Fundamental Mechanisms for Single ParticleInduced Soft Errors

1
Fundamental Mechanisms for Single
Particle-Induced Soft Errors

• Robert A. Reed
• Vanderbilt University

2
Radiation Events in Space-Based Imagers
Extreme ultraviolet Imaging Telescope
The Large Angle and Spectrometric Coronagraph
- C2 coronagraph - C3 coronagraph.
Solar Heliospheric Observatory (SOHO)
http//www.nasa.gov/mpg/61466main_eitlasco_fall200
3_320x240.mpg
3
Radiation Events in Space-Based Imagers
Fig.1

• Sun emits energetic particles
• Distortion due ionizing particle events passing
  through the imager
• Energy deposited by the ion
• Example of a transient soft-error

Solar Heliospheric Observatory (SOHO)
http//www.nasa.gov/mpg/61466main_eitlasco_fall200
3_320x240.mpg
4
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Collection

5
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Collection

6
Ionizing Radiation Event Across a Diode
Fig.2
Potential Drop Across a Diode

• Ionization from incident particle via
  interactions with the target electrons
• Soft-error is a change of logic state from single
  ionizing radiation event
• This state change can be self- recoverable or
  stable until the circuit is reset
• Sensitive Volume Model

Courtesy Jonathan Pellish
7
Sensitive Volume Defined

• The sensitive volume (SV) is contained within the
  active region of the device
• Energy deposited in the SV will contribute to
  soft errors
• Primary factors that determine the SV shape
• Doping profiles
• Insulating structures
• Its shape may also be determined
• Energy deposited by the ion
• Location of energy deposition

SV1
SV2
8
Transients from Single Particle Event
Fig.3
Energy Deposition
Transient Current Pulse
Ei
Eo
fs to ps
Time
Soft Error Examples

• Single Event Transient A current pulse occurring
  at a circuit node due to single energetic
  particle event
• Single Event Upset A change in a circuits logic
  state induced by a single energetic particle event

9
Outline

• Introduction
• Interaction of Radiation with Matter
• Ion Stopping (Stopping Force)
• Nuclear Reaction (Nuclear Forces)
• Space Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Collection

10
Important Interactions of Radiation with Matter

• Primary radiation types responsible for Single
  Event Upsets
• Protons, Neutrons, and Heavier Ions
• Two Classes of Interaction Mechanisms
• Ion Stopping (Stopping Force)
• Electronic stopping
• Coulomb recoil of target atom (nuclear
  stopping)
• Electromagnetic radiation
• Projectile excitation and ionization
• Electron capture by incident particle
• Chemical reactions
• Nuclear Reaction (Nuclear Forces)
• Inelastic
• Fission
• Elastic

11
Important Interactions of Radiation with Matter

• Primary radiation types responsible for Single
  Event Upsets
• Protons, Neutrons, and Heavier Ions
• Two Classes of Interaction Mechanisms
• Ion Stopping (Stopping Force)
• Electronic stopping
• Coulomb recoil of target atom (nuclear
  stopping)
• Electromagnetic radiation
• Projectile excitation and ionization
• Electron capture
• Chemical reactions
• Nuclear Reaction (Nuclear Forces)
• Inelastic
• Fission
• Elastic

12
Important Interactions of Radiation with Matter

• Primary radiation types responsible for Single
  Event Upsets
• Protons, Neutrons, and Heavier Ions
• Two Classes of Interaction Mechanisms
• Ion Stopping (Stopping Force)
• Electronic stopping
• Coulomb recoil of target atom (nuclear
  stopping)
• Electromagnetic radiation
• Projectile excitation and ionization
• Electron capture
• Chemical reactions
• Nuclear Reaction (Nuclear Forces)
• Inelastic
• Fission
• Elastic

