MC Implant 1D/2D Monte Carlo Implantation Simulator Seamlessly - - PowerPoint PPT Presentation

mc implant 1d 2d monte carlo implantation simulator
SMART_READER_LITE
LIVE PREVIEW

MC Implant 1D/2D Monte Carlo Implantation Simulator Seamlessly - - PowerPoint PPT Presentation

MC Implant 1D/2D Monte Carlo Implantation Simulator Seamlessly Integrated into ATHENAs SSuprem4 Summary MC Implant is an advanced 2-D ion-implantation physics based simulator for modeling of ion stopping and implant ranges in


slide-1
SLIDE 1

MC Implant 1D/2D Monte Carlo Implantation Simulator

Seamlessly Integrated into ATHENA’s SSuprem4

slide-2
SLIDE 2

MC Implant 1D/2D Monte Carlo Implantation Simulator

Summary

MC Implant is an advanced 2-D ion-implantation physics based

simulator for modeling of ion stopping and implant ranges in amorphous and crystalline media

Comparisons to measured data have shown that MC Implant is

still accurate and predictive even below 1keV

With the ATHENA framework, MC Implant provides seamless bi-

directional integration with SSuprem4 and Flash simulators allowing modeling of the implantation process for all available impurity/target material combinations and for any arbitrary geometry

  • 2 -
slide-3
SLIDE 3

MC Implant 1D/2D Monte Carlo Implantation Simulator

Key Benefits

Easy to use, self writing (menu driven) input files Predictive and experimentally verified results Fully integrated into Silvaco’s industry leading visualization tool Fully inter-active run time environment History file creation at every step allows real time modifications to

input files

Continuous, in house, customer driven development Fully integrated into Silvaco’s process simulation tool, SSuprem4

in ATHENA

  • 3 -
slide-4
SLIDE 4

MC Implant 1D/2D Monte Carlo Implantation Simulator

Applications

Process optimization for performance enhancement Analysis of material/topology effects on implant profile Implant damage and subsequent TED investigations Failure analysis Process robustness, manufacturability and yield analysis Investigation of mask (cost) reduction viability Novel devices Patent proposals and legal defense thereof

  • 4 -
slide-5
SLIDE 5

MC Implant 1D/2D Monte Carlo Implantation Simulator

Usage - When to use a Monte-Carlo Simulator

Unusual surface topology - eg shadowing effects Investigation of unusual phenomenon Implantation energy extremes - low or high When beam divergence is important - highly channeled Investigation of dominant channeling directions When amorphization effects are important Multi-layer material structures Whenever accuracy is more important than simulation time

  • 5 -
slide-6
SLIDE 6

MC Implant 1D/2D Monte Carlo Implantation Simulator

Contents

Topology effects - shadows, reflections and re-implantation Non intuitive phenomenon Calibration - how good is Silvaco’s simulator? 3D simulations Damage profiles and amorphization Increasing the simulation speed Some basic physics Concluding

  • 6 -
slide-7
SLIDE 7

MC Implant 1D/2D Monte Carlo Implantation Simulator

Unusual Topology - Shadowing, Reflection & Re-implantation

10keV, 1e14/cm2 Boron

implanted at 7 degrees

Shows shadowing Ions initially reflected from

the right hand side wall are re-implanted into the left hand wall.

Shows expected increase

in re-implanted dose with depth

Implanted dose is reduced

for high implant angles on the trench wall (CosØ effect)

  • 7 -
slide-8
SLIDE 8

MC Implant 1D/2D Monte Carlo Implantation Simulator

Unusual Topology - Shadowing, Reflection & Re-implantation

The same implant using

ATHENA’s SIMS verified look-up tables (SSuprem4)

Correctly shows shadowing

and CosØ effect in right hand trench wall

However, reflection and re-

implantation effects are absent from the look-up table approach

Monte-Carlo required for

correct simulation of dopant distribution in this case

  • 8 -
slide-9
SLIDE 9

MC Implant 1D/2D Monte Carlo Implantation Simulator

Unusual Effects - 2 Counter Intuitive Examples

Increasing surface screen

  • xide thickness increases ion

channeling - why ?

