Analytical method for the determination of O,O-diethyl phosphate and O,O-diethyl thiophosphate in faecal samples

Article abstract of DOI:10.1016/S0045-6535(00)00020-5

A residue analytical method was developed for the determination of the dialkylphosphate metabolites of parathion in faecal samples obtained from rabbits. The faecal pieces were homogenised in water and highly water-soluble O,O-diethyl phosphate (DEP) and O,O-diethyl thiophosphate (DETP) were subsequently alkylated to pentafluorobenzyl esters by a phase transfer reaction. Derivatisation yields depend on the reaction time. The recovery rates were determined over the complete procedure using authentic reference standards in matrix solution. The reference standards allow to observe an effect of the sample matrix on the area of signals while GC-FPD is used. The recoveries over the concentration range from 0.05 to 5 μg/g were 47-62% for O,O-diethyl phosphate and 92-106% for O,O-diethyl thiophosphate potassium salt with FPD. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0045-6535(00)00020-5

Chemosphere 41 (2000) 1313±1320
Analytical method for the determination of O,O-diethyl
phosphate and O,O-diethyl thiophosphate in faecal samples
Detlef Schenke *
Federal Biological Research Centre for Agriculture and Forestry, Institute of Ecotoxicology in Plant Protection, Stahnsdorfer Damm 81,
1
4532 Kleinmachnow, Germany
Received 17 November 1999; accepted 20 January 2000
Abstract
A residue analytical method was developed for the determination of the dialkylphosphate metabolites of parathion
in faecal samples obtained from rabbits. The faecal pieces were homogenised in water and highly water-soluble O,O-
diethyl phosphate (DEP) and O,O-diethyl thiophosphate (DETP) were subsequently alkylated to penta¯uorobenzyl
esters by a phase transfer reaction. Derivatisation yields depend on the reaction time. The recovery rates were deter-
mined over the complete procedure using authentic reference standards in matrix solution. The reference standards
allow to observe an e€ect of the sample matrix on the area of signals while GC-FPD is used. The recoveries over the
concentration range from 0.05 to 5 lg/g were 47±62% for O,O-diethyl phosphate and 92±106% for O,O-diethyl thio-
phosphate potassium salt with FPD. Ó 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Dialkylphosphates; Faecal samples; GC-NPD; GC-FPD
1
. Introduction
developed with di€erent success since 1968. Table 1 gives
a chronological overview of the essential methods de-
scribed in the literature. The recovery rates given in
these papers are only relative amounts from forti®cation
reactions without usual primary, authentic reference
standards. Reference standards were used only for the
analysis of O,O-diethyl phosphate (DEP) and O,O-di-
ethyl thiophosphate (DETP) in spinach (Archer, 1974).
Weisskopf and Seiber (1989) estimated derivatisation
yields by comparing molar phosphorus response with
trimethyl phosphate standard, but the yield of dimethyl
phosphate (DMP) could not be determined due to in-
terferences. O,S,S-trimethyl phosphorothioate was only
an internal standard for gas chromatography but not for
the whole method (Drevenkar et al., 1994). Aprea et al.
(1996) shows recovery rates limited at the stage of pu-
ri®cation. In this study the penta¯uorbenzyl (PFBz) es-
ter of the parathion metabolites O,O-diethyl phosphate
and O,O-diethyl thiophosphate were synthesised,
isolated and characterised. An analytical method
Thiophosphorus and dithiophosphorus compounds
applied as insecticide have been suspected of causing
sublethal and lethal toxic e€ects in wildlife (Joermann
and Gemmeke, 1994). Exposure normally leads to the
inhibition of various esterases. This fact is used for as-
sessment of exposure, but it is also understood that the
inhibition of esterases is also in¯uenced by other factors
and environmental impacts (Mineau, 1991). Today the
real exposure of wildlife can be roughly estimated.
Examination of dialkylphosphate metabolites in
faeces originating from organophosphate insecticides
could be an alternative tool as noninvasive indicator of
exposure in the environment. Analysis of the highly
polar and water-soluble dialkylphosphates has been
*
Tel.: +33203-480; fax: +33203-48-200.
E-mail address: d.schenke@bba.de (D. Schenke).
