NE201A: XPS Oct 2024

X-Ray Photoelectron Spectroscopy

Sushobhan Avasthi
NE201

INDIAN INSTITUTE OF SCIENCE

Measurement system

Oct 2024 S. Avasthi, NE201A 2

S. Avasthi 1
NE201A: XPS Oct 2024

Reference Level Issue

Electron out Electron in

K.E = hν – BEsample- φDetector

K.E = hν – BEsample- φAg
Vacuum level
Vacuum level
Detector Work function
Sample Work function (φAg) (φDetector)
Fermi level Fermi level

Detector
Binding Energy
(BEAg3p) Same potential, so
same Fermi level
XPS photon in
(hν) - -
Sample
A: No. But we do need to know the work-
Q: Do we need to know the function of the detector.
work-function of the sample? BEsample = hν – measured K.E.- φDetector
Oct 2024 S. Avasthi, NE201A 3

How to Measure Detector Work-Function?

• Option 1: Measure binding energy on a material with known work-function

– Practically this is tricky. Work-function of even noble metals drifts with
time/contamination
– Need UPS or a very reliable standard

• Option 2: Use adventitious C peak

– Easy to do. Common practice.
– More on this a little later

Oct 2024 S. Avasthi, NE201A 4

S. Avasthi 2
NE201A: XPS Oct 2024

Issue of Sample Charging

-
- -
Charged surface changes the
Flow of KE (and hence calculated BE) of
electrons + + + the photoelectrons
-
KE’=KE - 1
KE’=KE - 3
Conducting samples
KE’=KE - 2
- +1V +3V
- - +2V
Insulating sample surface
NO flow of
electrons + + +
Microscopically, the charging is
Positively charged surface
non-uniform
Insulating samples So peaks are broader

Oct 2024 S. Avasthi, NE201A 5

Charge Compensation: Neutralizer (E-Gun)

5 eV 5 eV 5 eV
Electrons from Electrons from Electrons from
neutralizer - neutralizer - neutralizer -

+5V +2V 0V

Charged sample Charged sample Neutral sample

Electrons from
5 eV 5 eV neutralizer are 5 eV
Electrons from Electrons from reflected
neutralizer - neutralizer - -

0V -2V -5 V

Neutral sample Charged sample Charged sample

Oct 2024 S. Avasthi, NE201A 6

S. Avasthi 3
NE201A: XPS Oct 2024

Effect of Charging & Neutralizer

Si 2p peak of SiO2 substrate

Without charge
compensation
(no neutralizer)

With charge compensation

(using neutralizer)

Oct 2024 S. Avasthi, NE201A

Courtesy Varadhu, MNCF 7

How to Detect Charging? -- Adventitious C

Adventitious C 1s
• XPS chambers typically have some
Without charge
compensation adventitious C contamination*.
(neutralizer) – From hydrocarbons in air, plastics, etc.
With charge – Usually at 284.7 ± .2 eV
compensation • This adventitious C is often used to detect
(using neutralizer)
charging.
– Is the C 1s peak shifted?

• Also used to check the calibration on the

electron detector.

*Barr & Seal, J. Vacuum Sci. Technol. A 13, 1239 (1995)

Oct 2024 S. Avasthi, NE201A 8

S. Avasthi 4
NE201A: XPS Oct 2024

Need for UHV: Mean Free Path & Incident Flux

• MFP: Average distance between molecular

collisions
6.5×10"#
𝜆!! 𝑖𝑛 𝑐𝑚 =
𝑃 (𝑖𝑛 𝑚𝐵𝑎𝑟)

• IF: Number of molecules impacting a

surface atom
𝐹!! 𝑖𝑛 𝑠 "$ = 2.9×10%𝑃 (𝑖𝑛 𝑚𝐵𝑎𝑟)
Pressure (mBar) Vacuum level MFP Incident flux Time/ML

103 $
Atmosphere 10-7 m 108 10-9 s
• Time to form monolayer
-6
(ML) ∝ & Too small
10 High vac. 102 m 10-1 1s

10-10 UHV 106 m 10-5 104 s 3 hours

Oct 2024 S. Avasthi, NE201A 9

Need for Ultra High Vacuum (UHV)

XPS by Winograd & Gaarenstroom

Cleaned Ni surface

Exposed to O2 at
10-6 Torr for 100s

Oct 2024 S. Avasthi, NE201A 10

10

S. Avasthi 5
NE201A: XPS Oct 2024

But Samples Always Come from Ambient!

