1
AP 5301/8301 Instrumental Methods of Analysis and Laboratory Lecture 11 Ion Beam Analysis
Prof YU Kin Man E-mail: [email protected] Tel: 3442-7813 Office: P6422 2 Lecture 8: outline . Introduction o Ion Beam Analysis: an overview . Rutherford backscattering spectrometry (RBS) o Introduction-history o Basic concepts of RBS – Kinematic factor (K) – Scattering cross-section – Depth scale o Quantitative thin film analysis o RBS data analysis . Other IBA techniques o Hydrogen Forward Scattering o Particle induced x-ray emission (PIXE) . Ion channeling o Minimum yield and critical angle o Dechanneling by defects o Impurity location 3 Analytical Techniques
100at%
10at% AES XRD IBA
1at%
0.1at% RBS XRF 100ppm TOF- 10ppm SIMS
1ppm TXRF
100ppb
Detection Range Detection 10ppb Chemical bonding DynamicS 1ppb Elemental information IMS Imaging information 100ppt Thickness and density information
10ppt
Analytical Spot size But IBA is typically quantitative and depth sensitive http://www.eaglabs.com 4 Ion Beam Analysis: an overview
Inelastic Elastic Incident Ions Rutherford backscattering (RBS), resonant scattering, Nuclear Reaction Analysis channeling (NRA) p, a, n, g
Particle induced x-ray emission (PIXE)
Defects generation Elastic recoil detection
Ion Beam Absorber foil Modification 5 Ion Beam Analysis
1977 2009 2014
1978 6 Why IBA?
. Simple in principle . Fast and direct . Quantitative (without standard for RBS) . Depth profiling without chemical or physical sectioning . Non-destructive . Wide range of elemental coverage . No special specimen preparation required . Can be applied to crystalline or amorphous materials . Simultaneous analysis with various ion beam techniques (RBS, PIXE, NRA, channeling, etc.) . Can obtain a lot of information in one measurement 7 A typical Ion Beam Analysis Facility
Experimental chamber
bending focusing Accelerator Ion source Beam Fixed particle PIXE collimators detector detector
f
Moveable particle detector 8 IBA Equipment: accelerator
Van de Graaff (single ended) Tandem Pelletron 9 IBA Equipment: tandem pelletron 10 Equipment: particle detector 11
Rutherford Backscattering Spectrometry (RBS) . Relatively simple physics . Requires some experimental skills . Requires some data simulation to obtain the most out of the spectrum . The only quantitative and absolute method to determine ─ atomic concentration with no matrix effect ─ composition variation with depth ─ layer thickness ─ damage distribution and impurity lattice location of multi-elemental, multi-layered films Rutherford Backscattering Spectrometry (RBS) a brief history The original experiment
Alpha particles at a thin gold foil experiment by Hans Geiger and Ernest Marsden in 1909.
1) Almost all of the alpha particles went through the gold foil 2) Some of the alpha particles were deflected only slightly, usually 2° or less. 3) A very, very few (1 in 8000 for platinum foil) alpha particles were turned through an angle of 90° or more. "We shall suppose that for distances less that 10-12 cm the central charge and also the charge on the alpha particle may be supposed to be concentrated at a point." (1911) 13 RBS: The Surveyor V experiment
Surveyor V, first of its spacecraft family to obtain information about the chemical nature of the Moon's surface, landed in Mare Tranquillitatis on September 11, 1967.
