Laser Diode Thermal Desorption-Atmospheric Pressure Chemical ...
Laser Diode Thermal Desorption-Atmospheric Pressure Chemical Ionization Tandem Mass Spectrometry Analysis of Selected Steroid Hormones in Wastewater: Method Optimization and Application Paul B. Fayad†, Michèle Prévost‡ and Sébastien Sauvé†*
† ‡
Department of Chemistry, Université de Montréal, Montreal, QC, Canada Department of Civil, Geological and Mining Engineering, École Polytechnique de
Montréal, Montreal, QC, Canada
*Corresponding author. Phone: 514-343-6749. Fax: 514-343-7586. E-mail: [email protected].
S-1
Table of content: Page S-3 to S-4: Principles of the LDTD-APCI source. Page S-5, Figure S-1: Molecular structures of selected steroid hormones and their acronyms. Page S-6, Table S-1: MS/MS Parameters for the Analysis of Selected Steroid Hormones Analytes in Both Negative (NI) and Positive (PI) Ionization Mode. Page S-7, Figure S-2: Choice of deposition solvent in the plate wells on peak area intensities for the selected steroid hormones. Page S-8, Figure S-3: Effect of carrier gas flow on peak area intensities in NI mode of the selected steroid hormones. Page S-9, Figure S-4: Effect of laser pattern on peak area intensities for PROG in PI mode. Page S-10, Figure S-5: SPE recovery values for matrix-free (pure water) and in matrix (spiked effluent wastewater) samples of the selected steroid hormones in PI and NI mode.
S-2
Principles of the LDTD-APCI Source. The Arrhenius plots comparing the logarithm of the kinetic rates of the temperature dependencies of vaporization (dissociation of intermolecular bonds) and molecular decomposition (dissociation of intramolecular bonds) against the inverse temperature (1/T) would identify an intersection point at some value of 1/T, above which the vaporization of the uncharged molecular species would be favored and below which decomposition would be observed.1,2 The heating rate of the LDTD laser is 3000C/sec which allows the samples to be quickly heated at high temperatures, minimizing the time spent in the decomposition region and favoring vaporization which generates a greater amount of the uncharged molecular species. The characterization of the thermal desorption processes associated with rapid sample heating and thin film deposition for the LDTD-APCI source have been shown on prednisone, a corticosteroid steroid hormone, that decomposed at a temperature of 234C but was fully analyzed by LDTD-APCI-MS/MS at a desorption temperature of 170C. Also sulfadiazine, a sulfonamide antibiotic, shows a desorption temperature observed between 95C and 140C though its bulk melting point is 252-256C.3 Unlike traditional LC-APCI, the LDTD-APCI ionization is performed in the absence of solvent reacting molecules as no liquid phase is introduced into the corona discharge region. Similarly to classical chemical ionization, the species present in the APCI region of the LDTD originate from gas-phase reactions involving proton transfer, governed by proton affinity in positive mode (PI) and gas-phase acidity in negative mode (NI), as well as charge exchange.4 The solvent used for analyte deposition in LDTD-APCI will not induce signal suppression due to competition for protonation as can be the case in classical LC-APCI mobile-phase components.5-7 In the absence of any mobile phase, it is the water traces in
S-3
the carrier gas that will generate ionization in LDTD-APCI, since proton transfer in the PI mode mostly occurs between water cluster, H3O+(H2O)n where n = 0-4,8 and more basic analytes (higher proton affinity). Higher efficiency protonation in LDTD-APCI is accomplished in the presence of the smallest hydronium species because of their low proton affinity, i.e. H3O+ and (H2O)H3O+. Larger water clusters will form when the humidity (water saturation) in the ionization region is increased and negatively impact the analyte ionization process.9 The larger hydronium cluster, with higher proton affinity, will affect sensitivity towards low proton affinity molecules and decrease their fragmentation because of an excess of energy in protonation reactions.10 The main advantage of using APCI with the LDTD is that it is less susceptible to matrix effects and ionization suppression than with electrospray ionization.11
S-4
OH
O
H
H
H
H
H
H
HO
H
OH
H
HO
H
HO
Estrone (E1)
Estradiol (E2)
Estriol (E3)
OH
OH
OH
H
H
H
H
OH
H
H
H
H O
O
HO
Norethindrone (NOR)
Levonorgestrel (LEVO)
17-ethynylestradiol (EE2)
H
O O
OH H H
H
H
O
H
CH 3
Medroxyprogesterone (MPROG)
H
O
Progesterone (PROG)
Figure S-1. Molecular structures of selected steroid hormones and their acronyms.