0045-6535/00/$ - see front matter Ó 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 0 0 ) 0 0 0 2 0 - 5

1
314
D. Schenke / Chemosphere 41 (2000) 1313±1320

D. Schenke / Chemosphere 41 (2000) 1313±1320
1315

1
316
D. Schenke / Chemosphere 41 (2000) 1313±1320
containing a phase transfer reaction developed by
Bradway et al. (1981) was improved and tested with
faeces from rabbits.
bath. Three ml hexane was added to the sample, soni-
cated to redissolve the residue and transferred to the
puri®cation column. The mobile phase was hexane±ac-
etone (v/v ˆ 95/5). The ®rst 25 ml of eluant was dis-
carded. DEP-PFBz and DETP-PFBz were eluted with
the other 275 ml. The solvent was evaporated to dryness
as described above. The residue was re®lled with a cer-
tain volume hexane. Aliquots were transferred into
autosampler vials for quanti®cation by gas chromatog-
raphy with di€erent detectors. The vials were stored at
)20°C until measurement.
2
. Material
O,O-diethyl phosphate (DEP-K) and O,O-diethyl
thiophosphate (DETP-K), both as potassium salt with a
purity of about 98%, were produced by Studio Chimico
Dr. M. Giuntini (Firenze, Italy). The stock solutions
were prepared by dissolving 20 mg salt in 50 ml meth-
anol. The working solutions were diluted with methanol.
Penta¯uorobenzyl bromide (PFBzBr) was purchased
from Sigma.
3
.2. Gas chromatography
The analysis was performed using a gas chromato-
graph HP 5890 equipped with an NPD; injection port:
50°C, split/splitless injection (0.5 min splitless), 1 ll
The penta¯uorobenzyl ester of O,O-diethyl phos-
phate (DEP-PFBz) was synthesised in boiling acetonit-
rile for 3 h with one equivalent 18-Crown-6 and 2.5
equivalent potassium carbonate. Correspondingly O,O-
diethyl thiophosphate (DETP-PFBz) was prepared in
acetone with a magnetic stirrer at room temperature.
After the reactions the solvents were evaporated. Fifty
ml water were added to the oily residue. The suspensions
were extracted with hexane. The penta¯uorobenzyl es-
ters of DEP and DETP were obtained as colourless
liquids by vacuum distillation. The stock solutions were
made at 0.2 mg/ml in hexane. Working solutions in
hexane for calibration curves in the range of 50±1500 pg/
ll were prepared daily. Salts, esters and the stock solu-
tions were stored at )20°C.
2
injected volume; column: 30 m DB±Wax (J&W Scien-
ti®c) 0.252 mm i.d., ®lm thickness 0.25 lm; column
temperature program: 60°C (1 min) ± 20°C/min ± 150°C
(
15 min) ± 30°C/min ± 180°C (10 min) ± 30°C/min ±
250°C (10 min); ¯ow rates: nitrogen 2.4 ml/min, make up
24.1 ml/min, hydrogen 3.5 ml/min, air 78 ml/min; de-
tector: 280°C.
Gas chromatograph HP 5890 series II equipped with
an FPD and phosphorus ®lter (525 nm), injection port:
2
00°C, split/splitless injection (0.75 min splitless), 1 ll
injected volume; columns: 15 m SE-54 0.259 mm i.d.,
lm thickness 0.25 lm; column temperature program:
0°C (1 min) ± 5°C/min ± 250°C (5 min); ¯ow rates:
®
9
Tetrahexyl-ammonium hydrogen sulphate (THA)
helium 100 kPa, nitrogen 34 ml/min, hydrogen 103 ml/
min, air 73 ml/min; detector: 200°C. 30 m HP-5 0.32 mm
i.d., ®lm thickness 0.25 lm; column temperature pro-
gram: 90°C (1 min) ± 10°C/min ± 200°C ± 30°C/min ±
(
Fluka) solution was made with 4.5 g salt in 100 ml 0.2
N hydrous sodium hydroxide (Roth) solution. The pu-
ri®cation columns (15 mm i.d.) were packed with 5 g
Florisil (60±100 mesh, Fluka, activated at 560°C for 6 h
and conditioned with 10% water) slurred in hexane and
topped with 1 cm anhydrous sodium sulphate (Roth).