"Surveyor V carried an instrument to determine the principal chemical elements of the lunar-surface material," explained ANTHONY TURKEVICH, Enrico Fermi institute and Chemistry Department, University of Chicago. "After landing, upon command from Earth, the instrument was lowered by a nylon cord to the surface of the Moon …” 14 General applications of RBS
. Quantitative analysis of thin films . thickness, composition, uniformity in depth . solid state reactions . interdiffusion . Crystalline perfection of homo- and heteroepitaxial thin films . Quantitative measurements of impurities in substrates . Defect distribution in single-crystal samples . Surface atom relaxation in single crystals . Lattice location of impurities in single crystals 15 RBS: basic concepts
. Kinematic factor: elastic energy transfer from a projectile to a target atom can be calculated from collision kinematics mass determination . Scattering cross-section: the probability of the elastic collision between the projectile and target atoms can be calculated quantitative analysis of atomic composition . Energy Loss: inelastic energy loss of the projectile ions through the target perception of depth
These allow RBS analysis to give quantitative depth distribution of targets with different masses 16 Kinematic factor: K
Conservation of energy : Conservation of momentum:
111 2 22 m11 vm 12 vm 2 vcoscosf m11 vm 12 vm 2 v 222 m1 vm 12sinsin 2v f
2 2 2 2 E1 (m2 m1 sin ) m1 cos K K,, m21 m m2 E (m m ) o 2 1 Kinematic Factor
2 2 2 2 E (m m sin ) m cos K 1 2 1 1 m2 E (m m ) o 2 1 2 (m m ) K ( 180o ) 2 1 m2 (m2 m1) (m m ) K ( 90o ) 2 1 m2 (m2 m1)
When m2>>m1: K ( 180o ) ~ 1 m2 18 Element identification
2.5 MeV He ion with =170o
K(197Au) = 0.923 K(109Ag) = 0.864 K(107Ag) = 0.862 K(65Cu) = 0.783 K(63Cu) = 0.777 Mass Resolution, dm2
2 E (m m ) For 180o scattering: K 1 2 1 m2 E0 (m2 m1)
2 d E mm 1 d 21 E 2 0 mm12
For a fixed dE1/E0 (~0.01), heavier projectiles result in better mass resolution
However, dE1 for heavier projectiles is higher 20 Mass Resolution: examples
With system energy resolution dE = 20keV and E0=2MeV (40 4)3 20 For m =40 dm2 1.48a.m.u. 2 4 4(40 4) 2000 (70 4)3 20 For m = 70 dm 3.84a.m.u. 2 2 4 4(70 4) 2000 Isotopes of Ga (68.9 and 70.9 a.m.u.) cannot be resolved
4 MeV 3 MeV 2 MeV 1 MeV
We can also improve mass resolution by increasing Eo
3 mm d E d m 21 1 2 E 4m1 m 2 m 1 0 21 RBS Quantification
Backscattering Yield . For a given incident number of particles Q, a greater amount of an element present (N ) should result in Q s a greater number of particles scattered. . Thus we need to know how often scattering events should be detected (A) at a characteristic energy (E =
KE0) and angle , within our detector’s window of solid angle W. Scattering cross-section
If particles are coming in with impact parameters between b and b +db , they will be scattered through angles between and +d. For central forces, we have circular symmetry.
2(bdbd ) 2 sin Rutherford cross-section:
2 2 2 2 2 1/ 2 2 d Z1Z2e cos (1 A sin ) Z1Z2 m1 ~ A dW 2E 4 2 2 1/ 2 E m o sin (1 A sin ) o 2 2 23 Scattering Yield
1/ 2 2 2 22 Example: 2 cos1sin A Z12 Z e Yield,Y 2E 422 1/ 2 0 sin1sin A 2 Z1Z1 2 ( ) ~ 1024 cm 2[barn] E0