S-5
Table S-1. MS/MS Parameters for the Analysis of Selected Steroid Hormones Analytes in Both Negative (NI) and Positive (PI) Ionization Mode. Compound E1 E2
NI
271 [M–H]‾
PI
255 [M-H2O+H]+
133 144 159
31 ± 3 21 ± 3 100
68 68 68
18 40 19
NI
273 [M–H]‾
145 147 185
57 ± 9 82 ± 12 100
-67 -67 -67
61 43 47
PI
257 [M-H2O+H]+
E3
NI
287 [M–H]‾
EE2
NI
295 [M–H]‾
135 146 161 143 171 143 145 183
52 ± 5 22 ± 1 100 67 ± 12 100 63 ± 12 100 58 ± 13
65 65 65 -97 -97 -82 -82 -82
17 37 16 54 40 70 45 44
PI
279 [M-H2O+H]+
133 159
100 69 ± 11
46 46
16 20
NI
297 [M–H]‾
143 145
72 ± 15 100
-74 -74
43 43
PI
281 [M-H2O+H]+
133 159
100 22 ± 6
48 48
23 22
LEVO
PI
313 [M+H]+
109 185 245
83 ± 19 45 ± 8 100
64 64 64
27 20 18
MPROG
PI
345 [M+H]+
97 123
26 ± 4 100
74 74
23 25
NOR
PI
299 [M+H]+
91 109
85 ± 16 100
69 69
35 26
PROG
PI
315 [M+H]+
E2-13C2
EE2-13C2
Precursor ion (m/z) 269 [M–H]‾
Product ion (m/z) 143 145 143 145 183
Relative intensity ratioa (%) 45 ± 6 100 51 ± 13 66 ± 20 100
Ionisation mode NI
TL (V) -84 -84 -74 -74 -74
CE (eV) 57 41 47 37 49
97 98 ± 17 65 20 109 100 65 24 a The most abundant product ion was used for quantification whereas the other product ions were used for to confirm the presence of the steroid hormones.
S-6
E1 E3
6.0E+04
EE2 E2
5.0E+04
Peak Area
4.0E+04
3.0E+04
2.0E+04
1.0E+04
0.0E+00 5
10
15
20
25
30
40
50
60
80
Laser Power (%)
Figure S-2. Negative ionization mode (NI) maximum peak intensity for a concentration of 2 mg/L was observed at a maximum 20% laser power for steroid hormones EE2 and E2 and was significantly different from other laser power (n=3; P 0.5). The optimum laser power for E1 was 20% and 25%, while for E3 it was 15% and 20%, which were not statistically different from each other in both cases (n=3; P 0.5). A laser power of 20% was chosen in NI mode, in order to have a single method applicable to all selected compounds and because it was common to all four selected steroid hormones.
S-7
E3 E1
1.2E+05
EE2 E2
1.0E+05
Peak Area
8.0E+04
6.0E+04
4.0E+04
2.0E+04
0.0E+00 1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
Gas Flow (L/min)
Figure S-3. Carrier gas flow was optimal between 2 to 3 L/min and gave significantly higher peak area response (n=3; P 0.5) in NI mode. A gas flow of 3 L/min provided the best signal-to-noise ratio combined with small signal variability (RSD 10%; n=3). Estriol (E3) was the only exception with an optimal gas flow of 2 L/min.
S-8
1.4E+06 1.2E+06
Peak Area
1.0E+06 8.0E+05 6.0E+05 4.0E+05 2.0E+05 0.0E+00
3 sec – 20%
1 sec – 20%
2 sec – 20%
3 sec – 20% 3 sec – 2% 1 sec
0.5 sec
0.5 sec
1 sec
1 sec
Laser Pattern
Figure S-4. Four different laser patterns (n=3) were used and compared for PROG in PI mode.
The optimal conditions giving the maximum peak area intensity and good
variability (RSD 10%) for desorption laser power and patter consisted of a 1 sec initial ramp from 0 to 20% and held for 3 sec at 20% before shutting off.
