depends on the fabrication process of the field oxide of CCDs, a sudden increase in dark signal occurs due to a merge between the space charge region and the additional depletion region induced by the oxide trapped charges.
Total ionizing dose hardness analysis of transistors in commercial 180 nm CMOS technology
2021, Microelectronics Journal
The very large scale integrated (VLSI) circuits become susceptible to soft errors mainly due to exposure to harsh environmental conditions. In this paper, the analysis of total ionizing dose (TID) effects on single gate NMOS, enclosed gate NMOS (ELT NMOS), and single gate PMOS are carried out for 180 nm CMOS technology. The various MOS architectures were exposed to radiation using Cobalt-60 $(C o^{60}$ ) radiation source. The charge density distribution at pre-radiation and post-radiation condition for all the three above said structures are observed. It is observed that the threshold voltage shift for the single gate NMOS device is about 25 times more in comparison to the ELT NMOS device after radiation at 30 Å (Å) gate oxide thickness. The transfer characteristics (drain current versus gate to source voltage) of single gate NMOS is compared for TID from 0 krad to 500 krad with a step size of 100 krad at different oxide thicknesses from 15 Å to 75 Å. It is observed that the drain current increases with an increase in total dose in the sub-threshold region for lower gate oxide thickness (15 Å to 45 Å) while with the increase of gate oxide thickness (beyond 60 Å), the drain current increases significantly for the higher gate to source voltage also. The drain saturation current decreases with the increase of oxide thickness from 15 Å to 75 Å. The impact of TID on single gate PMOS is analysed and no significant effect of radiation observed on threshold voltage or leakage current. The leakage current shift is the least for single gate PMOS among the single gate NMOS, ELT NMOS, and single gate PMOS after radiation of 500 krad total dose. The layouts are designed in Cadence virtuoso software and simulated in Visual TCAD software.
Ionizing radiation damage in 65 nm CMOS technology: Influence of geometry, bias and temperature at ultra-high doses
2021, Microelectronics Reliability
We studied the radiation response of 3 different 65 CMOS planar technologies at the ultra-high doses expected to be reached in the HL-LHC, the upgraded large hadron collider of CERN. All the processes studied are sensitive to radiation and show similar degradation mechanisms and, albeit with different intensities, similar dependencies on device geometry, applied polarization and temperature. The results obtained confirmed that the performance of MOS transistors exposed to ultra-high doses is mainly affected by radiation-induced charge trapped in auxiliary oxides such as shallow trench isolation oxides and spacers. The extensive data collected has been used to develop guidelines to qualify to ultra-high doses ASICs designed in the 65 technology node.
Thermo-electrical modelling of the ATLAS ITk Strip Detector
2020, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
In this paper we discuss the use of linked thermal and electrical network models to predict the behaviour of a complex silicon detector system. We use the silicon strip detector for the ATLAS Phase-II upgrade to demonstrate the application of such a model and its performance. With this example, a thermo-electrical model is used to test design choices, validate specifications, predict key operational parameters such as cooling system requirements, and optimize operational aspects like the temperature profile over the lifetime of the experiment. The model can reveal insights into the interplay of conditions and components in the silicon module, and it is a valuable tool for estimating the headroom to thermal runaway, all with very moderate computational effort.

View all citing articles on Scopus

View full text

Total ionizing dose effects in shallow trench isolation oxides

Abstract

Introduction

Section snippets

Test structures

Radiation-induced defects in the STI oxide

Subthreshold swing technique

Discussion

Summary and conclusion

Acknowledgements

Nucl Instrum Meth Phys Res Section A

Microelectron Reliab

Radiation-induced edge effects in deep submicron CMOS transistors

IEEE Trans Nucl Sci

Total ionizing dose effects on the noise performances of a 0.13 μm CMOS technology

IEEE Trans Nucl Sci

Total-dose radiation and annealing studies: implications for hardness assurance testing

IEEE Trans Nucl Sci

Prediction of low dose-rate effects in power MOSFETs based on isochronal annealing measurements

J Appl Phys

Experimental validation of an accelerated method of oxide-trap-level characterization for predicting long term thermal effects in metal oxide semiconductor devices

IEEE Trans Nucl Sci

Characterization of various stress-induced oxide traps in MOSFET’s by using a subthreshold transient current technique

IEEE Trans Electr Dev

Semiconductor material and device characterization

A reliable approach to charge-pumping measurements in MOS transistors

IEEE Trans Electr Dev

Estimation and verification of radiation induced Not and Nit energy distribution using combined bipolar and MOS characterization methods in gated bipolar devices

IEEE Trans Nucl Sci

Nature of interface defect buildup in gated bipolar devices under low dose rate irradiation

IEEE Trans Nucl Sci

Observation of near-interface oxide traps with the charge pumping technique

IEEE Electr Dev Lett

Simple technique for separating the effects of interface traps and trapped-oxide charge in metal-oxide-semiconductor field-effect transistors

Appl Phys Lett

Wafer-level radiation testing for hardness assurance

IEEE Trans Nucl Sci

Estimation and verification of radiation induced N_ot and N_it energy distribution using combined bipolar and MOS characterization methods in gated bipolar devices