X - R AY E N E R G Y D I S P E R S I V E S P E C T RO S C O P Y: OV E RV I E W, A P P L I C AT I O N S , & A N A LYS I S

D AV E S TA L L A , M . S .

S E N I O R R E S E A R C H T E C H N I C I A N , E L E C T RO N M I C RO S C O P Y C O R E

& P H . D. C A N D I D AT E , P H Y S I C S

U N I V E R S I T Y O F M I S S O U R I

Electron Microscopy CoreUniversity of Missouri

ABOUT ME

• Ph.D candidate in physics (ant. grad: June 2017) with Prof. Peter Pfeifer

– Dissertation focuses on H2 and CH4 storage via adsorption in carbonaceous materials

– Characterization: XPS, XRD, U/SAXS, U/SANS, PGAA, FTIR, N2 sorption (BET)

– Experience with synthesizing raw (high-Σ carbons) and functionalized (B-doped carbons, bulk C3N4)

materials

– Technical experience includes Mathematica programming and Arduino microcontroller

design/implementation

• Hired as Senior Research Technician 9/2016

– EMC user since 2010

– Travelled to Bruker HQ 11/14-17 for “Nano Analytics QUANTAX User School”

E-BEAM INTERACTIONQuanta/Hitachi/Scios

JEOL/F30

F30

Quanta/Hitachi/Scios/F30

CHARACTERISTIC X-RAYS

e- beam

incident electron

secondary electron

generation

Incident electrons scatter

electrons from the

sample, resulting in

secondary electron

generation for imaging.

The electron ejection

results in a vacancy in it’s

former orbital.

CHARACTERISTIC X-RAYS

e- beam

Ex-ray = Eshell1 – Eshell2

Electrons from higher orbitals

will fill this vacancy as is

energetically favorable.

As the electron jumps to a

lower energy state, it emits the

excess energy (ΔE) in the form of an x-ray photon.

The energies associated with these transitions are unique and well-tabulated.

CHARACTERISTIC X-RAYS

SI-LI VS. SDD

SDD design – significant improvements to collection efficiency!

capable of detecting elements as low as B

Si-Li detectors: ~10,000-20,000 CPS, above which data becomes unreliable

SDD detectors: routinely >100,000 CPS

specifications list upwards of 1,500,000 CPS possible

HIGH THROUGHPUT ON BRUKER EDS SYSTEM AT EMC

High count rate (routinely work at ~150-200kcps) generates remarkable

data output - with the exception of mapping, quantifiable EDS data

collection comprises a marginal amount of instrument time.

Bruker Esprit – powerful microanalysis suite associated with the

QUANTAX detector. Unlimited network license allows users to the

software on their own machines and process data offsite: all instrument

time can be dedicated to maximizing the amount of data collected!

ESPRIT: 4 MAIN ANALYSES

Increasing

spatial

information

‘Higher order’ analyses retain

functionalities of earlier ones

Windowed mode

Fullscreen mode

I. SPECTRUM COLLECTION

• Collects a spectrum

across the entire

imaged area

• ~20s / scan

• Peak deconvolution

aids in identification

from overlapping

features

• Standards-based &

standardless quant.

available

• Many options for

easy plot generation

and exporting

SPECTRUM ACQUISITION

• Total counts in spectrum

• Fixed real/live time

• Counts across a particular element’s

energy region

• Manual: collect until requisite statistics are

achieved

PEAK IDENTIFICATION

• Manual selection of elements

• Auto ID for well-resolved features

• Finder tool: highlight a region, and the

software will list lines that fall within that

region in order of ‘likelihood’: α > β > γ,

distance from line from the center of the

region, etc

ONLINE PEAK DECONVOLUTION

Initial assignments appear

to successfully account

for all features present

ONLINE PEAK DECONVOLUTION

Reconstruction of elemental lines

shows that Kβ signals (calculated from associated Kα lines) don’t account for all detected counts

ONLINE PEAK DECONVOLUTION

Deconvolution indicates

the presence of Mn and V

II. OBJECT ANALYSIS

User-defined

regions:

• Point

• Rectangle

• Oval

• Freeform

polygon

• Iterative

auto grid

III. LINE ANALYSIS

• Collects elemental

profiles across a

user-defined line

• Good for analyzing

phase interfaces

and non-discrete

boundaries

• ~5min / scan

• Can generate/

analyze/quantify a

spectrum across

entire line or a

selected segment

IV. HYPERMAPPING

• Distinct from traditional

elemental maps, which

was essentially ‘filtered’

imaging: elements

predetermined &

individually collected

• Collected as a data cube:

spatial information from

pixels in x & y, each with

associated spectral

information (“z-axis”)

• A single collection yields

virtually limitless data

upon post-processing

IV. HYPERMAPPING

• MASSIVE amount of data

(600x400 image has 240k

individual spectra) – most

versatile method

• Elemental maps

• Object analysis

• Line analysis

• Phase deconvolution

• ~10-20 min / region

INDIVIDUAL/COMPOSITE MAPS

Can generate individual and

composite maps for any

combination of elements

PHASE ANALYSIS

Identify discrete chemical phases via CPU-aided

analysis or manual assignment

ELEMENTAL HEAT MAP

BSE Image Standard Al Map

ELEMENTAL HEAT MAP

BSE Image Standard Al Map

MAXIMUM PIXEL SPECTRUMReference BSE Image Associated elemental map

MAXIMUM PIXEL SPECTRUM

Maximum pixel spectrum

Indicates the presence of several

additional elements, each of which may

only exist in single-digit no. of pixels

Net hypermap spectrum

Represents the sum of each energy

channel’s counts across ALL analyzed

pixels – high spatial concentrations

dominate

MAXIMUM PIXEL SPECTRUMReference BSE Image ‘Trace’ features map

Most features comprise significantly less than 0.1wt% in the bulk, below the accepted detection limit of EDS

PARTICLE ANALYSIS

• Image analysis utility independent of X-ray data

• Particles are partitioned from CPU-aided filtering and binarization

• A number of different geometric characteristics can be quantified

• May be followed by EDS measurements in defined particle regions for joint particle/phase

analysis

PARTICLE ANALYSIS

VARIABLE Z

15 mm4 mm

WD = 9mm WD = 50mm

Field of view is limited by working distance; most EDS systems

require a fixed WD for sufficient counting (detector must be

pointed at sample surface). EMC QUANTAX system is capable of

alignment to longer working distances for wider analysis regions