920 Acta Chim. Slov.

2009, 56, 920–926

Scientific paper

Modification of Method for the Determination

of Organochlorine Pesticides in Meat Samples
Gregor Arh,1,* Simona Fras,1 Toma` Polak,2 Bo`idar @lender,2
Marjan Veber1 and Matev` Pompe1
1
Faculty of Chemistry and Chemical Technology, University of Ljubljana, A{ker~eva 5, 1000 Ljubljana, Slovenia
2
Biotechnical Faculty, University of Ljubljana, Jamnikarjeva 101, 1000 Ljubljana, Slovenia

* Corresponding author: E-mail: gregor.arh@fkkt.uni-lj.si;

Fax: ++386 1 2419 220

Received: 01-04-2009

Abstract
In the paper the modification of standard method SIST EN 1528 1-4: 1998: Method D for measurements of organochlo-
rine pesticides (OCPs) in food products is presented. The modifications were made in the extraction step in which cold
extraction technique was replaced by Soxhlet extraction. For the clean-up step smaller columns were introduced and for
the concentration step a rotary evaporation was replaced by Kuderna-Danish concentration technique. Introduced modi-
fications improved the efficiency of the procedure for samples with high fat content. Recovery values for all analyzed
pesticides were over 60% and the reproducibility expressed as relative standard deviation was in the range of 10%. The
method is suitable for the determination of OCPs in meat products with high content of fat from low ppb concentration
range onward. The limits of detection for examined OCPs were in the range from 0.1 ppb to 2 ppb for lindane and α-en-
dosulfan, respectively.

Keywords: Organochlorine pesticides; soxhlet extraction; florisil; ECD; meat

Organochlorine pesticides are known to be very re-

1. Introduction sistant in the environment. Their lipophilic nature is the
Pesticides present a great danger for human health reason for their concentration and bioaccumulation in the
and significantly influence the environment owing to food chain, therefore they can be found in foodstuffs of
their extensive and uncontrolled use. The negative im- animal origin, mostly in meat and tissues that contain fat,
pact on human organism depends on concentration, sus- in milk and dairy products, eggs and fish. It was ascertai-
ceptibility to degradation, stability in the environment, ned that because of their long term usage they have accu-
possibility of bioaccumulation and bioconcentration, mulated in the environment and have reached critical con-
ability of insertion into the human food chain etc. Basic centrations even in regions where they have not been pro-
classes of pesticides are insecticides, herbicides, rodenti- duced or used for longer periods. Because of these negati-
cides, fungicides and fumigants. Regarding their chemi- ve effects their application was abandoned in many coun-
cal structure, pesticides can be divided into organop- tries.2,3 The Stockholm convention is a global treaty to
hosphorus pesticides (OPP), organochlorine pesticides protect human health and the environment from persistent
(OCP), carbamates, and pyrethroids. The chlorine-con- organic pollutants (POPs). In implementing the conven-
taining pesticides include dichloro-diphenyl-trichloroet- tion, governments take measurements to eliminate or re-
hane (DDT) and metabolites, hexachlorocyclohexane duce the release of POPs into the environment.4
(HCH) isomers, hexachlorobenzene (HCB), aldrin, diel- There are several procedures developed for the de-
drin, endrine, heptachlor and heptachlor epoxide. Their termination of pesticide residues in matrices such as vege-
classification as organochlorine pesticides is based on tables, honey, beer, baby food, and meat.5–8 However,
their common chemical properties and similar impact on their analysis in fatty matrices such as meat, represents
the environment.1 much more difficult analytical tasks due to problems of fat

Arh et al.: Modification of Method for the Determination of ...

