Experimental design for the optimization of
multi-residual analysis of oxygenated
metabolites of PAHs (hydroxylated, quinones)
in sediments
I. BERGER , N. MACHOUR, C. MORIN, F. PORTET-KOLTALO
University of Rouen
France
ICCE 2017 - 21/06/2017
1
TABLE OF CONTENTS
1) Presentation of contaminants and environmental matrix
2) Simultaneous MAE extraction and analyzes of two families of oxygenated
PAHs
a) Choice of chromatographic analytical tools
b) Optimization of MAE extraction by experimental design
3) Conclusion and perspectives
2
Contaminants
Hydroxy-PAHs (OH-PAHs):
DANGEROUS!!
Quinones:
No standardized methods for oxygenated PAHs (oxy PAHs)
3
Matrix
Sediment modelling a natural sediment from a Normand harbors
Silt: ~70% Clay: ~20% Sand: <5% Organic matter:
2,5-10%
Organic matter
Difficult
Finest particles
4
OBJECTIVES
Develop a method to extract simultaneously a mixture of four hydroxylated
PAHs (OH-PAHs) and six carbonyl PAHs (quinones) from sediments (MAE)
Develop a method to analyze these compounds at trace levels (GC-MS and
HPLC-FLD/UV)
5
2) Simultaneous MAE extraction
and analyzes of two families of
oxygenated PAHs
a) Choice of chromatographic
analytical tools
6
Choice of analytical tools
HPLC – UV/FLD (coupled) Low limit of detection (LOD) and limit of quantification
(LOQ)
Quinones Hydroxy-PAHs
Curve calibration 2-Naphthol Quinones Hydroxy-PAHs
3000000
LOD:2,4 - 4,3 µg/L LOD: 0,2 - 0,3 µg/L
2500000 y = 2E+08x + 27972
R² = 0,9953
LOQ: 8,0- 14,2 µg/L LOQ: 0,6- 1,0 µg/L
2000000
1500000
Legend:
LOD = 3.3 Sy / k
1000000 K: slope of the calibration curve
500000
LOQ = 10 Sy / k Sy: standard error of the predicted
0
y-value for each x-value
0 0,002 0,004 0,006 0,008 0,01 0,012 0,014
7
GC-MS
Without derivatization (quinones) With derivatization (quinones)
7,819 : 1,4-Benzoquinone acetylated
6,438 : 1,4-
Benzoquinone
7,900 : 1,4- 12,075 : 1,2-Naphthoquinone
19,412: 9,10-Phenanthrenequinone
Naphthoquinone acetylated
acetylated
8,077 :
Coumarin
13,583 : 9,10-anthracenequinone
11,292 : 12,768 : 1,4-
Phenanthre Naphthoquinone 25,247 :
11,675 :
ne D10 acetylated 19,654: 9,10- Perylene
Phenanthrene
Anthracenequinone D12
D10 acetylated
25,284 : Perylene D12 8,061 :
Coumarin
Silylation
GC-MS
Acetylation
8
Silylation of hydroxy-PAHs
Catalysts: Pyridine and ethyl acetate
Best conditions: BSTFA+ TMCS, pyridine and ethyl
acetate in 5 minutes of reaction
Time(min): 5, 15, 30, 45 and 60
Sensitivity improved by a
LOD: 90,0-220,0 µg/L LOQ: 300,0-720,0 µg/L With derivatization
factor 3
LOD: 180,0- 600,0µg/L LOQ: 610,0-2000,0 µg/L Without derivatization
9
Acetylation of quinones
- 1000µL solution
of quinones
- 0,1g Zn Cooling to room Centrifugation 2000 rpm,
Heating 80°C, 15 min
- 400µL acetic temperature 10 min
anhydride
0,1g Zn 1°) 1000µL H2O Take the above
2°) 3,0mL organic phase
Heating 80°C, 15 min dichloromethane
Add 60µL octanol Aqueous phase
LOD: 190,0-290,0µg/L LOQ: 640,0-960,0µg/L With derivatization
Organic phase evaporated discarted
Without (under N2 flow)
LOD: 560,0-10000,0µg/L LOQ: 850,0-33000,0µg/L
derivatization Redissolved in acetonitrile
until 1000µL+10µL internal
Sensitivity improved by a factor 4-53 Ortho quinones standard
10
2) Simultaneous MAE extraction
and analyzes of two families of
oxygenated PAHs
b) Optimization of MAE
extraction by experimental design
11
Microwave assisted extraction
MAE Soxhlet Sonication
Time of extraction 3- 30 min 3-48hrs 10-60min
Sample amount 1-10g 1-30g 1-30g
Solvent volume 10-40mL 100-500mL 30-200mL
MAE never tested for
quinones and hydroxy-PAHs
ADVANTAGES!!