Courtesy Varadhu, MNCF

Samples usually have contamination on the surface.

Usually carbon and water but could be more is surface is not well-prepared
Oct 2024 S. Avasthi, NE201A 11

11

Ion Gun

10x difference

• Argon ion gun is used to sputter the

sample surface
– Mechanical etching
– 500V to 5 kV
• Beam can be raster scanned
– Raster size should be 10x XPS spot size
From Kratos Ultra manual
Oct 2024 S. Avasthi, NE201A 12

12

S. Avasthi 6
NE201A: XPS Oct 2024

Ion Gun: Surface Cleaning

Courtesy Varadhu, MNCF

Samples usually have contamination on the surface.

Usually carbon and water but could be more is surface is not well-prepared
Oct 2024 S. Avasthi, NE201A 13

13

Depth profiling Study: ZnO Film on Si

Ion gun can also be used to do depth profiles of the samples

Survey C 1s Etching time

180 sec (Bulk of Sample)

150 sec

120 sec

90 sec
More
etching
60 sec

30 sec

0 sec (As deposited)

Oct 2024 S. Avasthi, NE201A Courtesy Sampada, MNCF 14

14

S. Avasthi 7
NE201A: XPS Oct 2024

Depth profiling Study: ZnO Film on Si

Zn2p O1s Si 2p Etching
time

180 sec

150 sec

120 sec

90 sec

60 sec

30sec

0 sec

Courtesy Sampada, MNCF

Oct 2024 S. Avasthi, NE201A 15

15

XPS System: Diagram

UHV System

http://chemwiki.ucdavis.edu/@api/deki/files/9549/

Oct 2024 S. Avasthi, NE201A 16

16

S. Avasthi 8
NE201A: XPS Oct 2024

Source Optics
Area of analysis defined
by electron lens
Area of analysis
defined by X-ray spot

X-ray spot size of 10 um X-ray spot size of 1 mm

This is what we have in the Kratos system below

Oct 2024 S. Avasthi, NE201A

From Kratos Ultra manual 17

17

Transfer Lens Assembly

Assembly complicated because it defines the analysis area

Analyser slit plate

Projector lens Projector lens slows down the electrons.

Why? Wait for a few slides

Objective lens
Analysis-area defining aperture

Angle defining iris Scan the spot in X-Y plane

Charge
neutralizer Sample

From Kratos Ultra manual

Oct 2024 S. Avasthi, NE201A 18

18

S. Avasthi 9
NE201A: XPS Oct 2024

Transfer Lens Assembly: Magnetic Lens

Kratos uses a magnetic lens to focus the Remember, focal length needs to
photoelectrons match the KE of the electron
Oct 2024 S. Avasthi, NE201A From Kratos Ultra manual 19

19

Transfer Lens Assembly: Neutralizer

Filament for Properly focussed
thermionic electrons photoelectrons

Charge balance
plates (-ve)
Under-focussed (high-energy)
photoelectrons emitting
Over-focussed (low- secondary electrons
energy) photoelectrons
emitting secondary
electrons
Neutralizer reduces
Thermionic electrons
for neutralizing surface charging.
But secondary emissions
Magnetic lens from unfocussed
photoelectrons also help.