 Acta Chim. Slov. 2009, 56, 920–926 921

removal. Matsumoto et al.9 investigated residues of poly- graphy was chosen because it is suitable for the determi-
chlorinated biphenyls (PCBs) and organochlorine pestici- nation of 20 OCPs, 6 OPPs and PCBs in fatty foodstuffs.
des in different meats (beef, pork and poultry) and in pro- Each part of this method was thoroughly examined and
cessed meat products in Osaka, Japan. They indicated re- improvements were made in extraction, clean-up and con-
ducing trend in the concentrations of poly chlorinated bip- centration steps.
henyls, HCH isomers, DDT analogues and dieldrin in all
meats during the past 35 years. However, the residual va-
lues of organochlorine (chlorinated) pesticides in proces- 2. Experimental
sed meat products remained at the same level for the past
15 years. It was also determined that concentration of 2. 1. Chemicals and Reagents
each individual pesticide is lower in the processed meat Pesticide standards of lindane, α-endosulfan, 4,4’-
products in comparison to raw meat.9 Lazaro et al.10 DDE, aldrin, dieldrin and HCB were purchased from
analyzed different meals of the average diet consumed in PolyScience (Niels, IL, USA) and Serve (Hiedelberg,
Aragon (Spain) for organochlorine pesticide residues. Of Germany) and their purities were > 99%. The pesticide
the 21 OCPs only HCB, lindane, 4,4’-DDD, 4,4’-DDE, mixture EPA CLP/625 (aldrin, a-BHC, b-BHC, d-BHC,
4,4’-DDT and β-endosulfan were detected in samples and g-BHC, 4,4’-DDD, 4,4’-DDE, 4,4’-DDT, dieldrin, endo-
their levels were below the limits set by current European sulfan I, endosulfan II, endosulfan sulfate, endrin, endrin
regulations.10 Frenich et al.3 optimized and validated the aldehyde, heptachlor, heptachlor epoxide) was purchased
method for simultaneous determination of residues of from Supelco and was of analytical grade. Organic sol-
OCPs and OPPs in chicken, pork and lamb meat samples. vents (acetone, n-hexane, petroleum benzene, diethyl et-
Extraction was carried out by a Polytron mixer, gel per- her, dichloromethane) were purchased from Merck (Ger-
meation chromatography was applied for the clean-up step many) and were of HPLC grade. Florisil and anhydrous
and the final determination was performed by gas chroma- Na2SO4 (purities in both cases > 99%) were purchased
tography coupled to a triple quadrupole mass spectrometry from Fluka (Germany) and Merck (Germany), respecti-
detection system.3 Juhler optimized the method for the de- vely. Partially deactivated Florisil was prepared with hea-
termination of OPPs in meat and fatty matrices.11 ting to 550 °C and left overnight. After cooling it was sto-
The European Committee for Standardization re- red in a sealed container. Before usage it was heated for 5
commends several methods for the determination of pesti- hours (130 °C) and then MilliQ water (Molsheim, France)
cide residues and PCBs in fatty foodstuffs. They proposed was added to obtain 3% final mixture.8 Stock standard so-
eight methods (from A to H) and each of them is suited lutions were prepared by weighing appropriate amounts
for determination and quantification of different sets of of standard substances, which were dissolved in 50 mL of
OCPs and OPPs in fatty foods. All methods consist of four n-hexane. Concentrations of stock solutions were 500 μg
steps: extraction, clean-up, identification and quantifica- mL–1. Liquid pesticide mixture was dissolved in 25 mL of
tion.12 Several approaches can be applied in the extraction n-hexane. All standard solutions were kept in the refrige-
step. The most frequently used techniques are cold extrac- rator at 4 °C. Nitrogen (99.996%, Messer, Austria) was
tion (recommended by the standard Method D), Soxhlet used for GC analysis and solvent evaporation.
extraction,13 solid-phase extraction,14,15 sonification16 and The glassware used for analysis was thoroughly
more recently supercritical fluid extraction,17 microwave washed: first with detergent and rinsed with Milli-Q water
assisted extraction,13 fluidized-bed extraction18 and acce- and ethanol. After drying it was rinsed with acetone, dried
lerated solvent extraction.19 An important step in the up and again rinsed with n-hexane.
analytical procedure is clean-up of the extracts. In this
step interferences should be eliminated and the analyte is
2. 2. Extraction
prepared for chromatographic analysis. Several techni-
ques such as gel permeation chromatography, adsorption Meat samples and their products were stored at –31
chromatography on different sorbents (Florisil – recom- °C until analysis. Approximately 100 g of meat sample
mended by the standard Method D, silica etc.) have been was taken and chopped with a food chopper. Ten grams of
applied for sample clean-up prior to the chromatographic chopped homogenized sample were weighed into a glass
detection.11,20–24 For the detection of organochlorine pesti- beaker and 10 g of anhydrous Na2SO4 were added. When
cides usually gas chromatography with either electron spiking samples, 1.0 mL of standard solution (concentra-
capture detector or mass selective detector is proposed.25 tion range from 1 to 10 μg mL–1) was added in this phase.
Among newly established techniques the application of The sample was transfered on a filter paper (type 388,
capillary electrophoresis is described.26 Sartorius, Germany) which was folded and put into a cel-
The aim of our work was to modify the standard lulose thimble. On top of the thimble a piece of cotton
SIST EN 1528 1-4: 1998: Method D in order to improve wool soaked with n-hexane, which was used for cleaning
recovery values and reliability of the results. From eight (collecting remaining parts of sample) the glass beaker,
proposed methods,12 method D based on gas chromato- used for sample weighing, was placed. The thimble was