S-9
Matrix-free In matrix
160 140
Recovery (%)
120 100 80 60 40
PROG (PI)
NOR (PI)
MPROG (PI)
LEVO (PI)
EE2-13C2 (PI)
EE2-13C2 (NI)
EE2 (PI)
EE2 (NI)
E3 (NI)
E2 (PI)
E2 (NI)
0
E1 (NI)
20
Compound (ionisation mode)
Figure S-5. Recovery values (n=3) for matrix-free (pure water) and in matrix (spiked effluent wastewater) samples at 50 ng/L in both NI and PI modes using Strata-X SPE cartridges for all selected steroid hormones including IS and surrogate.
S-10
REFRENCES
(1) Beuhler, R. J.; Flanigan, E.; Greene, L. J.; Friedman, L. J. Am. Chem. Soc. 2002, 96 , 3990-3999. (2) Daves, G. D. Acc. Chem. Res. 1979, 12 , 359-365. (3) Picard, P.; Tremblay, P.; Paquin, E. R. Laser Diode Thermal Desorption Ionization Source (LDTD): Fundamental Aspects. 56th ASMS Conference on Mass Spectrometry, Denver, CO. 2008. Available at: http://www.ldtdionsource.com/eng/documentations/publications.asp (4) Harrison, H. G. Chemical Ionization Mass Spectrometry; CRC Press Inc: Boca Raton, FL, 1983.pp. 156. (5) Ma, Y. C.; Kim, H. Y. Journal of the American Society for Mass Spectrometry 1997, 8 , 1010-1020. (6) Geerdink, R. B.; Kooistra-Sijpersma, A.; Tiesnitsch, J.; Kienhuis, P. G. M.; Brinkman, U. A. T. J. Chromatogr. A 1999, 863 , 147-155. (7) Dams, R.; Benijts, T.; Duff, K.; Bell, D. Rapid Commun. Mass Spectrom. 2002, 16 , 1072-1077. (8) Zook, D. R.; Grimsrud, E. P. J. Phys. Chem. 1988, 92 , 6374-6378. (9) Sinha, V.; Custer, T. G.; Kluepfel, T.; Williams, J. Int. J. Mass Spectrom. 2009, 282 , 108-111.
S-11
(10) Tani, A.; Hayward, S.; Hansel, A.; Hewitt, C. N. Int. J. Mass Spectrom. 2004, 239 , 161-169. (11) Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Anal. Chem. 2003, 75 , 3019-3030.
S-12
† ‡
Department of Chemistry, Université de Montréal, Montreal, QC, Canada Department of Civil, Geological and Mining Engineering, École Polytechnique de
Montréal, Montreal, QC, Canada
*Corresponding author. Phone: 514-343-6749. Fax: 514-343-7586. E-mail: [email protected].
S-1
Table of content: Page S-3 to S-4: Principles of the LDTD-APCI source. Page S-5, Figure S-1: Molecular structures of selected steroid hormones and their acronyms. Page S-6, Table S-1: MS/MS Parameters for the Analysis of Selected Steroid Hormones Analytes in Both Negative (NI) and Positive (PI) Ionization Mode. Page S-7, Figure S-2: Choice of deposition solvent in the plate wells on peak area intensities for the selected steroid hormones. Page S-8, Figure S-3: Effect of carrier gas flow on peak area intensities in NI mode of the selected steroid hormones. Page S-9, Figure S-4: Effect of laser pattern on peak area intensities for PROG in PI mode. Page S-10, Figure S-5: SPE recovery values for matrix-free (pure water) and in matrix (spiked effluent wastewater) samples of the selected steroid hormones in PI and NI mode.
