plasma-particle interactions in a laser-induced …...laser-based diagnostics laboratory david w....
TRANSCRIPT
Laser-Based Diagnostics Laboratory
David W. Hahn Michael Asgill, Prasoon Diwakar
University of Florida
SPIE Laser Damage Sept. 26, 2012
Plasma-particle Interactions in a Laser-induced Plasma –
The Path Toward Quantitative Analysis
Laser-induced Plasma Spectroscopy
2
Multi-photon and cascade
ionization creates plasma
Ne ~ 1018 cm-3
T ~ 30,000 K
Plasma forms the sample
volume, dissociating
molecules & fine particles
Well suited as a rapid,
real-time analytical
scheme
0
500
1000
1500
2000
2500
3000
320 340 360 380 400 420 440 460Wavelength (nm)
Re
lative
In
ten
sity (
a.u
.)
Laser-induced
plasma
Nd:YAG
(1064-nm)
(10-ns)
Czerny-Turner
ICCD array
Atomic emission
spectroscopy
Plasma-particle interactions drive analyte signal
380 385 390 395 400 405
Inte
nsity (
a.u
.)
Wavelength (nm)
• Plasma-analyte interactions
• Our evolution of understanding
• Means to overcome matrix effects
3
Plasma
Analyte Discrete
mass
Atoms
& ions
Vaporization & Dissociation
Heat & Mass Transfer
Matrix Effects?
Overview of Talk
“Quantitative spectrographic analysis has proved to be impossible.”
H. Kayser, Handbuch (1910)
“I do not want to arouse exaggerated hopes. The spectrographic analysis
has its limitations as have all analytical methods….but the method can
compete with other quantitative methods….in the range of low
concentrations.” K. Kellerman, Metals and Alloys (1929)
V in steel
(1929)
Historic perspective: A continuous evolution
Courtesy of
Ben W. Smith
University of Florida
LIBS has demonstrated promising sensitivity
5
B. atrophaeous (davg ~ 1 mm)
10 mm
Single-shot, Single-spore spectrum of aerosolized B. atrophaeous
385 390 395 400 405 410In
ten
sity (
a.u
.)
Wavelength (nm)
Ca
Ca
LOD ~1.5 fg Ca
LIBS has demonstrated promising precision
6
Carbon in Steel
L. Barrette and S. Turmel, Spectrochimica Acta B, 56, 715-723 (2001)
Nonetheless, calibration remains an issue
7 Precision & Accuracy?
An integrated approach to quantitative LIBS
8
What are the puzzles for quantitative LIBS?
9
1
10
100
1 10 100
Sili
con E
mis
sio
n P
/B (
a.u
.)
Diameter Cubed (mm3)
Single SiO2
particles
~2.1 mm
Diameter Cubed (mm3)
Upper Size Limit
Carranza & Hahn, Anal. Chem. (2002)
S
ilic
on
Em
iss
ion
(a
.u.)
• ~5 mm for glucose
particles. E. Vors &
Salmon, Anal. Bioanal.
Chem. (2006)
• ~7 mm for copper
particles. Gallou et al.,
Aero. Sci. Tech. (2011)
Other Studies
Upper size limit: The marble theory
10
• Consider the physics:
Data suggests a rate limitation
rather than an energy limitation
Marble Molecule
Incomplete
Sampling Complete
Sampling
~5 mm
Physical
Limit
Upper size limits vs. residence time
11
Nd:YAG
Spectrometer
iCCD
Mixing-
drying
chamber
Flow
controller
(Air)
Exhaust
Flow
controller
(Argon)
Delay
generator
Delay:
15 to
70 ms Analyte:
Silica spheres
2.47 or 4.09 mm
Nd:YAG
~275 mJ @ 1064 nm
Quantify Si emission
12
Si I
288.16 nm
284 285 286 287 288 289 290 291 292
Em
issio
n I
nte
nsity (
a.u
.)
Wavelength (nm)
35/5 gate
70/20 gate
Si4.09 mm
2.47 mm
Quantify Si emission
13 (ms)
Residence time
does not
extend upper
size limit
Expect: Ratio of 4.5
for complete vaporization
Now explore
rate limits
Consider the physical processes: Limiting Rates
14
A) Rate-limited
processes:
Dissociation &
diffusion are slow
relative to analytical
time-scales
B) Infinite rates:
Dissociation &
diffusion are very
fast relative to time-
scales for analysis
Plasma
Imaging
study
Imaging study of single particles
Nd:YAG
(1064 nm)
Spectro-
meter iCCD
Pierced
Mirror
Fiber
Optic
Sample
Chamber:
6-way cross
iCCD
1064 mirror with
UV-grade substrate
396.2-nm line filter
(3-nm fwhm)
Aerosol Stream
(~2-mm glass particles)
Ca II emission
396.85 nm (0-25,192 cm-1)
2 ms 4 ms 8 ms 15 ms
Evidence of finite rates: Atomic calcium cloud
Plasma Residence Time
0
100
200
300
400
500
600
0 5 10 15 20 25 30 35
Tota
l Det
ecte
d C
alci
um M
ass
(fg)
Delay Time (ms)
Total
detected
calcium
D ~ 0.04 m2/s
~ 15-20 ms for dissociation
Localized plasma processes and effects
17
Calcium emission from a single particle @ 2 ms Plasma-analyte
interactions are initially limited to region
~ 1/1000 of total plasma volume
Mass Transfer
to Plasma
Heat Transfer
to Particle
Finite Rates lead
to localized
plasma
perturbations
& matrix effects
Analysis of matrix effects: Experimental details
18
Nd:YAG
Spectrometer
iCCD
Mixing-
drying
chamber
Flow
controller
(Air)
Exhaust
Flow
controller
(Argon)
Delay
generator
Delay:
100 ns to
100 ms
Analyte:
Al,Lu,Mn
&
Al,Lu,Mn + Na
Nd:YAG
~275 mJ @ 1064 nm
Spectral data: Corrected for relative response
19 19
Experimental details: Aerosol generation
20
200 nm
200 nm
Mn/Al/Lu
Mn/Al/Lu + Na
@ 8x mass
• ~50-500 nm particles following desolvation
What are the puzzles for quantitative LIBS?