Arh et al.: Modification of Method for the Determination of ...

922 Acta Chim. Slov. 2009, 56, 920–926

placed into a Soxhlet equipment assembled with a con- D for the determination of OCPs in fatty foodstuffs sho-
denser on top and a round bottom flask that contained wed several weaknesses. Very low recovery values (be-
100 mL of n-hexane and perforated glass boiling beads on low 10%) for all tested spiked samples were observed,
the bottom. The sample was allowed to reflux for at least which influenced the precision and accuracy of the meth-
18 hours on a water bath (95 °C), then it was cooled down od. Since the standard method is based on a cold extrac-
and a part of the solvent was evaporated on a water bath. tion technique which can be the reason for the low reco-
veries, firstly this part of the procedure was taken under
investigation. Cold extraction was replaced by Soxhlet
2. 3. Florisil Clean-up
extraction. To specify other sources of possible errors, al-
To remove matrix compounds, the clean-up proce- so all analytical steps were examined. Our aim was to
dure was performed on a glass column that was filled with modify an existing standard method with incorporating
10 mL of n-hexane. Then 3 g of Florisil (an activated other well-established concentration techniques in order
magnesium silicate) were slowly added. After approxima- to lower solvent consumption, gain higher extraction
tely 15 minutes, when Florisil was settled, the sample so- yields, obtain better reproducibility and lower limit of de-
lution was introduced into the column. The level of n-he- tection (LOD).
xane was adjusted (redundant n-hexane was washed
through the column) so that it was 1 cm above Florisil.
3. 1. Chromatography
The sample was slowly and quantitatively transferred into
a column. Elution was carried out at flow rate of 5 mL The standard procedure allows a wide range of dif-
min–1 with 30 mL of n-hexane and dichloromethane mix- ferent separation columns regarding the selection of sta-
ture (w/w 4:1). The resulting samples were collected in tionary phases as well as column dimensions (column
glass beakers. length, internal diameter and stationary phase thickness)
to be used. That is why no universal temperature program
2. 4. Kuderna-Danish Sample Concentration
The eluent of the sample from Florisil clean-up step
(around 35 mL) was transferred with a boiling chip into a
flask and receiving vessel of a Kuderna-Danish concen-
trator. A Snyder column was put on top and the concen-
trator was submerged (so that almost the whole receiving
vessel was under water) into a water bath with temperatu-
re close to the boiling point of water. Samples were con-
centrated to 1 mL and n-hexane was added to adjust the
final volume to 2 mL. The concentrated sample was then
transferred into a glass vial and kept in a refrigerator un-
til GC analysis.