13
Electronic Stopping - Ionization
Fig.4
Electron
Incident Ion
nucleus
14
Electronic Stopping - Ionization
Fig.4
15
Electronic Stopping - Ionization
Fig.4
Electromagnetically scattering through elastic
Coulomb collisions Ionization of the target
atoms electrons
16
Electronic Stopping - Ionization
Fig.4
Electromagnetically scattering through elastic
Coulomb collisions Ionization of the target
atoms electrons
All soft errors are a result of ionization -
Direct Ionization
17
Important Interactions of Radiation with Matter

• Primary radiation types responsible for Single
  Event Upsets
• Protons, Neutrons, and Heavier Ions
• Two Classes of Interaction Mechanisms
• Ion Stopping (Stopping Force)
• Electronic stopping
• Coulomb recoil of target atom (nuclear
  stopping)
• Electromagnetic radiation
• Projectile excitation and ionization
• Electron capture
• Chemical reactions
• Nuclear Reaction (Nuclear Forces)
• Inelastic
• Fission
• Elastic

18
Nuclear Stopping
Fig.5
19
Nuclear Stopping - Recoil of Target Atom
Fig.5
Coulomb collisions between the ion and the atom
nucleus field Screened by the atom electrons
20
Nuclear Stopping - Recoil of Target Atom
Fig.5

• Deflection of incoming ion transfers momentum to
  the target nucleus
• Target ion remains fixed for an electronic
  stopping event
• Indirect ionization
• Ionization from secondary particle

21
Stopping Force (LET)

• Mean energy loss (E) per path length (x)
• Stopping Force
• Stopping Power
• Linear Energy Transfer (LET)

• Fluctuations in LET (energy-loss straggling)
  determine
• Penetration depth (range) and its fluctuation
  (range straggling)
• energy-deposition profile

Journal of the ICRU Vol 5 No 1 (2005) Report 73
22
Stopping Force (LET)

• Mean energy loss (E) per path length (x)
• Stopping Force
• Stopping Power
• Linear Energy Transfer (LET)

• Mass Stopping Force
• MeV-cm2/mg
• density of target
• Energy Deposited (E) over path length (l)

• Fluctuations in LET (energy-loss straggling)
  determine
• Penetration depth (range) and its fluctuation
  (range straggling)
• energy-deposition profile

Journal of the ICRU Vol 5 No 1 (2005) Report 73
23
Model Results for Stopping Force
Fig.6

• Trend over energy occurs for all ions in all
  target materials
• Details depend on physical quantities like
  charge of incident ion

www.srim.org
www.srim.org
24
Model Results for Stopping Force
Fig.6

• Trend over energy occurs for all ions in all
  target materials
• Details depend on physical quantities like
  charge of incident ion
• Bragg Peak
• Different magnitude and energy for other ions and
  materials

www.srim.org
www.srim.org
25
Model Results for Stopping Force
Fig.6

• Trend over energy occurs for all ions in all
  target materials
• Details depend on physical quantities like
  charge of incident ion
• Bragg Peak
• Different magnitude and energy for other ions and
  materials
• LET LETelec LETnuc

www.srim.org
www.srim.org
26
Model Results for Stopping Force
Fig.6

• Trend over energy occurs for all ions in all
  target materials
• Details depend on physical quantities like
  charge of incident ion
• Bragg Peak
• Different magnitude and energy for other ions and
  materials
• LET LETelec LETnuc

www.srim.org
www.srim.org
27
Model Results for Stopping Force
Fig.6

• Trend over energy occurs for all ions in all
  target materials
• Details depend on physical quantities like
  charge of incident ion
• Bragg Peak
• Different magnitude and energy for other ions and
  materials
• LET LETelec LETnuc

t
LET determines Average Energy Loss
Ei gt Eo
www.srim.org
www.srim.org
28
Model Results for Stopping Force
Fig.6

• Trend over energy occurs for all ions in all
  target materials
• Details depend on physical quantities like
  charge of incident ion
• Bragg Peak
• Different magnitude and energy for other ions and
  materials
• LET LETelec LETnuc
• Energy-loss straggle produces a variation in
  energy deposition

t
LET determines Average Energy Loss
Ei gt Eo
www.srim.org
www.srim.org
29
Protons Stopping in Silicon
Fig.7