S-2
Principles of the LDTD-APCI Source. The Arrhenius plots comparing the logarithm of the kinetic rates of the temperature dependencies of vaporization (dissociation of intermolecular bonds) and molecular decomposition (dissociation of intramolecular bonds) against the inverse temperature (1/T) would identify an intersection point at some value of 1/T, above which the vaporization of the uncharged molecular species would be favored and below which decomposition would be observed.1,2 The heating rate of the LDTD laser is 3000C/sec which allows the samples to be quickly heated at high temperatures, minimizing the time spent in the decomposition region and favoring vaporization which generates a greater amount of the uncharged molecular species. The characterization of the thermal desorption processes associated with rapid sample heating and thin film deposition for the LDTD-APCI source have been shown on prednisone, a corticosteroid steroid hormone, that decomposed at a temperature of 234C but was fully analyzed by LDTD-APCI-MS/MS at a desorption temperature of 170C. Also sulfadiazine, a sulfonamide antibiotic, shows a desorption temperature observed between 95C and 140C though its bulk melting point is 252-256C.3 Unlike traditional LC-APCI, the LDTD-APCI ionization is performed in the absence of solvent reacting molecules as no liquid phase is introduced into the corona discharge region. Similarly to classical chemical ionization, the species present in the APCI region of the LDTD originate from gas-phase reactions involving proton transfer, governed by proton affinity in positive mode (PI) and gas-phase acidity in negative mode (NI), as well as charge exchange.4 The solvent used for analyte deposition in LDTD-APCI will not induce signal suppression due to competition for protonation as can be the case in classical LC-APCI mobile-phase components.5-7 In the absence of any mobile phase, it is the water traces in
S-3
the carrier gas that will generate ionization in LDTD-APCI, since proton transfer in the PI mode mostly occurs between water cluster, H3O+(H2O)n where n = 0-4,8 and more basic analytes (higher proton affinity). Higher efficiency protonation in LDTD-APCI is accomplished in the presence of the smallest hydronium species because of their low proton affinity, i.e. H3O+ and (H2O)H3O+. Larger water clusters will form when the humidity (water saturation) in the ionization region is increased and negatively impact the analyte ionization process.9 The larger hydronium cluster, with higher proton affinity, will affect sensitivity towards low proton affinity molecules and decrease their fragmentation because of an excess of energy in protonation reactions.10 The main advantage of using APCI with the LDTD is that it is less susceptible to matrix effects and ionization suppression than with electrospray ionization.11
S-4
OH
O
H
H
H
H
H
H
HO
H
OH
H
HO
H
HO
Estrone (E1)
Estradiol (E2)
Estriol (E3)
OH
OH
OH
H
H
H
H
OH
H
H
H
H O
O
HO
Norethindrone (NOR)
Levonorgestrel (LEVO)
17-ethynylestradiol (EE2)
H
O O
OH H H
H
H
O
H
CH 3
Medroxyprogesterone (MPROG)
H
O
Progesterone (PROG)
Figure S-1. Molecular structures of selected steroid hormones and their acronyms.
S-5
Table S-1. MS/MS Parameters for the Analysis of Selected Steroid Hormones Analytes in Both Negative (NI) and Positive (PI) Ionization Mode. Compound E1 E2
NI
271 [M–H]‾
PI
255 [M-H2O+H]+
133 144 159
31 ± 3 21 ± 3 100
68 68 68
18 40 19
NI
273 [M–H]‾
145 147 185
57 ± 9 82 ± 12 100
-67 -67 -67
61 43 47
PI
257 [M-H2O+H]+
E3
NI
287 [M–H]‾
EE2
NI
295 [M–H]‾
135 146 161 143 171 143 145 183
52 ± 5 22 ± 1 100 67 ± 12 100 63 ± 12 100 58 ± 13
65 65 65 -97 -97 -82 -82 -82
17 37 16 54 40 70 45 44
PI
279 [M-H2O+H]+
133 159
100 69 ± 11
46 46
16 20
NI
297 [M–H]‾
143 145
72 ± 15 100
-74 -74
43 43
PI
281 [M-H2O+H]+
133 159
100 22 ± 6
48 48
23 22
LEVO
PI
313 [M+H]+
109 185 245
83 ± 19 45 ± 8 100
64 64 64
27 20 18
MPROG
PI
345 [M+H]+
97 123
26 ± 4 100
74 74
23 25
NOR
PI
299 [M+H]+
91 109
85 ± 16 100
69 69
35 26
PROG
PI
315 [M+H]+
E2-13C2
EE2-13C2
Precursor ion (m/z) 269 [M–H]‾
Product ion (m/z) 143 145 143 145 183
Relative intensity ratioa (%) 45 ± 6 100 51 ± 13 66 ± 20 100
Ionisation mode NI
TL (V) -84 -84 -74 -74 -74
CE (eV) 57 41 47 37 49
97 98 ± 17 65 20 109 100 65 24 a The most abundant product ion was used for quantification whereas the other product ions were used for to confirm the presence of the steroid hormones.