Oct 2024 S. Avasthi, NE201A From Kratos Ultra manual 20

20

S. Avasthi 10
NE201A: XPS Oct 2024

Scanning in X-Y
Detector slit Scanning electrodes allow raster
scan of the analysis point

This is virtual scanning, because

mechanically nothing moves

Aperture

Scanning electrodes

Iris

Sample top view

Sample side-view
Oct 2024 S. Avasthi, NE201A 21

21

XPS Imaging

• Images acquired in less than 60s

– Virtual scan, not mechanical
• Field of view (FOV) changed by selecting predefined lens modes for a specific
magnification.
800 mm FOV 400 mm FOV 200 mm FOV
• Resolution = area of analysis (>10um)

Au
Oct 2024 S. Avasthi, NE201A
Courtesy Kratos 22

22

S. Avasthi 11
NE201A: XPS Oct 2024

XPS Imaging

Imaging can be especially

useful for probing spatially
non-uniform samples

Like EDS…

However, the resolution of XPS

imaging is 1000x worse than SEM
(EDS)

http://www.udel.edu/chem/beebe/surface.htm

Oct 2024 S. Avasthi, NE201A 23

23

Electron Energy Analyser

From Briggs & Seah

-ve voltage

-ve voltage
w
Hemispherical Analyser From transfer lens
To detector
For circular motion, electrical and centripetal forces need to balance → Works as band-pass filter

Δ𝐸 𝑤 To resolve a 1000 eV electron to a 0.5 eV resolution, R = 1.2m!!

= 0.63
𝐸' 𝑅' Hence, projection lens is use to slow down the electron!

Oct 2024 S. Avasthi, NE201A 24

24

S. Avasthi 12
NE201A: XPS Oct 2024

Effect of Pass Energy (w)

Lower the pass energy, higher

the resolution

But, at the expense of signal

PET sample

strength

For high-resolution, use 10-15 eV

For imaging, use 160 eV
For survey, use 160 eV

From lecture notes on XPS by R. Smart, S. McIntyre, M. Bancroft, Igor Bello & Friends
City University of Hong Kong & Surface Science Western, UWO
Oct 2024 S. Avasthi, NE201A 25

25

XPS for Measuring Valence Band Energy?

• Can X-Rays excite valence electrons?

Electron out
– Yes Photon in (K.E = hν – EV)
(hν)
• Si VB is 5.17 eV below vacuum level Evac = 0
Vacuum level

– Al Kα photon =1486.6 eV
– Expected KE = 1481.43 eV
– Subtracting two large numbers (photon & KE)to
get a smaller number (VB level)
Conduction band
– XPS resolution is only ± 0.5 eV EC
(empty)

EV Valence band
• Use smaller photon energy - +- - - (filled)
– Ultraviolet photoemission spectroscopy (UPS)
– Photon energy = 22 to 44 eV

Oct 2024 S. Avasthi, NE201A 26

26

S. Avasthi 13
NE201A: XPS Oct 2024

Ultra-Violet Photoelectron Spectroscopy (UPS)

Oct 2024 S. Avasthi, NE201A 27

27

UV Source

• He discharge lamp is used to generate UV

light
– At high pressure generate He(I) 21.22 eV Omicron UV source
– At low pressure generate He(II) 40.81 eV
– Typical discharge pressure ~ 1 Torr Analysis
chamber at
10-10 Torr
• Nothing is transparent to this UV
radiation UV + gas Discharge Gun
– Can’t have a window, so gas leaks into the out 1kV
chamber
– Need to use inert, non-interacting gas like
He.
2nd stage 1st stage He inlet
• To maintain low-pressure in chamber, Turbo rotary
discharge lamp is differentially pumped pump at pump at
10-6 Torr 0.1 Torr
Oct 2024 S. Avasthi, NE201A 28

28

S. Avasthi 14
NE201A: XPS Oct 2024

UPS vs. XPS

Parameter XPS UPS

Energy >1000 eV 21.22 (He I) or 40.81 eV (He II)
Source Target Al, Mg, etc. He
KE of photoelectron 100 – 1300 eV 10 – 41 eV
Energy resolution 0.5 – 1.0 eV ~0.1 eV
Inelastic mean free path 1 – 3 nm
Sampling depth 3 – 10 nm ~ 1 – 2 nm
Used for Core level measurement Valence band & work function
measurement

Oct 2024 S. Avasthi, NE201A 29

29

XPS & UPS: Impact of Secondary Electrons

From Rudy Schlaf’s tutorial on PES calibration

Oct 2024 S. Avasthi, NE201A 30

30

S. Avasthi 15
NE201A: XPS Oct 2024

UPS Spectrum for Semiconductors Measured signal

Intensity

VBM - EF
From Rudy Schlaf’s tutorial on PES calibration

Fermi level
Valence
φS secondary band

electrons

Binding energy

BE = 21.22 eV

BE = 21.22 - φS

BE = 0
UPS (or XPS) can be used to measure
work-function of the sample
Oct 2024 S. Avasthi, NE201A 31

31

Inverse Photo-eMISSION Spectroscopy (UPS)

Oct 2024 S. Avasthi, NE201A 32

32

S. Avasthi 16
NE201A: XPS Oct 2024

Inverse Photo-Emission Spectroscopy (IPES)

• Electron-in, photon out.