Monte Carlo Simulation

provides the answer - The dominant channeling direction is not vertical. The thicker

  • xide increases the probability
  • f ion scattering in this

preferred “off-axis” direction

  • 9 -
slide-10
SLIDE 10

MC Implant 1D/2D Monte Carlo Implantation Simulator

Unusual Effects - 2 Counter Intuitive Examples

Crystalline structure of

silicon in various crystal directions

Note the more open

structure in the preferred channeling <011> direction giving rise to the effect on the previous slide

  • 10 -
slide-11
SLIDE 11

MC Implant 1D/2D Monte Carlo Implantation Simulator

Unusual Effects - 2 Counter Intuitive Examples

An implanted ion’s eye view

when channeling

Simply a beautiful graphic

showing how ions become channeled for certain crystal directions

  • 11 -

Reference: http://www.ornl.gov/ORNLReview/v34_2_01/fermi.htm

In 1962 ion channeling was discovered

  • n a computer at ORNL.
slide-12
SLIDE 12

MC Implant 1D/2D Monte Carlo Implantation Simulator

Unusual Effects - 2 Counter Intuitive Examples

Significant lateral spread under

the gate

Lateral spread in crystalline

materials is much higher than in amorphous materials and is due to ion channeling in <011>, <110> and equivalent crystallographic directions

The implant direction is <001>

but because of dechanneling, ions enter the other planes

This ion redirection is increased

in the presence of oxide layers and with high dose implants

The resultant lateral spread will

affect device performance target such as puchthrough voltage

  • 12 -
slide-13
SLIDE 13

MC Implant 1D/2D Monte Carlo Implantation Simulator

Unusual Effects - 2 Counter Intuitive Examples

Implantation of two

similar mass ions, aluminum and phosphorus, results in greatly differing ion channeling depths by a factor of 3 - Why ?

  • 13 -

Random and (110) channeled atom depth distributions for Al and P implanted into crystalline Si at 200 keV and 3 x 1013 cm-2

Reference: R.G.Wilson, J.Appl.Phys., Vol.60, pp2797-2805, 1986

slide-14
SLIDE 14

MC Implant 1D/2D Monte Carlo Implantation Simulator

Unusual Effects - 2 Counter Intuitive Examples

Answer:- The electronic

stopping cross section for the two ions in the <110> direction is greatly different

The graphic shows

electronic stopping in the <110> direction for ions

  • f various atomic number

Al = 13 P = 15

  • 14 -

Reference: Ivan Chakarov “Atomic and Ion Collisions in Solids” Ed. R.Smith, Cambridge University Press, 1997

Electronic stopping for low energies Stereographic and schematic view of the (100) channel in silicon. The Zi-depence of the electronic cross-section (v = 1.5 x 108 cms-1 for the ){110} direction in crystalline silion. Experimental points are from Eisen (1968.

slide-15
SLIDE 15

MC Implant 1D/2D Monte Carlo Implantation Simulator

Calibration - Measured versus Simulated Data

  • 15 -

Very low energy 0.5keV Boron - 1e12/cm2 1keV Arsenic - 1e12/cm3 Both zero degree tilt

slide-16
SLIDE 16

MC Implant 1D/2D Monte Carlo Implantation Simulator

Calibration - Measured versus Simulated Data (con’t)

  • 16 -

Low energy 2 keV Boron - 1e12/cm2 Left - Zero Tilt Right - 7 Tilt, 45 Rotation

slide-17
SLIDE 17

MC Implant 1D/2D Monte Carlo Implantation Simulator

Calibration - Measured versus Simulated Data (con’t)

  • 17 -

Low energy 2 keV Arsenic - 1e12/cm2 Left - Zero Tilt Right - 7 Tilt, 45 Rotation

slide-18
SLIDE 18

MC Implant 1D/2D Monte Carlo Implantation Simulator

Calibration - Measured versus Simulated Data (con’t)

  • 18 -

Medium energy 15 keV Boron - 1e13/cm2 80 keV Boron - 1e13/cm2 Both - 7 Tilt, 30 Rotation

slide-19
SLIDE 19

MC Implant 1D/2D Monte Carlo Implantation Simulator

The Monte-Carlo Simulator is Actually 3D

Boron concentration

contributions from various channeling angles

  • 19 -
slide-20
SLIDE 20

MC Implant 1D/2D Monte Carlo Implantation Simulator

The Monte-Carlo Simulator is Actually 3D (con’t)

The same channeling contributions viewed from the surface

  • 20 -
slide-21
SLIDE 21

MC Implant 1D/2D Monte Carlo Implantation Simulator

The Monte-Carlo Simulator is Actually 3D (con’t)

3D boron point implant simulation integrated into 2D Also compared to other work fundamental atomistic calculations