• 28 experiments to measure stopping power of
  protons in silicon
• Theory agrees with experimental results energies
  greater than few 100s keV
• Large disagreement of experiment with theory near
  Bragg peak
• Can be significant soft-error source of error
  when attempting to test near Bragg peak
• Commercial SRAM sensitivity to low energy protons

www-nds.iaea.org/stoppinggraphs/stopp_bot.htm
www.srim.org
30
Variation in Energy Deposition
ED1 lt ED2 lt ED3
1
2
Energy straggle Large sensitive volume as
compare to the effects of multiple scattering
31
Example Transient Events in Silicon FPAs
Fig.8

• Focal Plane Array (FPA)
• Image sensing device consisting of an array of
  light-sensing pixels

Hybrid P-i-N FPA
one pixel
Si Detector Array
substrate
p
intrinsic
n
Indium bump
Si Readout IC

Vreadout
Vreadout
Vreadout
Vreadout




Note not to scale
Howe et al, TNS 2007
32
Experimental Results
Fig.9

• - 63 MeV protons
• Range 19 mm
• Fluence is protons/cm2

energy deposited

Vreadout
Vreadout
Vreadout
Vreadout




Convert electrical signal to energy deposited
Isolated pixels
Howe et al, TNS 2007
33
Average Energy Deposition
Fig.9
Comparison of Experiment to Maximum Energy
Prediction by Average LET
Pixel 1
l
Emax LETelec ? l ? density

• Average value for LET does not predict response

Ave
Howe et al, TNS 2007
34
Detailed Energy Deposition
Fig.9

• Monte Carlo tools contain comprehensive set of
  physical models for radiation transport
• Geant4, MCNPX, Fluka, PHITS, JQMD, etc..
• Useful for computing energy deposition from an
  ensemble of radiation events
• Ion Stopping
• Straggling
• Multiple scattering
• Nuclear reactions
• MRED is Vanderbilts Geant4 application

MRED
Howe et al, TNS 2007
35
Average and Atypical Energy Deposition
Fig.10

• Ion LET determines the average energy deposited
  in sensitive volumes
• Extreme events often determine the onset of an
  observed radiation effect
• Extreme events for lightly ionizing particles
  (e.g., protons and alphas) can dominate the
  response
• This will be important for circuits with low
  switching energies
• For example, commercial CMOS memories

MRED Prediction of 100 MeV proton events
Weller et al., TNS 2003
36
Important Interactions of Radiation with Matter

• Primary radiation types responsible for Single
  Event Upsets
• Protons, Neutrons, and Heavier Ions
• Two Classes of Interaction Mechanisms
• Ion Stopping (Stopping Force)
• Electronic stopping
• Coulomb recoil of target atom (nuclear
  stopping)
• Electromagnetic radiation
• Projectile excitation and ionization
• Electron capture
• Chemical reactions
• Nuclear Reaction (Nuclear Forces)
• Inelastic
• Fission
• Elastic

37
Nuclear reactions (Inelastic)
Fig.11
Proton
38
Nuclear reactions (Inelastic)
Fig.11
39
Nuclear reactions (Inelastic)
Fig.11
40
Nuclear reactions (Inelastic)
Fig.11
?
Indirect ionization
41
Mass Distribution of Fragments
Fig.12

• Experiment to measure the distribution of mass of
  fragments that occur for a large number of proton
  events
• 173 MeV protons
• Target was aluminum
• Differential production cross section (N1 / ?)
• N1 - Number of times a specific mass was observed
• ? - Fluence protons/cm2

K. Kwiatkowski, et al., Phys. Rev. Letters, Vol.
50, No. 21, 1983.
42
Mass, Energy and Angle Distribution
Fig.12

• Same Experimental conditions
• 173 MeV protons
• Target was aluminum
• Differential production cross section (N2 / ?)
• N2 - Number of times a specific fragment with a
  certain energy and angle was observed
• ? - Fluence ( protons/cm2)

Oxygen
K. Kwiatkowski, et al., Phys. Rev. Letters, Vol.
50, No. 21, 1983.
43
63 MeV Proton
Experimental Results for Protons
Fig.13

• Proton-induced soft-error response may depend on
  the angle of incidence of the proton
• Attributed to the forward directed nature of
  recoiling nuclei
• Experimental soft-error data show that
  silicon-on-insulator (SOI)
• Soft-error cross-section (N3 / ?)
• Number of SEUs (N3)
• Fluence ( protons/cm2)

Soft Error Response (Cross Section - cm2)
Reed, et al. TNS 2000
44
Geometry and Trajectory Matters!