Solvents with the SupraSolv quality were obtained from
Merck.
3
3
2
00°C (1 min); ¯ow rates: helium 100 kPa, nitrogen
4 ml/min, hydrogen 103 ml/min, air 73 ml/min; detector:
00°C.
Mass spectral analysis was made using a GCQ,
Finnigan MAT. The electron impact ionisation mass
spectra were obtained at 70 eV, with mass range scanned
from 50 to 650 amu. A 30 m DB±XLB 0.25 mm i.d., ®lm
thickness 0.25 lm was used for con®rmation analysis.
3
. Analytical method
3
.1. Extraction and clean-up
Five g faeces (f.w.) was dissolved in 30 ml distilled
water and homogenised for 1 h on a rotary shaker
Heidolph). Eighty ml dichloromethane, 10 ml THA-
4
. Results and discussion
(
4
.1. Standard compounds
solution and 250 ll PFBzBr were added to this sus-
pension. For the derivatisation the Erlenmeyer ¯ask was
shaken only 10 min at room temperature. The solution
was ®ltrated through glass wool in a separation funnel.
The residue and the funnel were rinsed with 40 ml di-
chloromethane. The combined organic layer was dried
over anhydrous sodium sulphate and evaporated to
dryness with a rotary evaporator (B u chi) in a 30°C water
The authentication of gas chromatographic standard
compounds is summarised in Table 2. The fragmenta-
tion in the mass spectra and the shift and methylene
coupling constants in the NMR-spectra characterised
DETP-PFBz as the thiolate in accordance with the
results found by Li et al. (1991).

D. Schenke / Chemosphere 41 (2000) 1313±1320
1317
Table 2
Characterisation of gas chromatographic standards
a
DEP-PFBz
141
DETP-PFBz
145
Boiling point, bpꢀ8† (°C)
1
4 3
H-NMR (ppm) d(Me Si) (in CDCl )
±
±
±
OCH
OCH
XCH
2
2
2
CH
CH
3
1.34 (t, 6H, J ˆ 7 Hz)
4.14 (m, 4H, J ˆ 7 Hz)
5.14 (d, 2H, J ˆ 7.5 Hz)
1.36 (t, 6H, J ˆ 7 Hz)
4.18 (m, 4H, J ˆ 7 Hz)
4.09 (d, 2H, J ˆ 12 Hz)
3
±C
6
F
5
GC±MS EI (m/z) (in %)
O) P(O)XCH
H±XCH
H±CH
Important masses
334 (11%)
198 (20%)
(
C
2
H
5
2
2
C
6
F
5
350 (14%)
214 (5%)
2
C
6
F
5
2
C
6
F
5
182 (100%)
154 (53%)
125 (82%)
182 (40%)
170 (100%)
141 (17%)
(C
C
2
H
H
5
O) P(O)X±H
2
O)(HO)P(O)X
(
2
5
a
X ˆ O (DEP), S (DETP).
The purity of DEP-PFBz and DETP-PFBz was de-
termined by the peak areas in the gas chromatograms as
being about 95%.
pared to the other metabolites (Weisskopf and Seiber,
1989). So, derivatisation with penta¯uorobenzyl bro-
mide mentioned by Aprea et al. (1996) was found most
suitable. Bradway et al. (1981) developed a method with
a phase-transfer catalyst. Quaternary ammonium salt
extracts the anions of DEP and DETP into dichlorom-
ethane, where they react with PFBzBr (Rabinovitz et al.,
1986). The procedure is short versus the separate reac-
tions with di€erent temperature and puri®cations per-
formed by Aprea et al. (1996).
4
.2. Extraction
The properties of dialkylphosphates required con-
sideration of the pH-value (Lores and Bradway, 1977;
Weisskopf and Seiber, 1989) and of the variability of
the medium. Problems lie in the choice of the best ex-
traction solvents, the relation between solvent and
analysed matrix (Lores and Bradway, 1977; Richardson
and Seiber, 1993) and in the choice of additives to
produce a saturated solution (Daughton et al., 1976;
Weisskopf and Seiber, 1989). Drevenkar et al. (1994)
found the best recoveries by diluting plasma with de-
ionised water. In contrast to the samples with a high
water content (Table 1) faeces produced by rabbits are
a relatively compact matter. The pieces were shaken in
distilled water for 1 h to form a homogeneous pap so
that the dialkylphosphates were better available for
short derivatisation. The pH-value of the faeces±water
suspension was about 9±10. Both higher (pH ˆ 13) and
lower (pH ˆ 3) pH-values led to decreasing yields of
DEP-PFBz.