2. 5. Chromatographic Analysis
Chromatographic measurements were performed
using a HP 6890 gas chromatograph (Hewlett-Packard,
CA, USA) equipped with 63Ni electron capture detector
(ECD). The RTX-5MS column with dimensions 60 m ×
0.25 mm × 0.5 μm was purchased from Restek (USA).
For quantification an HP Chemstation software (Rev. A.
05. 04) was used. Before analysis 1 μL of n-hexane was
injected three times to stabilize the response of the de-
tector. Then 1 μL of sample was manually injected three
times and the result was the mean value of three replica-
tes. Figure 1. Segment of a chromatogram for separation of OCPs.
Peaks from left to right correspond to: α-lindane (10.063 min); β-
lindane (10.529 min); γ-lindane (10.935 min); heptachlor
3. Results and Discussion (11.544 min); aldrin (12.151 min); heptachlor epoxide isomer B
(12.796 min); 4,4’-DDE (13.495 min); dieldrin (13.586 min); 4,4-
DDD (14.006 min); endrin (14.514 min); α-endosulfan
Our preliminary results based on the application of (14.722 min); endrin aldehyde (15.125 min) and 4,4’-
the recommended standard EN 1528 1-4: 1998: Method DDT (15.652 min).

Arh et al.: Modification of Method for the Determination of ...

 Acta Chim. Slov. 2009, 56, 920–926 923

is suggested within these standard documents. The tempe- tes to low volumes and in addition, analyte losses or prob-
rature programs found in the literature13,17 using similar lems related to the contamination are more pronounced in
separation columns did not give satisfactory separation of comparison to other two above mentioned techniques.
4,4’-DDE and dieldrin. Both literature procedures used When low-volatility solvent xylene was added to serve as
shorter separation columns. In general longer separation a keeper, the recovery values were within the experimen-
column would enable better separation, however one tal uncertainty of previous measurements. However the
should not forget that we have used just equivalent statio- presence of the keeper hindered the subsequent concentra-
nary phase. It was already shown that equivalent separa- tion of the extract, therefore it was excluded in further ex-
tion columns can give different resolution for the particu- periments. From Table 1 we can see that satisfactory reco-
lar pair of chromatographic peaks, even more they can very values were obtained only for concentrations greater
even cause the change in the elution order of different than 0.2 μg mL–1. At lower concentrations the contamina-
chromatographic peaks.27 Therefore the temperature pro- tion problems are evident. For all examined concentra-
gram was optimized for our experimental configuration. tions the relative standard deviation for three separate runs
Satisfactory separations of OCPs under investigation were was in the range of 20%.
obtained with the following parameters: temperature of The concentration efficiency when using nitrogen
injector 250 °C, temperature of detector 320 °C, tempera- purging (not shown) gave overall lower recovery values
ture program for column: initial temperature 70 °C, hea- (around 50%), however the reproducibility throughout the
ting with rate of 30 °C min–1 to 250 °C, with rate of 5 °C concentration range in comparison to rotary evaporation
min–1 to 270 °C and with rate of 10 °C to 300 °C (hold ti- was improved. The relatively low recovery values can be
me 5 min). From Figure 1 it is evident that under such attributed to losses caused by formation of an aerosol.
conditions, compounds are well separated, which enables The results obtained with Kuderna-Danish concen-
their reliable quantitative determination. At selected con- trators showed the best results regarding reproducibility as
ditions the precision of measurements for all compounds well as recovery values. Table 2 represents average reco-
expressed as relative standard deviation was below 10% very values for lindane, α-endosulfan, 4,4’-DDE and diel-
for manual injections of 1 μL of standard solution with drin throughout the concentration range from 0.2 to
concentration of 1 μg mL–1. For most of the of investiga- 1.0 μg mL–1 (in case of lindane from 0.02 to 0.1 μg mL–1)
ted compounds the linearity was obtained in concentration with corresponding RSD values. In comparison with ro-
range from 0.08 to 7 μg mL–1 (lindane: y = 15538 x – tary evaporation and nitrogen purging, Kuderna-Danish
163.4, R2 = 0.993, n = 7, α-endosulfan: y = 13733 x + concentration technique yields higher recovery values and
20.5, R2 = 0.995, n = 7, dieldrin: y = 14869 x – 388.5, R2 better reproducibility (RSD below 10%). This technique
= 0.990, n = 7). For 4,4’-DDE linear range was even broa- also provides better pre-concentration rates, since it is
der, that is from 0.08 to 10 μg mL–1 (y = 8370 x – 1721.9, possible to reduce final sample volume to 1 mL or less.
R2 = 0.998, n = 8).
3. 3. Extraction Step
3. 2. Concentration Step
The cold extraction technique (as proposed in the
For concentration of extracts, besides rotary evapo- standard method) was replaced by a Soxhlet extraction.
ration, other procedures were tested: purging with nitro- For the study of this step standard mixture of lindane, α-
gen and the application of Kuderna-Danish concentrators. endosulfane, dieldrin and 4,4’-DDE with concentration of
For this purpose standard solutions of lindane, α-endosul- 1 μg mL–1 was used (1 mL of mixture was added to a filter
fane, dieldrin and 4,4’-DDE in concentration range from paper, which was then put in a cellulose thimble). Diffe-
0.2 to 1.0 μg mL–1 (for lindane 10 times lower) were cho- rent extraction times were investigated. Initially we let the
sen. In all cases 35 mL of initial solution of standards we- sample to reflux for 3 hours. The yielded recoveries were
re concentrated to 5 mL. very low (around 5%) so the time was prolonged to 5
For rotary evaporation concentrating procedure se- hours and further to 10 hours, 18 hours and 24 hours. Re-
veral disadvantages were observed. Due to large apparatus covery values increased with the prolonged time and the
surfaces it is very difficult to efficiently concentrate analy- maximum was achieved after approximately 18 hours. Se-

