FLUORESCENCE AS A TOOL FOR THE CHARACTERIZATION OF WATER AquaLife 2010 Martin Wagn Technologiezentrum Wasser (T Außenstelle Dres
Mar 29, 2015
FLUORESCENCE AS A TOOL FOR THE CHARACTERIZATION OF WATER
AquaLife 2010Martin Wagner,
Technologiezentrum Wasser (TZW)Außenstelle Dresden
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Outline
Principles of fluorescence spectroscopy
Characterization of DOC
Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Measurement of fluorescence
Design of a fluorescence spectrometer
I. Principles of fluorescence spectroscopy
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Measurement of fluorescence
Emission spectrum: λEx = const., λEm
I. Principles of fluorescence spectroscopy
0
50
100
150
200
250
240 340 440 540 640
Flu
ores
cenc
e in
tens
ity [
a.u.
]
Emission wavelength [nm]
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Measurement of fluorescence
Variation of λEx produces an excitation-emission-
matrix, called EEM
I. Principles of fluorescence spectroscopy
λEmission λExcitation
Intensity
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Characterization of water
II. Characterization of DOC
Chl A Chlorophyll APC PhycocyaninPE PhycoerythrinFC FucoxanthinTyr TyrosineTrp TryptophanePhe PhenylalanineEPS extracellular
polymericsubstances
FS fulvic acidlike
HS humic acidlike
Q-o Quinone (oxidized)Q-s SemiquinoneQ-h HydroquinoneBak bacteria like
fluorescence
HS
Chl a
Chl a
Chl a
PC
PE
Tyr
Tyr
Trp
Trp
EPS
EPSQ-oQ-o
Q-s
Q-s
Q-s
Q-s/h
Q-s/hFS
FC
FC
Bak
Biopolymers
Humic substances
Algae pigments
Phe
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Characterization by LC-OCD
Most parameters used to describe DOC are sum parameters (like BOD, COD, UV254, UV436)
LC-OCD (Liquid chromatography – Organic carbon detection) and fluorescence allow the characterization of DOC LC-OCD separates DOC by molecular weight Fluorescence separates DOC by chemical structure or rather chemical
properties
II. Characterization of DOC
0
20
40
60
80
100
120
10 30 50 70 90Retention time [min]
OC
D [
rela
tive
hei
gh
t o
f si
gn
al]
low molecular compounds
Polysaccharides
Humic Substances
Building Blocks
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Quantification of fluorescence
Fluorescence is easy to use and is appropriated for the
characterization of the DOC
The quantification isn’t easy, because of Influence of stray light
Inner – Filter - Effects
Quenching of fluorescence signals
Portability: standardization between different
spectrometers
Spectral overlapping of signals
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Stray light
Caused by scattering of exciting light in sample
Differentiation between Rayleigh- and
Ramanscattering
Rayleigh: elastic scattering without loss of energy
Appears at excitation wavelength
Raman: inelastic scattering with loss of energy
Appears at longer wavelengths
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Stray light
III. Problems in quantification of fluorescence signals
stray light in pure water
0
50
100
150
200
250
300
260 360 460 560 660
emission wavelength [nm]
fluor
esce
nce
inte
nsity
[a.u
.]
Rayleigh peaks
Raman peaks
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Stray light
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of stray light problem
Best method is the use of cutoff filters
III. Problems in quantification of fluorescence signals
0
20
40
60
80
100
120
200 300 400 500 600 700
T [%
]
wavelength [nm]
Cutoff filter
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of stray light problem
Best method is the use of cutoff filters
III. Problems in quantification of fluorescence signals
0
0,2
0,4
0,6
0,8
1
1,2
300 350 400 450 500
norm
aliz
ed in
tens
ity [-
]
emission wavelength
Quinine sulfate without filterQuninine sulfate with cutoff filter (290 nm)
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Inner – Filter – Effects (IFE)
Primary IFE: absorption of excitating light by
sample
Secondary IFE: absorption of emitted light
III. Problems in quantification of fluorescence signals
Calibration
020406080
100120140160180
0 2 4 6 8 10 12
Concentration [mg/L]
fluor
esce
nce
inte
nsity
[a.u
.]
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of the IFE-Problem
Generally there are two methods: Additionally measurement of absorption spectrum of sample
Correction via stray light peaks of the sample (Raman peak)
Absorption
0
0.5
1
1.5
2
2.5
3
200 300 400 500 600 700
wavelength [nm]
ab
sorp
tion
III. Problems in quantification of fluorescence signals
LAKOWICZ (2006):
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of the IFE-Problem
III. Problems in quantification of fluorescence signals
correction of Inner-Filter-Effects by ratio between raman peak of pure water and sample
0
100
200
300
400
500
600
240 290 340 390 440 490 540 590 640 690
emission wavelength [nm]
flu
ore
scen
ce in
ten
sity
[a.
u.]
pure water sample
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of the IFE-Problem
Result of IFE-correction is a linear relationship
III. Problems in quantification of fluorescence signals
Calibration
0
100
200
300
400
500
0 2 4 6 8 10 12
Concentration [mg/L]
fluor
esce
nce
inte
nsit
y [a
.u.]