Mixed Si and W target analzed by a 2MeV He ion beam at 165o scattering angle.angle. (Si)=2.5x10-25cm2/str/str (W)=7.4x10-24cm2/str/str Total incident ions Q=1.5x1014 ions W=1.8mstr Area under Si, A(Si)=200 counts Area under W, A(W)=6000 counts 15 2 (Nt)(Nt)Si=3x10 atoms/cm 15 2 (Nt)(Nt)W=3x10 atoms/cm Energy Loss
Electronic stopping dE dx ele
Nuclear dE
Stopping dx nucl
When an He or H ion moves through matter, it loses energy through . interactions with e- by raising them to excited states or even ionizing them. . Direct ion-nuclei scattering Since the radii of atomic nuclei are so small, interactions with nuclei may be neglected dE dE dE dE dx total dx ele dx nucl dx ele 25 How to read a typical RBS spectrum
A thin film on a light substrate . Most projectile ions experience electronic stopping that results in a gradual reduction of the particle’s kinetic energy (dE/dx). . At the same time a small fraction of projectile ions come close enough to the nucleus for large- angle scattering (KE). . A detected backscattered particle has lost some energy during initial penetration, then lost a large fraction of its remaining energy during the large-angle scattering event, then lost more energy in leaving the solid. Depth Scale
E1 K(E0 Ein ) Eout dE dE t E K(E t) ( ) 1 0 dx dx cos E0 KE0
KEo Total energy loss E=KE0-E1
K dE 1 dE E ( ) t [S ] t cos dx cos dx o 1 E0 2 KE0
E1 KE E 0 [So] is the effective stopping power Stopping cross-section
1 dE eV cm3 eVcm2 : N dx cm atom atom
K 1 E N N t N[ ] t E in out o cos1 cos2 Depth scale: thin film
180o- Thickness of film: Eo KEo dE E dE dx KE 1 E (K 0 ) t [S ]t t dx cos o E0
E1 KE E E o t So N[ o ]
E obtained from TRIM code or empirical energy loss data fitting
E 28 Example: layer thickness
consider the Au markers
EAu=EAuF - EAuB
= [KAu dE/dxEo + 1 / (cos10º) • dE/dx EAuB ] • t
dE/dx 3MeV = NAl Al 3MeV = 6.02x1022 • 36.56x10-15 = 2.2x109eVcm-1
dE/dx EAuB = NAl Al 2.57MeV = 6.02x1022 x 39.34x10-15 = 2.37x109 eVcm-1 t = 3945Å
EAl = 165keV consider the Al signals EAl = [KAl dE/dxEo + 1 / (cos10º) • dE/dx KEo ] • t EAu = 175keV
KAu = 0.9225 t = 3937Å
KAl = 0.5525 29 Energy Loss: Bragg’s rule
For a target AmBn, the stopping cross- section is the sum of those of the constituent elements weighted by the abundance of the elements.
AmBn = m A + n B
Example: Al2O3 the stopping cross-section of Al2O3. Given: Al = 44 x 10-15eVcm2 O = 35 x 10-15eVcm2 Al2O3 = 2/5 x Al +3/5 x O = (2/5 x 44 + 3/5 x 35) x 10-15 = 38.6 x 10 -15 eV-cm2/atom Al2O3 22 -15 dE/dx(Al2O3)=N =(1.15x10 )(38.6x10 )eV/cm =44.4 eV/Å 30 Quantitative analysis: composition and thickness
AA=AWQ(Nt)A
AB=BWQ(Nt)B A (Nt) Z m A ( A ) ( A ) ( A )2 A (Nt) n B B B ZB m A Z ( A ) ( B )2 n AB Z A
(E) (E) t A B AmBn AmBn [So ]A [So ]B (E) (E) t A B AmBn AmBn N[o ]A N[o ]B 31 Example: As implanted Si
KAs=0.809; KSi=0.566 Si 15 2 [o]Si 92.6x10 eV cm / atom Si 15 2 [o]As 95.3x10 eV cm / atom
Si EAs 68keV (FWHM )As 60keV E Si R As 1420Å p Si N[o]As (FWHM ) R As 540Å p Si 2.355 N[o]As
Total As dose: AAs (Nt)As ( As W Q) 32 RBS Application: impurity profile
Ge inplanted region 1.95 MeV 4He+
Si SiO2
I. Sharp et al., LBNL (2004) 33 Thin film analysis
YxBayCuzO
y AY ZBa 2 ( ) y HY Z Ba 2 x ABa ZY ( ) x H Ba ZY 34 Example: silicide formation (1970-80s)
Before reaction After reaction
Wei-Kan Chu, James W. Mayer and Marc-A. Nicolet, Backscattering Spectrometry, (Academic Press, New York 1978). RBS application: silicide formation
Reaction kinetics:
K. N. Tu et al. Thin Solid film 25, 403 (1975).
K. N. Tu et al. Japn, J. Appl. Phys. Suppl. 2 Pt 1 669 (1974). J. O. Olowolafe et al. Thin Solid film 38, 143 (1976). 36 IBA Data Analysis
. Starts with a simple guess sample structure (film thickness and composition) . Simulates spectrum and directly compare to experimental data . Iterates simulated structure to fit experimental data (either automatically or manually) . An ideal software: – able to handle different data files – have large data base for various ion-solid interaction: resonant scattering, nuclear reaction, etc. – User friendly interface – Fast simulation – Can simulate sample non-uniformity (roughness, compositional gradient, etc.) – Able to correct for electronic errors (pile up, dead time, charge up, etc) 37 Thin Film Analysis: single layer
3000 GaN Bi on sapphire 1-x x
550 nm GaN0.9Bi0.1
2000 3.04 MeV a
Ga
1000
Bi
0 200 700 1200 1700 2200 Energy (keV) A.X. Levendar, LBNL 38 Thin film analysis: multilayers
8000 150ITO nm ITO CIGS solar cell structure i35- ZnO nm ZnO 3.05 MeV a 45 nm CdS 6000 CdS 2 mm Mo CIGSCuGa0.16In0.84Se2 4000
In O 1Mo mm Mo 2000 Se In Zn Cu Cd SGa Sn 0 0.6 1.0 1.4 1.8 2.2 2.6 glass Energy (MeV) 39 Impurity profile
Ge inplanted region 1.95 MeV 4He+
Si SiO2
I. Sharp et al., LBNL (2004) 40 Strengths of RBS
. Simple in principle . Fast and direct . Quantitative without standard . Depth profiling without chemical or physical sectioning . Non-destructive . Wide range of elemental coverage . No special specimen preparation required . Can be applied to crystalline or amorphous materials . Simultaneous analysis with various ion beam techniques (PIXE, PIGE, channeling, etc.) 41 Weaknesses of RBS
. Poor lateral resolution (~0.5-1mm) . Moderate depth resolution (>50Å) . No microstructural information . No phase identification . Poor mass resolution for target mass heavier than 70amu (PIXE) . Detection of light impurities more difficult (e.g. Li, B, C, O, etc) (non-Rutherford scattering, Nuclear Reaction Analysis) . Data may not be obvious: require knowledge of the technique (simulation software) 42
Hydrogen Forward Scattering (HFS) (Elastic Recoil Detection Analsyis ERDA) 43 Hydrogen Forward Scattering (HFS)
Generally known as elastic recoil detection analysis (ERDA) 44 Hydrogen Forward Scattering (HFS)
Generally known as Elastic Recoil Detection (ERD)
Charles Evans and Assoc., RBS APPLICATION SERIES NO. 3 . Quantitative hydrogen and deuterium profiling . Good sensitivity (~0.01at% of H) . Can be perform simultaneously with RBS and PIXE . Profiling with any light element in solid (using heavy ion beam, ERD) 45 HFS: a-SiN:H film
Energy (MeV) 0.4 0.6 0.8 1.0 1.2 25
20
d
l
e
RBS i
Y 15
d
e
z
i
l
a
m 10
r
o
N
5 177 nm SiN1.15H0.013
0 100 200 300 400 500 600 700 Channel Energy (MeV) 0.5 0.6 0.7 0.8 0.9 1.2
1.0 Si
d l 0.8
e
i
Y
d
HFS e z 0.6
i
l
a
m
r
o 0.4
N
0.2
0.0 Courtesy: F. Hellman group, 2008 200 250 300 350 400 450 500 Channel Particle Induced X-ray Emission (PIXE) 47 PIXE: Light impurity in heavy matrix
GaAs 2 MeV 4He+
Ga Mn As RBS 1-x x PIXE
K. M. Yu et al. 2002. 48 PIXE Application: Geology, Art, Archeology, Biology
External Beam 49 Ion channeling 50 Ion Channeling
Kobelco Steel Group Ion Channeling
random Planar channel Axial channel 52 Ion Channeling: minimum yield and critical angle