Firstly, the method was tested with the individual
extraction each of the dialkylphosphates from water
(Fig. 1). For DETP-K the highest rates were obtained at
4.3. Derivatisation
From the beginning of dealing with ultratrace de-
tection of the dialkyl phospates it was clear that a gas
chromatographic method would be successful. The in-
jector block derivatisation developed by Moody et al.
(
1985) and Weisskopf and Seiber (1988) could not be
realised with our equipment. Additionally, this derivat-
isation of DETP was strongly in¯uenced by the urine
concentration. Problems with the analytical quality for
DETP were assumed with a complex matrix like faeces.
The recoveries for DEP were exceptionally low com-
Fig. 1. Recovery rates for the derivatisation of DEP-K (A, h)
and DETP-K (± ± ±, s) and the transformation from DETP to
DEP (- - -, d) in water as a function of the reaction time mea-
sured with authentic reference standards (2.5 and 5 lg K-salts
in 10 ml water).

1
318
D. Schenke / Chemosphere 41 (2000) 1313±1320
1
0 min. The derivatisation carried out by Bradway et al.
1981) takes 25±30 min, more than the double time.
4.5. Recovery
(
With the authentic reference standard, it was shown that
in this way only 60% DETP-K can be recovered as
DETP-PFBz. The yield of the derivatisation of DEP-K
was hardly in¯uenced by the reaction time, with a
maximum amount of about 50% for 1 h. These form a
contrast to the recovery rates for DETP (80%) and DEP
The faecal samples were forti®ed with an appropriate
amount of both K-salts together in methanol and was
kept at room temperature for 3 h until the analysis
started with dilution. The recovery rates obtained are
more realistic than the forti®cation of diluted samples,
which suppresses the interaction of dialkylphosphates
with the sample (Drevenkar et al., 1994).
(
10%) obtained by Bardarov and Mitewa (1989) with
di€erent ion-pair reagents after 1 h. Fig. 1 suggests the
possibility of side-reactions to more than one alkylation
with PFBzBr. Surprisingly DEP-PFBz was observed to
develop from DETP-K. In accordance with the de-
creasing yields of DETP-PFBz, formation of DEP-PFBz
from DETP-K increased with longer reaction times. A
transformation has been known in the case of DMTP
during the reaction with PFBzBr but only at higher
temperatures (Reid and Watts, 1981). The analysis of
faeces was carried out with a derivatisation time of
Table 3 shows the recovery rates over the complete
procedure from control faeces forti®ed in the range from
0.05 to 5 lg/g and measured by a primary, authentic
reference standard. The results of recoveries listed in the
table were obtained over half a year as a routine check
during the analysis of a feeding study. So, they also re-
¯ect the robustness of the method. Neither of the hy-
drolytical metabolites was found when faecal samples
were spiked with parathion at the beginning of analyti-
cal procedure.
1
0 min.
It seems that an all-round method does not exist for
all six metabolites listed in Table 1. The recovery rates
of these methods were calculated with standards ob-
tained by the same analytical procedure with forti®ed
untreated samples or with forti®ed solvents. This is
basically not comparable with the results developed
using authentic reference standards, as described in this
paper (Table 3).
4
.4. Gas chromatography
Analytical results were veri®ed using two GC-systems
with di€erent capillary columns and detectors. The NPD
and FPD responses to injections ranging from 50 to
1
500 pg were linear for both esters. The minimal de-
The method from Bradway et al. (1981) is suitable to
determine the concentration of DETP and DEP as me-
tabolites of parathion. Observance of the reaction time
will enhance the validity of the results. This method was
adapted to faecal samples. In particular for DEP re-
covery rates at the low level of 0.05 lg/g with coecients
of variation below 20% were achieved. The decomposi-
tion of the polar metabolites in faeces was investigated
under ®eld condition over 4±8 days. The concentrations
for DEP were in the range between 1 and 0.2 lg/g and for
DETP from 2 to 0.5 lg/g (Schenke and Gemmeke, 1999).
tectable quantities (noise-signal-ratio:1/10) of DEP-
PFBz and DETP-PFBz were 25 pg/ll. Matrix e€ects
were assessed by comparing the peak area from the es-
ters spiked in untreated faecal samples before the in-
jection to that of the standards in hexane. Using NPD,
interferences from the matrix occurring in some un-
treated samples in¯uenced the results in the low con-
centration range (80 pg/ll), in particular for DEP-PFBz.