• Silicon on Insulator Device
• Consider sensitive volume for SOI to be a simple,
  high aspect ratio box

SiO2
Silicon Substrate
Reed, et al. TNS 2000
45
Ion-Ion Interactions
Fig.14
http//th.physik.uni-frankfurt.de/weber/Movies/in
dex.html
46
Complex Material Systems
Fig.15
Experimental evidence for this effect on a CMOS
SRAM
http//images.dailytech.com/nimage/4621_21476.jpg

• Ion-Ion nuclear reactions in non-silicon material
  near the sensitive volume contribute to the soft
  error response
• To date, this is only important for radiation
  harden circuits

Warren et al. 2005, Dodd et al., TNS 2007, Reed
et al. TNS 2007
47
Important Interactions of Radiation with Matter

• Primary radiation types responsible for Single
  Event Upsets
• Protons, Neutrons, and Heavier Ions
• Two Classes of Interaction Mechanisms
• Ion Stopping (Stopping Force)
• Electronic stopping
• Coulomb recoil of target atom (nuclear
  stopping)
• Electromagnetic radiation
• Projectile excitation and ionization
• Electron capture
• Chemical reactions
• Nuclear Reaction (Nuclear Forces)
• Inelastic
• Fission
• Elastic

48
Nuclear Fission
Fig.16
Thermal Neutron
49
Nuclear Fission
Fig.16
50
Nuclear Fission
Fig.16
Indirect ionization
51
A Word on Scale

• Diameter of a nucleus is few Fermi (10-15
  meter)
• Diameter of an atom is few angstroms (10-10
  meter)
• Atomic spacing in Silicon 2.3 angstroms

52
A Word on Scale

• Diameter of a nucleus is few Fermi (10-15
  meter)
• Diameter of an atom is few angstroms (10-10
  meter)
• Atomic spacing in Silicon 2.3 angstroms

Distance to the next golf ball 10 km
Diameter of Nucleus size of a golf ball
53
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Space Radiation Environment
• Terrestrial Radiation Environment
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Collection

54
Space Radiation Environment
Fig.17
Galactic Cosmic Rays
Nikkei Science, Inc. of Japan, image by K. Endo
55
Space Radiation Environment
Fig.17
Galactic Cosmic Rays
Nikkei Science, Inc. of Japan, image by K. Endo
56
Galactic Cosmic Rays
Fig.18
https//creme96.nrl.navy.mil/
57
Galactic Cosmic Rays
Fig.18
https//creme96.nrl.navy.mil/
58
Iron Dominates the LET Environment
Fig.19
https//creme96.nrl.navy.mil/
59
Iron Dominates the LET Environment
Fig.19
Range lt 80 ?m
Range lt 45 ?m
https//creme96.nrl.navy.mil/
60
Space Radiation Environment
Fig.17
Galactic Cosmic Rays
Nikkei Science, Inc. of Japan, image by K. Endo
61
Sunspot Cycle with Solar Proton Events
Fig.20
Proton Event Fluence
Protons (/cm2)
Sunspot Number
Year
Sunspot cool planet-sized areas on the Sun where
intense magnetic loops poke through the star's
visible surface
J. Barth, Notes from 1997 IEEE NSREC Short
Course
62
Solar Event-Integrated Spectra
Fig.21

• Solar spectra varies from one event to the next
• Certain events contain increased heavy ion (Zgt2)
  component
• Coronal Mass Ejection
• Dominated by protons