Two important parameters to characterize channeling results:
1. Minimum yield:
Ychanneled min Yrandom ~0.02-0.06
2. Critical half-angle, y1/2 indicates presence of defects responsible for beam dechanneling Ion Channeling: minimum yield and critical angle
Minimum Yield, min Critical half-angle, y1/2
1 2 1 2 1 2Z1Z2e Ca 2 Y c ( ) 2 ln[( ) 1] channeled 2 Ed min Yrandom where d is the distance of atoms in a row, a is the Thomas-Fermi screeing distance, r r 2 min ~ 0.02 0.05 is rms thermal vibration min 2 ro 2Z Z ~ ~ ( 1 2 )2 ~ 0.5 1o 1/ 2 c E 54 Dechanneling by defects
Small angle dechanneling- line defects
Perfect crystal channeling
Direct scattering- point defects Channeling: Impurity Lattice Location
3 0.0 r 0.2
ro 2 0.4 Atomic row
Flux 0.6 y/y1 0.8 1 =1.0
Channel center
0 0.2 0.6 1.0 Distance from atom row, r/ro 56 Experimental techniques : combined channeling RBS/PIXE
Rutherford backscattering (RBS) Mn
Particle-induced x-ray emission (PIXE)
c-RBS: crystalline quality of film (from GaAs backscattering yields) c-PIXE: substitutionality of Mn atoms w.r.t. host GaAs 57 Channeling: Ga1-x-yBeyMnxAs
Channeling RBS/PIXE: -presence of interstitial Mn in GaAs
-[MnI] increases with Be doping -Mn reduces T <100> I C
<110>
<111>
K. M. Yu et al. 2003. 58 Homo- and Heteroepitaxy 59 Channeling: Heteroepitaxy
500 nm
Z. Liliental-Weber ~530nm InN on 210 nm GaN
K. M. Yu et al., LBNL 2004. 60 Amorphous layer analysis 61 Strengths of Ion Beam Analysis Techniques
. Simple in principle . Fast and direct . Quantitative (without standard for RBS) . Depth profiling without chemical or physical sectioning . Non-destructive . Wide range of elemental coverage . No special specimen preparation required . Can be applied to crystalline or amorphous materials . Simultaneous analysis with various ion beam techniques (RBS, PIXE, channeling, etc.) 62 Pitfalls in IBA 풅흈 푬, … Yield: 풀 = 푵 ( ) ∗ 푸훀 풊 풊 풅훀
. Q-total charge ─ accurate charge integration . W-solid angle ─ d/dW- detection angle, double/plural scattering . Other issues: ─ Radiation damage ─ Sputtering ─ Detector response ─ Surface roughness ─ Non-uniformity ─ Charging on insulators ─ Count rate effects 63 Charge integration
. Charge Integration . Accurate charge integration is important for absolute quantitative measurements . Good faraday cup design 64 Deviation from Rutherford scattering
Correction factor F, which describes the deviation from pure Rutherford scattering due to electron screening + for He scattering from atoms, Z2, at variety of incident kinetic energies.
Electronic screening Rutherford Nuclear Interaction Energy . At very high energy and very low energy, scattering will deviate from the Rutherford type. . At low energy : screening of e- must be considered . At high energy : nuclear short range force will enhance the cross-section, the so-called “resonance scattering.” 65 Charging effect for insulating samples
Severely distort the RBS spectrum . Provide a supply of low-E electrons from a small, hot filament located nearby . Coating the surface with a very thin layer of conducting material 66 Target non-uniformity
. Surface roughness and interface roughness cannot be distinguished . Target non-uniformity will resemble diffusion
Campisano et al., 1978