21
Matrix Effects
25 ms
Analyte enhancement
& continuum reduction
with sodium addition
Are finite process rates related to matrix effects?
22 Plasma Residence Time (ms)
Strong matrix
effects early in
plasma evolution
Matrix effects
diminish with time
Are finite process rates related to matrix effects?
23
~20-30 ms
Break in T
decay slope….
Suggests
completion of
vaporization
and
equilibration
Plasma Residence Time (ms)
Understanding plasma / analyte dynamics
24
Early times Later times
Analyte T & Ne
Localized effects
Mass/matrix effects
Bulk plasma
T & Ne Equilibration
between plasma
& analyte
?
The path forward for quantitative analysis
25
Must allow sufficient plasma residence time for dissociation, diffusion
of heat & mass, and equilibration of analyte species with bulk plasma
Local perturbations:
Matrix effects
Diffuse analyte:
Bulk analytical plasma
provides more ideal response
Residence
time
What about direct analysis of solids?
26
• LIBS combines the target
sampling with the analytical
measurement
• Different plasma evolution for
different materials
• Necessitates matrix-matched
standards
Consider
LA-ICP-OES
• Uncouples the
laser sampling
event from the
analytical plasma
Laser-Ablation LIBS (LA-LIBS)
27
• Separate the laser-ablation process and
analytical plasma to uncouple these effects
LA-LIBS: Transport efficiency considerations
28
Positive
Pressure
Reduced
Pressure
Analytical
Plasma
Laser
Ablation
Single-shot crater
Ablation
Cell
LIBS
Cell
• Minimize ablation cell volume
• Transport directly to LIBS
plasma via carrier gas flow
• Vacuum vs. positive pressure?
LA-LIBS: Transport efficiency considerations
29
• Clear maximum in analyte signal
• 50% improvement with suction
LA-LIBS: Experimental Sample Matrix
30
Glass 2
Copper-
Nickel
Cobalt-
Chrome
Sample Fe (%)
Mn (%)
SM-10 Al-alloy
1.96 0.30
1276-a Cu-Ni alloy
0.56 1.01
1242 Co-Cr alloy
1.80 1.58
1297 Fe-Cr alloy
69.4 7.11
1761 High Fe
95.3 0.68
Glass 1* Si-K-Ca
0.64 1.12
Glass 2* Si-K-Ca
0.27 0.66
*Courtesy of Anna Matiaske & Ulrich Panne – BAM (Berlin, Germany)
Wide range
of sample
matrices
LA-LIBS: Sample spectra
31
Mn
Mn
LA-LIBS: Mn/Fe calibration curve
32
NIST
1242
Co-Cr
Cr line
interference
with Mn line • Al alloy
• Cu-Ni
• Fe-Cr
• High Fe
• Glass
LA-LIBS: Absolute calibration?
33
Long-standing interest in
LIBS for analysis of soils
*In collaboration with Prof. Alejandro Molina and Jhon Pareja
University of Colombia - Medellin
• Use LA-LIBS for analysis
of soils
• Dope with fertilizer to
different concentrations
of N, P and K
• Use a single-laser
configuration for total
concentration analysis
LA-LIBS: Experimental set-up
34
Laser beam
Lens
Mirror
Fiber optic
Ablationspark
LensLens
Pierced Mirror
Beam splitter
Carrier gas inlet
Shaft
Laser ablation cellAnalytical
LIBS plasma
Soil sample
Split the laser to produce both beams
Analytical
Plasma
Laser
Ablation
LA-LIBS: Spectral data for soils
35
K K
LA-LIBS Direct LIBS
LA-LIBS: Calibration results
36
LA-LIBS Direct LIBS
R2=0.8843 R2=0.6954
Superior correlation
& near-zero intercept
Summary remarks
37
• Particle dissociation, atomic diffusion, & heat transfer have similar, finite time-scales that result in localized plasma perturbations. Temporal considerations are important to minimize matrix effects for quantitative analysis.
• LA-LIBS approach can improve LIBS by uncoupling the laser-ablation process from the analytical plasma processes. Moves us closer to the use of non-matrix matched standards.
• Understanding of the fundamental processes
will improve LIBS as an analytical method, with implications to the larger analytical community (e.g. ICP-AES & LA-ICP-MS).
L I
B S
The future of LIBS: Applications
38
The future of LIBS: Advanced spectral analysis
39
Acknowledgements
40
Graduate students:
Michael Asgill
Prasoon Diwakar
Bret Windom
Kris Loper
Vince Hohreiter
Jorge Carranza
Collaborators:
Prof. Nico Omenetto
Prof. Kay Niemax
Prof. Ulich Panne
Prof. Alejandro Molina
Funding:
• NSF (CHE 0822469)
• NSF (CBET 0317410)
Thank you
41
2009 North American
Symposium on LIBS
Mississippi River, New Orleans
Laser-Based Diagnostics Laboratory
Florida Museum of Natural History