R.A. Mewaldt, et al., 29th International Cosmic
Ray Conference
63
Space Radiation Environment
Fig.17
Figure 2.1
Nikkei Science, Inc. of Japan, image by K. Endo
64
Van Allen Proton and Electron Belts
Fig.22
Inner Belt - Protons Electrons
Outer Belt - Electrons - Protons
(Low Energy)
Slot Region
J. Barth, Notes from 1997 IEEE NSREC Short
Course
65
Van Allen Proton Belts
Fig.23
Protons (/cm2/s)
L-Shell
J. Barth, Notes from 1997 IEEE NSREC Short
Course
66
Important Interactions of Radiation with Matter

• Primary radiation types responsible for Single
  Event Upsets
• Protons, Neutrons, and Heavier Ions
• Two Classes of Interaction Mechanisms
• Ion Stopping (Stopping Force)
• Electronic stopping
• Coulomb recoil of target atom (nuclear
  stopping)
• Electromagnetic radiation
• Projectile excitation and ionization
• Electron capture
• Chemical reactions
• Nuclear Reaction (Nuclear Forces)
• Inelastic
• Fission
• Elastic

Space Environment
67
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Space Radiation Environment
• Terrestrial Radiation Environment
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Collection

68
Alpha Particles for Radioactive Decays

• Alpha Helium ion
• Energy lt 10 MeV
• Range in silicon lt 100 ?m
• Traces of alpha emitters in semiconductor process
  and packing material
• Purification production materials and shielding
  will mitigated alpha environment

69
Terrestrial Neutron Environment
Fig.24

• Neutron showers produced from interaction of GCR
  ions with the atmosphere
• Flux depends on strongly altitude
• Denver, CO has higher flux than Hawthorne, NY
• Neutrons interact with target nuclei
• Fission E lt ? 1 MeV
• Nuclear reaction E gt ? 1 MeV
• Similar concerns as proton induced nuclear
  reactions

Paul Goldhagen, private communication.
70
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Crystal Structure
• Energy Band Diagram
• Carrier Motion
• p-n junction
• Charge Generation
• Charge Collection

71
Crystal Structure and Energy Bands
Fig.25

• Crystal structure imposes a periodic force on
  free electrons
• Periodic potential

72
Crystal Structure and Energy Bands
Fig.25

• Crystal structure imposes a periodic force on
  free electrons
• Periodic potential
• The quantum nature of electrons and the periodic
  crystal potential cause the electrons to be in
  well defined energy states
• Each energy band has a finite number of electrons

Conduction Band
EC
EG EC - EV
EV
Valence Band
73
Crystal Structure and Energy Bands
Fig.25

• Crystal structure imposes a periodic force on
  free electrons
• Periodic potential
• The quantum nature of electrons and the periodic
  crystal potential cause the electrons to be in
  well defined energy states
• Each energy band has a finite number of electrons
• At T 0 K no electrons in the conduction band
• Above T 0 K electrons and holes are generated

Conduction Band
EG EC - EV
Valence Band
74
Fermi Level
Fig.25

• Fermi level (EF) is the energy level that has 50
  probability of being occupied by an electron
• EF Ei ? EV EG/2 for an intrinsic
  semiconductor

Conduction Band
Intrinsic semiconductor
EF Ei
Valence Band
75
Determining Carrier Density
Fig.26
Energy Band Diagram
Density of Available Energy States
Density of Carriers
Probability of Occupancy
intrinsic semiconductor
Courtesy of Alan Doolittle
76
Determining Carrier Density
Fig.26
Energy Band Diagram
Density of Available Energy States
Density of Carriers
Probability of Occupancy
n-type semiconductor
Courtesy of Alan Doolittle
77
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Crystal Structure
• Energy Band Diagram
• Carrier Motion
• p-n junction
• Charge Generation
• Charge Collection