S-6
E1 E3
6.0E+04
EE2 E2
5.0E+04
Peak Area
4.0E+04
3.0E+04
2.0E+04
1.0E+04
0.0E+00 5
10
15
20
25
30
40
50
60
80
Laser Power (%)
Figure S-2. Negative ionization mode (NI) maximum peak intensity for a concentration of 2 mg/L was observed at a maximum 20% laser power for steroid hormones EE2 and E2 and was significantly different from other laser power (n=3; P 0.5). The optimum laser power for E1 was 20% and 25%, while for E3 it was 15% and 20%, which were not statistically different from each other in both cases (n=3; P 0.5). A laser power of 20% was chosen in NI mode, in order to have a single method applicable to all selected compounds and because it was common to all four selected steroid hormones.
S-7
E3 E1
1.2E+05
EE2 E2
1.0E+05
Peak Area
8.0E+04
6.0E+04
4.0E+04
2.0E+04
0.0E+00 1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
Gas Flow (L/min)
Figure S-3. Carrier gas flow was optimal between 2 to 3 L/min and gave significantly higher peak area response (n=3; P 0.5) in NI mode. A gas flow of 3 L/min provided the best signal-to-noise ratio combined with small signal variability (RSD 10%; n=3). Estriol (E3) was the only exception with an optimal gas flow of 2 L/min.
S-8
1.4E+06 1.2E+06
Peak Area
1.0E+06 8.0E+05 6.0E+05 4.0E+05 2.0E+05 0.0E+00
3 sec – 20%
1 sec – 20%
2 sec – 20%
3 sec – 20% 3 sec – 2% 1 sec
0.5 sec
0.5 sec
1 sec
1 sec
Laser Pattern
Figure S-4. Four different laser patterns (n=3) were used and compared for PROG in PI mode.
The optimal conditions giving the maximum peak area intensity and good
variability (RSD 10%) for desorption laser power and patter consisted of a 1 sec initial ramp from 0 to 20% and held for 3 sec at 20% before shutting off.
S-9
Matrix-free In matrix
160 140
Recovery (%)
120 100 80 60 40
PROG (PI)
NOR (PI)
MPROG (PI)
LEVO (PI)
EE2-13C2 (PI)
EE2-13C2 (NI)
EE2 (PI)
EE2 (NI)
E3 (NI)
E2 (PI)
E2 (NI)
0
E1 (NI)
20
Compound (ionisation mode)
Figure S-5. Recovery values (n=3) for matrix-free (pure water) and in matrix (spiked effluent wastewater) samples at 50 ng/L in both NI and PI modes using Strata-X SPE cartridges for all selected steroid hormones including IS and surrogate.
S-10
REFRENCES
(1) Beuhler, R. J.; Flanigan, E.; Greene, L. J.; Friedman, L. J. Am. Chem. Soc. 2002, 96 , 3990-3999. (2) Daves, G. D. Acc. Chem. Res. 1979, 12 , 359-365. (3) Picard, P.; Tremblay, P.; Paquin, E. R. Laser Diode Thermal Desorption Ionization Source (LDTD): Fundamental Aspects. 56th ASMS Conference on Mass Spectrometry, Denver, CO. 2008. Available at: http://www.ldtdionsource.com/eng/documentations/publications.asp (4) Harrison, H. G. Chemical Ionization Mass Spectrometry; CRC Press Inc: Boca Raton, FL, 1983.pp. 156. (5) Ma, Y. C.; Kim, H. Y. Journal of the American Society for Mass Spectrometry 1997, 8 , 1010-1020. (6) Geerdink, R. B.; Kooistra-Sijpersma, A.; Tiesnitsch, J.; Kienhuis, P. G. M.; Brinkman, U. A. T. J. Chromatogr. A 1999, 863 , 147-155. (7) Dams, R.; Benijts, T.; Duff, K.; Bell, D. Rapid Commun. Mass Spectrom. 2002, 16 , 1072-1077. (8) Zook, D. R.; Grimsrud, E. P. J. Phys. Chem. 1988, 92 , 6374-6378. (9) Sinha, V.; Custer, T. G.; Kluepfel, T.; Williams, J. Int. J. Mass Spectrom. 2009, 282 , 108-111.
S-11
(10) Tani, A.; Hayward, S.; Hansel, A.; Hewitt, C. N. Int. J. Mass Spectrom. 2004, 239 , 161-169. (11) Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Anal. Chem. 2003, 75 , 3019-3030.
S-12