Table 1. Average recovery values for lindane, α-endosulfan, 4,4’-DDE and dieldrin after evaporating the
solvent with rotary evaporator and corresponding RSD values (n = 3) in brackets.

Concentration η (%) η (%) η (%) Concentration η (%)

μg mL–1 )
(μ α-endosulfan 4,4’-DDE dieldrin μg mL–1 )
n (μ lindane
0.2 113 (21%) 119 (18%) 110 (17%) 0.02 118 (23%)
0.6 54 (16%) 47 (17%) 67 (14%) 0.06 72 (17%)
1.0 59 (18%) 50 (14%) 62 (15%) 0.1 64 (18%)

Arh et al.: Modification of Method for the Determination of ...

924 Acta Chim. Slov. 2009, 56, 920–926

Table 2. Average recovery values for lindane, α-endosulfan, 4,4’-DDE and dieldrin after evaporating the
solvent with Kuderna-Danish concentrator and corresponding RSD values (n = 3) in brackets.

Concentration η (%) η (%) η (%) Concentration η (%)

μg mL–1 )
(μ α-endosulfan 4,4’-DDE dieldrin μg mL–1 )
(μ lindane
0.2 82 (4%) 74 (8%) 80 (6%) 0.02 74 (9%)
0.6 69 (5%) 66 (6%) 68 (7%) 0.06 60 (6%)
1.0 87 (5%) 79 (4%) 84 (9%) 0.10 78 (8%)