  • 21 -

G.Hobler, G.Betz (Inst. F.Allg.Physik, TU Wien)

slide-22
SLIDE 22

MC Implant 1D/2D Monte Carlo Implantation Simulator

Damage Profiles and Amorphization - Dose Effects

3 keV Arsenic Doses 1e13, 1e14, 1e15

and 1e16/cm2

Note increasing depth of

amorphization but damage tail remains fairly constant after onset of amorphization

  • 22 -
slide-23
SLIDE 23

MC Implant 1D/2D Monte Carlo Implantation Simulator

Damage Profiles and Amorphization - Dose Effects

  • 23 -

100keV Phosphorus, (0 tilt) Measured (diamonds) and

simulated data over-layed

Doses of 1e13, 5e13, 2e14,

5e14 and 1e15/cm2

R.J. Schreutelkamp et al., “Channeling Implantation

  • f B and P in Silicon”, Nuclear Instruments and

Methods, B55, pp. 615–619, 1991

slide-24
SLIDE 24

MC Implant 1D/2D Monte Carlo Implantation Simulator

Damage Profiles and Amorphization

  • 24 -

3keV Phosphorus Dose 1e13/cm3 Showing phosphorus (red),

interstitials (green), vacancies (dark blue) and net defects (light blue)

slide-25
SLIDE 25

MC Implant 1D/2D Monte Carlo Implantation Simulator

Using Statistics to Accelerate Simulation Times

  • 25 -

100keV Phosphorus, (0 tilt) Measured (diamonds) and

simulated data over-layed

Doses of 1e13, 5e13, 2e14,

5e14 and 1e15/cm2

Statistics increase chances of

rare event occurrences improving tail profiles for a given number of simulated ions

slide-26
SLIDE 26

MC Implant 1D/2D Monte Carlo Implantation Simulator

Basic Physics (con’t)

  • 26 -

ion lattice nuclear collisions e- e- e-

For predictive modeling of II we need to have

physically realistic treatment of: Nuclear stopping & interatomic potentials Local & non-local electronic stopping Damage buildup & amorphization

slide-27
SLIDE 27

MC Implant 1D/2D Monte Carlo Implantation Simulator

Basic Physics (con’t)

  • 27 -

BCA calculates paths Po, Pa

and Pf for incoming ion and To, Ta and Tf for impacted ion

It is not necessary to calculate

the exact curved path, just the initial and final ion path directions and distance. This is the approximation in the method

  • M. T. Robinson and I. M. Torrens, Phys. Rev., B9 (1974) p.5008
slide-28
SLIDE 28

MC Implant 1D/2D Monte Carlo Implantation Simulator

Basic Physics (con’t)

  • 28 -

1 2 3 4 5 6 7 8 9 10 1 10 100 1000 10000 100000 Boron energy, keV Nuclear, eV/A 20 40 60 80 100 120 Electronic, eV/A Nuclear Electronic

Nuclear and

Electronic stopping for boron in amorphous silicon up to 100MeV (Mega electron volts)

slide-29
SLIDE 29

MC Implant 1D/2D Monte Carlo Implantation Simulator

Conclusions – General Capabilities

Highly accurate implanted ion profiles verified by SIMS Accurate for any surface topology Surface reflection, shadowing, re-implantation and free flight

through voids all correctly calculated

Any multi-layer combinations of materials known to Athena Crystalline/amorphous material combinations correctly calculated Correctly calculates profiles for a very wide range of ion energies

from fractional keV up to the MeV range

Damage accumulation and amorphization effects (dose

dependency) correctly calculated

Statistical options for simulation speed up

  • 29 -
slide-30
SLIDE 30

MC Implant 1D/2D Monte Carlo Implantation Simulator

Conclusion – De-channeling

Accurate calculation of de-channeling effects caused by:

  • 1. Damage buildup and previous implant damage
  • 2. Surface oxides polysilicon and other materials
  • 3. Beam divergence variations
  • 4. Implant angle and energy
  • 5. Amorphous material in the structure

3-D Channeling effects included in the generic solution of ion

propagation and stopping

  • 30 -
slide-31
SLIDE 31

MC Implant 1D/2D Monte Carlo Implantation Simulator

Conclusions – Process Applications

The comprehensive capabilities of MC Implant enable accurate

simulation of critical process issues such as shallow junction implants, multiple implants and pre-amorphization, HALO implants, retrograde well formation by high energy implants

Advanced damage accumulation algorithms and seamless

integration with SSuprem4 allow the application of novel diffusion models to implanted species

  • 31 -