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Quenching
Is also a decrease of fluorescence intensity
Results from contact between fluorophor and quenching
molecule Dynamic Quenching: collision between molecules in excited
state High temperatures and high concentrations increase the probability
of collisions
Static Quenching: formation of complex between fluorophore
and quencher Fluorophore isn‘t able to fluoresce any more
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Quenching: an example
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of Quenching-Problem
Relationship between fluorophore and Quencher
can be described by the Stern-Volmer-Law
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of Quenching-Problem
Definition of the most important Quenchers in respect
of raw and drinking water
O2, Cl-, NO3- (surface- and groundwater)
Fulvic acid
Humic acid
Methodical laboratory tests to derive the single
quenchingconstants for every fluorophore-quencher-
pair
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Portability
III. Problems in quantification of fluorescence signals
Proteinfluorescence at two spectrometers
0
50
100
150
200
250
300
350
400
240 340 440 540 640emission wavelength [nm]
fluore
scence
inte
nsi
ty [
a.u
.]
LS50 LS55
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Portability
The reason for the differences is the missing
reference photomultiplier for the emission channel
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of Portability-Problem
Standardization in three steps:
Correction of exciting light: is included in all
spectrometers (reference photomultiplier)
Correction of deformed peaks: via derivation of
correction-function with the use of reference dyes
Normalization of signals via external standard (sealed
pure water cuvette)
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of Portability-Problem
Correction of deformed peaks via reference dyes
III. Problems in quantification of fluorescence signals
LAKOWICZ (2006)
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Solution of Portability-Problem
III. Problems in quantification of fluorescence signals
before standardization after standardization
LS50
LS55
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Short summary
We have learned How a spectrometer does work
How the DOC is characterized by Fluorescence
LC-OCD method
How a quantification is complicated by Stray light
Inner – Filter – Effects
Quenching
Portability
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Spectral overlapping
III. Problems in quantification of fluorescence signals
Component I
0.00
50.00
100.00
150.00
200.00
250.00
220.0 320.0 420.0 520.0 620.0
emission wavelength [nm]
fluor
esce
nce
inte
nsity
[a.
u.]
Component II
0.00
50.00
100.00
150.00
200.00
250.00
220.0 320.0 420.0 520.0 620.0
emission wavelength [nm]
fluor
esce
nce
inte
nsity
[a.u
.]
Mix of both Components
0.00
50.00
100.00
150.00
200.00
250.00
220.0 320.0 420.0 520.0 620.0
emission wavelength [nm]
fluor
esce
nce
inte
nsity
[a.
u.] Component I Component II Mix
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Spectral overlapping
Existing multivariate methods are:
Principal components regression (PCR)
Parallel factor analysis (PARAFAC)
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Principal components regression (PCR)
Need for set of EEM‘s for decomposition
DOC: 1,2 mg/LDOC: 0,4 mg/LDOC: 0,6 mg/L
… … …
training dataset
Collection of samples about one year
New matrix
Quantification of the new matrix
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Principal components regression (PCR)
Comparison between „classical“ calibration and calibration using principal components
Principal components are difficult to interpret
Appropriate for quantification of well known waters, not for characterization
III. Problems in quantification of fluorescence signals
LC-OCD-fraction
R²
(classical)
R²(PC-Regression)
Number of principal
components
TOC 0,86 0,95 4
Biopolymers 0,00 0,73 10
Humic Substances
0,78 0,91 4
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Parallel factor analysis (PARAFAC)
Some kind of „extended“ principal components analysis
III. Problems in quantification of fluorescence signals
20 to ~ 200
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Multivariate analysis
Lack of interpretation (PCA/PCR)
No universal application possible
New calibration for every location or water necessary
High number of samples necessary
PCR mainly applied in process-monitoring (e.g. brewery), where water
always has the same defined composition and may only exhibits
fluctuation of concentration
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Approach of TZW to solve the problem
Target is decomposition based on one EEM
Extended curve fitting approach is used
Allows to remove stray light, if cutoff filters weren‘t able to
remove them
III. Problems in quantification of fluorescence signalsTryptophan fitted with an asymmetric curve and stray light with symmetric curves
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Approach of TZW to solve the problem
Main problem is finding the truth, because several
solutions are possible
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Approach of TZW to solve the problem
Principle of fluoresence: λem = constant
Usage of pattern recognition (DTW)
III. Problems in quantification of fluorescence signals
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
Summary
Advantages
Quick
Little sample preparation
Very sensitive
Disadvantages
Complexity of data evaluation and interpretation
Technologiezentrum Wasser (TZW) – Außenstelle Dresden
The End
Thank you for your attention