The more selective FPD leads to an increased response
of the analytes in the spiked matrix (c
the hexane standard (c ). The increased response can
be calculated over the described concentration range
M
) in relation to
S
4.6. Stability
b
S
with the equation c
M
ˆ ac (DEP-PFBz:a ˆ 2.3419;
2
b ˆ 0.9483;
r ˆ 0.9999/
DETP-PFBz:a ˆ 1.4151;
The stock solutions of the salts and of the esters were
prepared freshly after three months. The stability of the
2
b ˆ 0.9801; r ˆ 1).
Table 3
Average recoveries (REC) and coecients of variation (CV) of DEP and DETP from forti®ed rabbit faeces by the complete procedure
K-salt
Concentration (lg/g)
Number
FPD (%)
REC
NPD (%)
REC
CV
CV
DEP-K
0.05
7
8
4
62
47
51
13
11
11
79
68
47
68
14
7
0
5
.5
DETP-K
0.05
7
8
4
104
92
18
10
16
98
81
76
29
11
11
0
5
.5
106

D. Schenke / Chemosphere 41 (2000) 1313±1320
1319
ꢀ
Table 4
Stability of DEP-K and DETP-K in faeces at )20°C (calculated
with respect to recoveries (FPD))
Drevenkar, V., Stengl, B., Fr o be, Z., 1994. Microanalysis of
dialkylphosphorus metabolites of organo-phosphorus pes-
ticides in human blood by capillary gas chromatography
and by phosphorus-selective and ion trap detection. Anal.
Chem. Acta 290, 277±286.
Days after
forti®cation
DEP-K (%)
DETP-K (%)
Fenske, R.A., Lengwell, J.T., 1989. Method for the determi-
nation of dialkyl phospate metabolites in urine for studies of
human exposure to Malathion. J. Agric. Food Chem. 37,
0
±
79
±
2
2
3
3
20
91
14
40
89
86
95
93
97
995±998.
116
75
Joermann, G., Gemmeke, H., 1994. Meldungen u ber P¯an-
zenschutzmittelvergiftungen von Wirbeltieren. Nac-
hrichtenbl. Deut. P¯anzenschutzd. 46, 295±297.
Li, H.P., Wong, S.S., Li, G.C., 1991. The analysis of organo-
phosphate metabolites in human urine samples. Plant Prot.
Bull. 33, 188±196.
stock solution of DEP-K and DETP-K was shown by
consistent recoveries carried out over this time. The
stability of the esters tested with atrazine as references
showed no deterioration. Table 4 shows the stability of
DEP-K and DETP-K in untreated faeces samples stored
at )20°C over nearly 1 year.
Lores, E.M., Bradway, D.E., 1977. Extraction and recovery of
organophosphorus metabolites from urine using an anion
exchange resin. J. Agric. Food Chem. 25, 75±79.
Mineau, P. (ed.), 1991. Cholinesterase-inhibiting insecticides ±
their impact on wildlife and the environment. In: Chemicals
in Agriculture, vol. 2. Elsevier, Amsterdam.
Moody, R.P., Franklin, C.A., Riedel, D., Muir, N.I., Green-
halgh, R., Hladka, A., 1985. A new GC on-column
methylation procedure for analysis of DMTP (O,O-dimeth-
yl phosphorothioate) in urine of workers exposed to
Fenitrothion. J. Agric. Food Chem. 33, 464±467.
Acknowledgements
I cordially thank B. Kunde for her experimental as-
sistance. Thanks also to B. Pusch and also H. Nowak
and G. Reese-St a hler (both BBA ± Institute of Ecolog-
ical Chemistry, Berlin) for GC-measurements. I thank
Dr. Heydenreich, University Potsdam, for the NMR
spectra of reference substances.