78
Carrier Motion (Drift)
Fig.27
Thermal Motion
3
4
1
2
79
Carrier Motion (Drift)
Fig.27
Thermal Motion
Thermal MotionDrift
2
3
4
3
4
1
2
1
Electrostatic potential gradient
For an ensemble of particles J - Current Density
(amps/cm2) C1 - Constant
80
Carrier Motion (Diffusion)
Fig.28
Diffusion
Carrier gradient
81
Carrier Motion (Diffusion)
Fig.28
Diffusion
Diffusion
Carrier gradient
For an ensemble of particles J - Current Density
(amps/cm2) C2 - Constant n - Carrier Density
82
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Crystal Structure
• Energy Band Diagram
• Carrier Motion
• p-n junction
• Charge Generation
• Charge Collection

83
p-n junction
Fig.29
Conduction Band
Conduction Band
p-type
EF
n-type
EF
Valence Band
Valence Band
84
Biasing a p-n Junction
Fig.29
85
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Ionization Energy
• Track Structure
• Charge Collection

86
Electronic Ionization of the Target

• Electronic ionization equates to generation of
  electron-hole pairs (or e-h pairs)

Conduction Band
Valence Band
87
Radiation Ionization Energy (Ee-h)
Fig.31

• Ee-h is the mean energy expended per
  electron-hole (e-h) pair generated in a given
  material by an ionizing radiation
• Experimental determined values of Ee-h for
  various crystalline semiconductors
• For Si the Ee-h(Si)
• 3.6 eV/e-h pair
• Bandgap energy 1.12 eV
• Average number charge generated (Qave) can be
  determined from the average energy loss from
  electronic stopping

Klein - J.A.P. 39 2029(1968)
88
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Ionization Energy
• Track Structure
• Charge Collection

89
Ion Track Structure - e-h pair cloud size

• Key issues when discussing the size of a
  radiation event and its effects on the soft-error
  response
• Initial 3D distribution of e-h cloud from a
  single charge particle,
• Limited experiment work
• Majority of the work has been theoretical (2
  techniques)
• Analytical models
• Monte-Carlo calculation of ion transport
• Variability of ion e-h cloud size produced by
  otherwise identical particles,
• Redistribution of the initial spatial
  distribution of carriers via thermal relaxation
  with the energy band structure,
• Transport of e-h pairs after the radiation event
  and thermalization has occurred, and
• Geometry of the components of the technology,
  e.g., active region versus insulator.

90
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Transport and Collection
• Fundamentals
• Selected Charge Collection Processes for Bipolar
  Transistor
• Small, Process-Isolated Diode
• Selected Charge Collection Processes for
  Sub-Micro CMOS Technologies

91
Charge Collection Large, Spatially-Isolated Diode
Fig.29

• Current induced on a contact by the movement (via
  drift and/or diffusion) of ion-induced free
  carriers within the semiconductor

• Current profile is determine by
• Type of carriers (electron or hole) and their
  motion (drift or diffusion)
• Impedance tied to the contact
• Electrostatic potential, in time, on the contact,
• Typically, events on a reversed bias p-n junction
  it the primary event type of concern for
  soft-errors

• Technology Computer Aided Design (TCAD) used by
  several research groups to study charge
  collection processes

C. M. Hsieh IEEE Trans. Electron. Devices, Dec.
1983. C. M. Hsieh, Proc. IEEE Int. Reliability
Phys. Symp., 1981
92
Bipolar Transistor
Fig.40

• SiGe Heterojunction Bipolar Transistor (HBT)
• SEE-critical features
• Deep trench isolation
• Subcollector junction
• Substrate doping
• 5 x 1015 cm-3

Courtesy of International Business Machines
Corporation. Unauthorized use not permitted.
93
Ionizing Radiation Event in an HBT
Figs. 2 and 29

• Ionizing event crossing a reversed bias p-n
  junction

Free carriers
Electric Field
94
Ionizing Radiation Event in an HBT
Fig.2

• Ionizing event crossing a reversed bias p-n
  junction
• Generates e-h pairs
• Potential and electric field modulation can occur
• Carriers are separated by the field and move by
  diffusion, inducing a current on the contact
• Potential drop returns to pre-event condition
• charge is collect across the diode by diffusing
  to the junction