veral different solvents with different polarities were also Danish concentration technique was even slightly impro-
tested: n-hexane, mixture of n-hexane and acetone (1:1) ved. With the reduction of the column size we also mana-
and diethyl ether. Among all tested solvents, n-hexane (as ged to lower the amounts of sample, Florisil and organic
the standard method suggests) showed best results. While solvents while at the same time the efficiency of Florisil
the use of n-hexane and acetone mixture (1:1) resulted in clean-up step was improved.
higher background of the chromatogram and slightly lower
peak areas, the diethyl ether gave significantly lower reco-
3. 5. Application of Method
very values. Further different temperatures and solvent vo-
lumes were also tested. For the extraction 200 mL round for Meat Samples
bottom flasks were used. At least 100 mL of n-hexane The developed analytical procedure as described in
should be used to prevent drying of the bottom part of flask the experimental part and shortly summarized in the para-
during cycles. When performing extractions at 70 and 80 graph below was critically evaluated regarding reproduci-
°C respectively the reflux was not satisfactorily, therefore bility and recovery values. The results for optimization of
the temperature was increased almost to the temperature of different steps of procedure (concentration, extraction and
boiling point of water (water bath was set to 95 °C). At tho- clean-up step) for the determination of OCPs in fatty
se conditions the best reflux was achieved. foodstuffs were initially focused on lindane, α-endosul-
fan, 4,4’-DDE and dieldrin because they represent a wide
range of polarities from non polar lindane to the more po-
3. 4. Clean-up
lar α-endosulfan with 4,4’-DDE and dieldrin in the midd-
For a clean-up agent, Florisil was applied, which is le range. After the optimization of individual parts using
also recommended by the standard method. Using stan- the standard mixtures without matrix, the same procedu-
dard mixture of lindane, α-endosulfane, dieldrin and 4,4’- res were tested with spiked samples of meat and different
DDE (1 mL of mixture with concentration of 1 μg mL–1 meat products (bacon, sausages, lean stag meat). Other or-
was added to 5 mL of n-hexane and elution was perfor- ganochlorine compounds present in the EPA CLP/625
med with 30 mL of eluting mixture) we have shown that standard mixture were also evaluated.
Florisil does not affect the overall losses of examined sub- Approximately 10 g of homogenized meat was
stances. However it has to be mentioned that it is very im- transferred into a glass beaker and approximately 10 g of
portant to prepare Florisil by partial deactivation before anhydrous Na2SO4 were added. Samples were spiked with
usage (this process is described in the experimental sec- 1 mL of EPA CLP/625 standard mixture with concentra-
tion). Chromatographic peaks of analytes obtained with tion of 1.0 μg mL–1. Soxhlet extraction with 100 mL of n-
clean-up on partially deactivated, compared to uncondi- hexane was performed for 18 hours. Sample clean-up pro-
tioned Florisil, are higher. This indicates possible losses cedure on Florisil was applied and the samples were con-
due to the adsorption of analytes on the unconditioned centrated using Kuderna-Danish concentrators to 2 mL.
Florisil. 1 μL of sample was injected into the gas chromatograph.
For the elution several compositions of n-hexane : Recovery values for all examined OCPs were over 60%
dichloromethane mixture: (w/w) 1:1, 1:2, 1:3 and 1:4 we- and the results presented in Table 3 are an example for
re tested. No improvement over the 1:4 (w/w) elution mix- analysis of lean stag meat. Reproducibilty of our modified
ture (as suggested in the standard method) was observed. method was tested by analyzing meat samples on different
One of the aims was also to reduce the amounts of days using different operators. In Table 3 results of mea-
organic solvents used in the analytical procedure. The ob- surements where individual runs were performed more
ject was to reduce the amount of n-hexane and dichloro- than 1 month apart are presented. The reproducibility
methane in the clean-up step without deteriorating the ef- among 9 different runs is satisfactory – the average error
ficiency of the procedure. Regarding the standard method of 9 different parallels among all tested analytes is around
approximately 10 times smaller columns were tested 10% (RSD). Similar results were also obtained for analy-
(amounts of sample, Florisil and solvents were appropria- sis of bacon and sausages. Compared to the standard SIST
tely adjusted). It was ascertained that in case of smaller EN 1528 1-4: 1998: Method D, where recovery values of
columns the efficiency of the clean-up step with Kuderna- all examined OCPs were below 10%, substantial improve-

Arh et al.: Modification of Method for the Determination of ...

 Acta Chim. Slov. 2009, 56, 920–926 925

Table 3. Recovery values (%) of α-lindane; β-lindane; γ-lindane; heptachlor; aldrin; heptachlor epoxide
isomer B; 4,4’-DDE; dieldrin; 4,4-DDD; endrin and α-endosulfan for sausage and corresponding RSD
values.

RSD
Analyte η11 η21 η32 η42 η52 η62 η73 η83 η93
(%)
α-lindane 67 69 71 79 84 84 71 66 72 9
β-lindane 70 67 68 77 81 81 70 68 72 8
γ-lindane 58 58 61 66 75 73 67 62 67 9
heptachlor 65 66 69 74 82 79 67 63 69 9
aldrin 64 67 69 74 79 76 76 70 75 7
heptachlor epoxide 59 64 63 70 75 74 73 69 73 8
isomer B
4,4’-DDE 66 67 65 67 71 69 80 79 82 9
dieldrin 60 66 64 70 75 76 74 68 73 8
4,4-DDD 69 80 88 93 95 99 79 77 83 11
endrin 55 59 55 63 71 73 68 64 69 10
α-endosulfan 58 62 59 67 74 78 64 60 64 11
1 2 3
Date of analysis: 1st February 2008, Date of analysis: 14th February 2008, Date of analysis: 5th
March 2008