12
Microwave assisted extraction
First trials
Best results for hydroxy-PAHs
Volume(mL): 10 and 20 100,000
90,000
Temperature(°C): 80, 100 and 120 80,000
70,000
Solvent: 60,000
50,000
- Acetonitrile 40,000
30,000
- 90%Acetonitrile/10%toluene* 20,000
10,000
- 90% Acetonitrile/10%dichloromethane 0,000
Acetonitrile/10%toluene - 100°C MAE 10min 20mL
- 50% Acetone/50%toluene** 2-Naphthol silylated 2-Hydroxyfluorene
Time(min): 10, 20 and 30 9-Phenanthrol silylated 1-Hydroxypyrene silylated
*Oriol et al., Anal. Methods, 2013, 5, 6297-6305
**Optimal for PAHs
13
Microwave assisted extraction
160,000 Best results for quinones
Volume(mL): 10 and 20 140,000 1,4-Benzoquinone acetylated
Coumarin acetylated
Temperature(°C): 80, 100 and 120 120,000
1,2- Naphthoquinone acetylated
1,4- Naphthoquinone acetylated
Solvent: 100,000
9,10-Phenanthrenequinone acetylated
Recoveries (%)
- Acetonitrile 80,000
- 90%Acetonitrile/10%toluene 60,000
- 90% Acetonitrile/10%dichloromethane 40,000
- 50% Acetone/50%toluene 20,000
Time(min): 10, 20 and 30 0,000
Acetonitrile/10%toluene- 100°C 20min 20mL
14
Microwave assisted extraction
Not the same conditions of extraction for the two families
Univariate optimization not appropriate chemometric approach to find the influent
factors, their interactions and a compromise for the two families
15
First experimental design: fractional
4-1
factorial design 2
Temperature Time
• Screening design
Tests Volume solvent Nature solvent
Extraction extraction
Influent factors and possible interactions?
1 80°C (-1) 10mL (-1) CH3CN/10%CH2CL2 (-1) 10min (-10)
• 2 levels + 0 center points
2 80°C (-1) 10mL (-1) CH3CN/10%toluene (+1) 30min (+1)
3 80°C (-1) 30mL (+1)
CH3CN/10%CH2CL2 (-1)
30min (+1)
• Results (recovery yields):
4 80°C (-1) 30mL (+1) CH3CN/10%toluene (+1) 10min (-10) 1. Most influent factors: Temperature and
volume
5 120°C (+1) 10mL (-1) CH3CN/10%CH2CL2 (-1) 30min (+1)
6 120°C (+1) 10mL (-1) CH3CN/10%toluene (+1) 10min (-10) 2. Not influent: Time set to 10 minutes
7 120°C (+1) 30mL (+1) CH3CN/10%CH2CL2 (-1) 10min (-10) 3. Solvent: compromise for the two
8 120°C (+1) 30mL (+1)
CH3CN/10%toluene (+1)
30min (+1)
families Acetonitrile/Dichloromethane
90/10
9 - 15 100°C (0) 20mL (0) CH3CN (0) 20min (0)
16
Second Experimental design: central
composite design 2 2
Temperature
Tests Volume solvent
Extraction
1 80°C (-1) 15mL (-1)
• Surface response design only to
2 80°C (-1) 35mL (+1) temperature and volume studied
3 120°C (+1) 15mL (-1)
• 5 levels non linear modeling
4 120°C (+1) 35mL (+1)
5 72°C (-α) 25mL (0)
6 128°C (+α) 25mL (0)
7 100°C (0) 11mL (-α)
8 100°C (0) 39mL (+α)
9- 13 100°C (0) 25mL (0)
17
Second Experimental design: central
composite design 2 2
Best conditions
T = + 1,41 (128ºC)
V = + 0,78 (33mL)
Response surface for quinones
18
Second Experimental design: central
composite design 2 2
Best conditions
T = + 0,46 (110ºC)
V = + 0,38 (29mL)
Response surface for hydroxy-PAHs
19
Second Experimental design: central
composite design 2
Influent factors
Tᶾ > V² > T x V
Best conditions for the two families
T= 128ºC (+ α)
V= 26mL (+0,12)
Together fitted
20
4) Conclusion and perspectives
21
CONCLUSION
Derivatizations before GC-MS improve the detection of the hydroxy-PAHs and quinones
(particularly ortho-quinones)
HPLC-UV/FLD is more sensitive than GC-MS but GC-MS allows unknown compounds
The best conditions for the extraction of two oxygenated families were found for MAE (
time, solvent, temperature and volume) need to validate the method MAE- GC-MS
22
PERSPECTIVES
Modeling of MAE - HPLC-UV/FLD to do
Comparison of the two methods MAE – GC-MS and MAE – HPLC – UV/FLD
23
Ingrid Brito Berger
Doctorante
Ingrid.berger@etu.univ-rouen.fr
24