Courtesy Jonathan Pellish
95
Angle Dependence for HBT

• Potential modulation is truncated by the deep
  trench

Courtesy Jonathan Pellish
96
Carrier Diffusion and Charge Collection
Fig.41
0.005 pC
0.5 pC
Courtesy Jonathan Pellish
97
Carrier Diffusion and Charge Collection
Fig.41
Substrate p-type 5 x 1015 cm-3
0.005 pC
0.5 pC
Courtesy Jonathan Pellish
98
Carrier Diffusion and Charge Collection
Fig.41
Substrate p-type 5 x 1015 cm-3
0.005 pC
0.5 pC
Courtesy Jonathan Pellish
99
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Transport and Collection
• Large, Spatially-Isolated Diode
• Small, Process-Isolated Diode
• Selected Charge Collection Processes for Bipolar
  Transistor
• Selected Charge Collection Processes for
  Sub-Micro CMOS Technologies

100
Equilibrium and Non-Equilibrium Condition

• Carrier equilibrium is defined as the careful
  self-balance of each fundamental process with its
  inverse process
• Non-equilibrium occurs when the self-balance does
  not exist
• Quasi-Fermi levels used to define carrier
  characteristics
• Concentration
• Total Current

Conduction Band
p-type Equilibrium
Ei
EF
Valence Band
Conduction Band
EFN
p-type Non-equilibrium
Ei
EFP
Valence Band
101
Small, Process-Isolated Diode
Fig.37

• Small n region on a p substrate
• Shallow trench isolation (STI)

Courtesy Sandeepan DasGupta
102
Energy Band and Potential During a Radiation Event

• Band diagrams for the region surrounding diode
  (0.5 ?m)
• The equilibrium Fermi level is shown in panel A
• The quasi-Fermi levels are show in panels B and C
• Just after event (1ps)
• Potential is modulated, high level of injection,
  non-equilbrium
• Quasi-Fermi level is sloped and carrier density
  is high
• After event (10 and 250 ps)
• Potential modulation deep into substrate
• Quasi-Fermi level nealy flat and carrier density
  is high

Equilb.
1ps
10 to 250 ps
103
TCAD Simulation Results
Fig.38 39
Potential Modulation Small, Process-Isolated
Diode

• Potential modulation controlled by process
  features
• Carrier density defines extent of modulation
• Carrier motion has implication on e-h cloud size

Before Event
gt10 ps lt 250 ps
1ps
104
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Collection
• Fundamentals
• Selected Charge Collection Processes for Bipolar
  Transistor
• Selected Charge Collection Processes for
  Sub-Micro CMOS Technologies

105
CMOS - Technology Trends
Fig.43

• Transistors Scale, ion-induced e-h pair cloud
  doesnt!
• For large scaled devices, e.g. 1 µm, the charge
  cloud created by the SE confined to drain region.
  
• For highly scaled devices, e.g., 90 nm, e-h pair
  cloud affects entire device plus well contact,
  plus other proximate devices.

S. DasGupta TNS 2007
106
CMOS - Charge Sharing

• A single ion event can induce current on more
  than one contact
• This example is for CMOS, other technologies will
  have similar effects, e.g., blooming in imagers

O. A. Amusan, TNS 2006 2007 - TCAD Simulation
Courtesy John Hutston
107
CMOS - Bipolar Effects

• Bipolar effects may exist, e.g.,
  diffusion-well-substrate
• Carrier transport after an ion event in the well
• e-h pairs modulate the souce/well potential
• The source/well/drain becomes a bipolar
  transistor
• The electrons emitted by the source will
  contribute to the total charge collected by the
  drain.

Courtesy of Paul Dodd
108
Outline

• Introduction
• Interaction of Radiation with Matter
• Radiation Sources
• Fundamental Semiconductor Physics
• Charge Generation
• Charge Collection
• Summary !