• IPES yields conduction band energy - Electron in
Photon out (K.E = E0)
level
(hν = E0 – EC)
Vacuum level
Evac = 0

Conduction band
EC
(empty)

EV Valence band
- +- - - (filled)

Oct 2024 S. Avasthi, NE201A 33

33

Case Study

How to put it all together in an actual sample

Oct 2024 S. Avasthi, NE201A 34

34

S. Avasthi 17
NE201A: XPS Oct 2024

Case Study: PQ on Silicon

• Fang et al. (2002) proposed that PQ

bonds to silicon
9,10-Phenanthrenequinone (PQ).

• Our goal:
– Confirm proposed bonding mechanism
– Band diagram of PQ/Si interface

Oct 2024 S. Avasthi, NE201A 35

35

PQ 10 nm
Confirm Bonding Mechanism p+ silicon
substrate

C=O peak should be between

287.7–289.3 eV
But, it is 286.2 eV
↓
Must have been modified!

Oct 2024 S. Avasthi, NE201A 36

36

S. Avasthi 18
NE201A: XPS Oct 2024

PQ 10 nm
p+ silicon
Si 2p Binding Energy substrate

• X-ray penetrate PQ layer

– Probe the Si core levels.
• Measured BE is different
– Expected: 98.9 ± 0.1 eV
– Measured: 99.1 eV ± 0.1 eV
• Due to surface band-bending!

Surface
Si EVAC 0.2 eV band-
Expectation Measured
EF bending
Si EVAC Si EVAC
EF EF 98.9 eV 99.13 eV
98.9 eV 99.13 eV
Si 2p
Si 2p Si 2p

Oct 2024 S. Avasthi, NE201A 37

37

PQ 10 nm
p+ silicon
Ultra-violet photoemission spectroscopy substrate

• Probes the PQ HOMO levels only

– UV rays do not penetrate deep enough to reach Si surface He I 21.22 eV radiation

Measured
Interface dipole value of p+ Si
Si EVAC
1.05 eV
PQ EVAC
φ 5.17 eV
4.12 eV CB
EF VB
1.21 eV
HOMO

PQ p+-Si
Work function (φ) = 4.12 eV
VBM - EF = 1.21 eV

Oct 2024 S. Avasthi, NE201A 38

38

S. Avasthi 19
NE201A: XPS Oct 2024

PQ 10 nm
p+ silicon
IPES for Empty Energy Levels substrate

Si EVAC
PQ EVAC
Φ =4.12 eV 5.17 eV
LUMO
1.97 eV CB
EF VB
HOMO

PQ p+-Si

EF - CBM= 1.97 eV

Oct 2024 S. Avasthi, NE201A 39

39

PQ/Silicon interface band alignment

• Band bending:
– There is 0.2 eV band-bending in silicon. 0.2 eV band
Interface bending in Si
dipole (ΔEVAC )
• PQ Bandgap
EVAC
– (VBM – EF) + (EF – CBM) = 1.21 + 1.97 eV = 1.05 eV
3.18 eV EVAC
ΔEC
• Interface dipole: LUMO
– Vacuum levels are not aligned at the 0.85 eV EC
PQ/p+-Si interface due to the interface 3.18 eV
dipole. EV
– ΔEVAC = 1.05 eV 1.21 eV
HOMO
ΔEV
• PQ/p+-Si is a type I heterojunction
– ΔEC= 0.85 eV PQ p+-Si
– ΔEV= 1.21 eV

Oct 2024 S. Avasthi, NE201A 40

40

S. Avasthi 20