ment of the method is evident. The main weakness of the sen because of its better characteristics compared to other
standard method is low efficiency of the cold extraction. tested solvents, such as a mixture of n-hexane/acetone
When Soxhlet extraction is applied, where the sample is (1:1) or diethyl ether. The time of extraction was prolon-
allowed to reflux at an elevated temperature (95 °C), the ged to 18 hours and the temperature of water bath during
distribution between solvent and analytes is more effecti- extraction was held at 95 °C. For the clean-up step the use
ve, resulting in higher recovery values. of smaller columns was proposed to reduce the amounts
With the proposed procedure better recovery, better of organic solvents, Florisil and sample weight. The elu-
reproducibility and lower limits of detection (LOD) in tion was carried out at an approximate rate of 5 mL min–1
comparison to the standard method were obtained. Calcu- with 30 mL of n-hexane and dichloromethane mixture
lations of LODs are based on the results of measurements (w/w 4:1). Partially deactivated Florisil has to be freshly
of the complete analytical procedure with a blank sample prepared before analysis. For sample concentration the
(only required solvents without matrix were used). LOD use of Kuderna-Danish concentrators was proposed. Also
values are expressed as a three times standard deviation rotary evaporation (as the standard method suggests) and
(3σ) of blank extract. The lowest LOD was found for lin- nitrogen purging were tested but in this case results did
dane (0.1 ppb) and the highest for α-endosulfan (2 ppb). not meet our expectations. It was shown that with use of
Results show that the method is suitable for the determi- Kuderna-Danish concentrators the reproducibility and re-
nation of wide spectrum of OCPs. Experiments with un- covery values are improved. The parameters for GC-ECD
spiked meat samples showed that concentrations of OCPs determination of OCPs are following: temperature of in-
were approximately in the 10–70 ppb (ng per kg of fat) jector 250 °C, temperature of detector 320 °C and tempe-
concentration range. From those results it is evident that rature program for column: initial temperature 70 °C, hea-
for all analyzed meat products concentrations of exami- ting with rate of 30 °C min–1 to 250 °C, with rate of 5 °C
ned OCPs were below the recommended maximum resi- min–1 to 270 °C and with rate of 10 °C to 300 °C (hold ti-
due levels (MRLs), which are legislated from country to me 5 min).
country differently (e.g. for UK and Germany 1 mg per kg The obtained recovery values of OCPs were from 60
of fat (1ppm) and for France 0.2 mg per kg of fat). to 90% and the reproducibility expressed as relative stan-
dard deviation was in the range of 10%. The limits of de-
tection for examined OCPs were in the range from
4. Conclusions 0.1 ppb to 2 ppb for lindane and α-endosulfan respecti-
vely.
The standard procedure SIST EN 1528 1-4: 1998:
Method D was modified in order to make it suitable for
the determination of organochlorine compounds in food 5. Acknowledgments
samples with high fat content. Extraction of OCPs from
the food matrix was thoroughly optimized and we propo- The authors acknowledge the financial support by
sed the use of a Soxhlet instead of a cold extraction tech- the European Commission through the project TRUE-
nique. Among different tested solvents n-hexane was cho- FOOD (Contract no. FOOD-CT-2006-016264). TRUE-

Arh et al.: Modification of Method for the Determination of ...