109
Summary (I)

• The basic physical mechanisms for SEEs are
• 1) ionizing radiation-induced energy deposition
• 2) initial electron-hole pair generation
• 3) carrier transport within the semiconductor,
  and
• 4) the response of the device and circuit
• The primary particles of interest when studying
  soft-errors are neutrons, protons and heavier
  ions.
• Important energy loss mechanisms
• stopping and nuclear (strong and weak).
• stopping force is dominated by ionization of the
  target
• direct ionization or indirect ionization.

110
Summary (II)

• Average LET calculated using computer codes
• agree with experimental data for the most part
  for LETs above the Bragg peak
• significant difference can occur for energies
  near or below the Bragg peak
• Fluctuations in LET (energy-loss straggling)
  determine the penetration depth (range) and its
  fluctuation (range straggling)
• Ion LET determines the average energy deposited
• Extreme events often determine the onset of
  radiation effect
• Implication of energy deposited by the
  fragmentation of nuclei as a result of nuclear
  reactions
• Angular dependence for high aspect ration SVs
• Soft-errors induced by incident heavy ions via
  nuclear reactions

111
Summary (III)

• Key factors that determine the e-h cloud size
  are
• Radial distribution of e-h pairs from a single
  charge particle
• Variability of this e-h pair distribution from
  particle to particle
• Transport of e-h pairs after the radiation event
  has occurred
• Simultaneous e-h pair production from primary and
  secondary particles
• Geometry of the components of the technology
• Modeling, simulation and experimental methods
  have been used to study charge collection for
  several years
• Drift, funneling, and diffusion models can
  describe the charge collection process for a
  large, spatially-isolated diode.
• For example, appropriate for gt 1 ?m technology
  node

112
Summary (VI)

• Charge collection processes for a small,
  process-isolated diode is much different than
  that for a large diode
• Structures defined by the process impact
  potential modulation around the stuck device.
• Region immediately surrounding the small diode is
  affected by the transport of carriers in the
  ion-induced e-h cloud.
• Basic fundamental mechanisms for soft-errors are
  easily understood
• Several advanced technology dependent issues show
  that the manifestation of those fundamental
  mechanisms from technology to technology are very
  complex.
• Primary mechanisms for large scale (gt 250 nm)
  CMOS is much different that small scale CMOS.
• Bipolar effects can be exhibited in bulk CMOS and
  in SOI.
• Devices manufactured on lightly doped substrates
  can collect charge deposited several micrometers
  deep into the substrate

113
Advanced Radiation Effects Analysis
Automated Connection Between New Models
Ground Based Experiments
Orbit
Space Radiation Environment Models
Space Radiation Transport Models

• Device Single Event Effects

• Circuit Single Event Effects

Modeled Response
System Single Event Effects
114
Advanced Simulation of a Radiation Event

• 63 MeV proton incident on a SiGe HBT
• Iso-charge surfaces following a nuclear reaction

115
Advanced Simulation of a Radiation Event

• 63 MeV proton incident on a SiGe HBT
• Iso-charge surfaces following a nuclear reaction

116
Advanced Simulation of a Radiation Event

• 63 MeV proton incident on a SiGe HBT
• Iso-charge surfaces following a nuclear reaction

117
Advanced Simulation of a Radiation Event

• 63 MeV proton incident on a SiGe HBT
• Iso-charge surfaces following a nuclear reaction

118
Advanced Simulation of a Radiation Event

• 63 MeV proton incident on a SiGe HBT
• Iso-charge surfaces following a nuclear reaction

119
Acknowledgements

• Friends, colleagues, and students at Vanderbilt
  University
• Profs. Robert Weller, Ronald Schrimpf, Marcus
  Mendenhall, Lloyd Massengill, Arthur Witulski,
• Kevin Warren, Jeff Black, Brian Sierawski, and
  Dennis Ball
• Jonathan Pellish, Christina Howe, Sandeepan
  DasGupta, and Nicholas Pate
• Sponsors
• NASA Electronics Parts and Packaging Program
• Defense Threat Reduction Agency Radiation
  Hardened Microelectronics Program