Rabinovitz, M., Cohen, Y., Halpern, M., 1986. Hydroxide ion
initiated reaction under phase transfer catalysis conditions:
mechanism and implications. Angew. Chem. Int. Ed. Engl.
25, 960±970.
Reid, S.J., Watts, R.R., 1981. A method for the determination
of dialkyl phosphate residues in urine. J. Anal. Toxicol. 5,
126±132.
Richardson, E.R., Seiber, J.N., 1993. Gas chromatographic
determination of organophosphorus insecticides and their
dialkyl phosphate metabolites in liver and kidney samples.
J. Agric. Food Chem. 41, 416±422.
References
Aprea, C., Sciarra, G., Orsi, D., Boccalon, P., Sartorelli, P.,
Sartorelli, E., 1996. Urinary excretion of alkyl-phosphates
in the general population (Italy). Sci. Total Environ. 177,
Schenke, D., Gemmeke, H., 1999. Behaviour of O, O-diethyl-
phosphate and O, O-diethylthiophosphate in faecel samples
under ®eld conditions. In: Proceedings of the 218th ACS
National Meeting, Symposium: Pesticide and wildlife, New
Orleans. Book of Abstracts (1) Agro, p. 118.
3
7±41.
Archer, T.E., 1974. Dissipation of Parathion and related
compounds from ®eldsprayed spinach. J. Agric. Food
Chem. 22, 974±977.
Sha®k, T., Bradway, D.E., Enos, H.F., Yobs, A.R., 1973.
Human exposure to organophosphorus pesticides. A mod-
i®ed procedure for the gas±liquid chromatographic analysis
of alkyl phosphate metabolites in urine. J. Agric. Food
Chem. 21, 625±629.
Bardarov, V., Mitewa, M., 1989. High-performance liquid and
gas chromatography of dialkylphosphates, dialkylthiophos-
phates and dialkyldithiophosphates as their penta¯uorob-
enzyl derivatives. J. Chromatogr. 462, 233±241.
Blair, D., Roderick, H.R., 1976. An improved method for the
determination of urinary dimethyl phosphate. J. Agric.
Food Chem. 24, 1221±1223.
St. John Jr., L.E., Lisk, D.J., 1968. Rapid, sensitive residue
determination of organophosphorus insecticides by alkali
thermoionic gas chromarography of their methylated alkyl-
phosphate hydrolytic products. J. Agric. Food Chem. 16,
408±410.
Bradway, D.E., Moseman, R., May, R., 1981. Analysis of alkyl
phosphates by extractive alkylation. Bull. Environ. Contam.
Toxicol. 26, 520±523.
Takade, D.Y., Reynolds, J.M., Nelson, J.H., 1979. 1-(4-
Nitrobenzyl)-3-(4-tolyl)triazene as a derivatizing reagent
for the analysis of urinary dialkyl phosphate metabolites of
organophosphorus pesticides by gas chromatography. J.
Agric. Food Chem. 27, 746±753.
Daughton, C.G., Crosby, D.G., Garnas, R.L., Hsieh, D.P.H.,
1
976. Analysis of phosphorus-containing hydrolytic prod-
ucts of organophosphorus insecticides in water. J. Agric.
Food Chem. 24, 236±241.
Daughton, C.G., Cook, A.M., Alexander, M., 1979. Gas
chromatographic determination of phosphorus-containing
pesticide metabolites via benzylation. Anal. Chem. 51,
Weisskopf, C.P., Seiber, J.N., 1989. New approaches to analysis
of organophosphate metabolites in the urine of ®eld
workers. In: Wang, R.G.M., Franklin, C.A., Honeycutt,
R.C., Reinert, J.C. (Eds.), Biological monitoring for pesti-
1
949±1953.

1
320
D. Schenke / Chemosphere 41 (2000) 1313±1320
cide exposure: measurement, estimation and risk reduction.
ACS Symposium Series 382, Am. Chem. Soc., Washington,
DC, pp. 206±214.
Weisskopf, C.P., Seiber, J.N., Maizlish, N., Schenker, M., 1988.
Personnel exposure to diazinon in a supervised pest eradica-
tion program. Arch. Environ. Contam. Toxicol. 17, 201±212.