926 Acta Chim. Slov. 2009, 56, 920–926

FOOD – “Traditional United Europe Food” is an Integra- 12. Fatty food – Determination of pesticides and polychlorinated
ted Project financed by the European Commission under biphenyls (PCBs) – Part 1, 2, 3, 4, Nov 1996, European
the 6th Framework Program for RTD. Committee for Standardisation.
The information in this document reflects only the 13. M. Gfrerer, E. Lankmayr, Anal. Chim. Acta 2005, 533,
authors views and the Community is not liable for any use 203–211.
that may be made of the information contained therein. 14. J. Cook, M. P. Beckett, B. Reliford, W. Hammock, M. Engel,
J. AOAC Int. 1999, 82, 1419–1435.
15. T. Ligor, W. Kebrowski, B. Buszewski, Pol J Environ Stud
6. References 2007, 16 (4), 571–578.
16. C. M. Lino, M. I. Noronha de Silveira, J. Chromatogr. A
1. D. Kipcic, J. Vukusic, B. Sebecic, Food Technol. Biotechnol. 1997, 769, 275–283.
2002, 40 (1), 39–47. 17. C. Gonçalves, J. J. Carvalho, M. A. Azenha, M. F. Alpendu-
2. D. Muir, E. Sverko, Anal. Bioanal. Chem. 2006, 386, rada, J. Chromatogr. A 2006, 1110, 6–14.
769–789. 18. M. Gfrerer, C. Fernandes, E. Lankmayr, Chromatographia
3. A. Garrido Frenich, J. L. Martinez Vidal, A. D. Cruz Sicilia, 2004, 60 (11–12), 681–686.
M. J. Gonzalez Rodriguez, P. Plaza Bolaños, Anal. Chim.Ac- 19. X. H. Yi, Y. T. Lu, J. AOAC Int. 2005, 88 (3), 729–735.
ta 2006, 558, 42–52. 20. V. Leoni, A. M. Caricchia, S. Chiavarini, J. AOAC Int. 1992,
4. http://www.pops.int/ (accessed on June 2007). 75 (3), 511–518.
5. J. W. Wong, M. G. Webster, D. Z. Bezabeh, M. J. Hengel, K. 21. A. M. Gillespie, S. L. Daly, D. M. Gilvydis, F. Schneider, S.
K. Ngim, A. J. Krynitsky, S. E. Ebeler, J. Agric. Food Chem. M. Walters, J. AOAC Int. 1992, 78 (2), 431–437.
2004, 52, 6361–6372. 22. F. Hernandez, R. Serrano, J. Beltran, F. J. Lopez, J. AOAC
6. A. Garrido-Frenich, M. J. Gonzalez-Rodriguez, F. J. Arrebo- Int. 1996, 79 (1), 123–131.
la, J. L. Martinez-Vidal, Anal. Chem. 2005, 77, 4640–4648. 23. A. Hussen, R. Westbom, N. Megersa, L. Mathiasson, E.
7. Y. Tahboub, M. F. Zaater, T. A. Barri, Anal. Chim. Acta 2006, Bjorkuland, J. Chromatogr. A 2007, 1152 (1–2), 247–253.
558, 62–68. 24. R. Doong, C. Lee, Analyst 1999, 124, 1287–1289.
8. K. Patel, R. J. Fussell, M. Hetmanski, D. M. Goodall, B. J. 25. F. J. Santos, M. T. Galceran, TRAC 2002, 21 (9–10),
Keely, J. Chromatogr. A 2005, 1068, 289–296. 672–685.
9. H. Matsumoto, K. Kuwabara, Y. Murakami, H. Murata, Food 26. A. K. Malik, W. Faubel, Crit. Rev. Anal. Chem. 2001, 31 (3),
Hyg. Soc. Jpn. 2006, 47 (3), 127–135. 223–279.
10. R. Lazaro, A. Herrera, A. Ariño, M. P. Conchello, S. Bayarri, 27. P. Korytár, A. Covaci, J. de Boer, A. Gelbin, A. Th. Brink-
J. Agric. Food Chem. 1996, 44, 2742–2747. mann, J. Chromatogr. A 2005, 1065 (2), 239–249.
11. R. K. Juhler, J. Chromatogr. A 1997, 786, 145–153.

Povzetek
V ~lanku je opisana prilagojena standardna metoda SIST EN 1528 1-4: 1998: Metoda D, ki je primerna za dolo~evanje
organokloriranih pesticidov (OCP) v mesnih prehrambenih izdelkih. Glavne spremembe so bile napravljene v stopnji
ekstrakcije, kjer smo hladno ekstrakcijo nadomestili s Soxhlet ekstrakcijo. Za ~i{~enje vzorcev predlagamo uporabo
manj{ih kolon, za koncentriranje analitov pa smo namesto tehnike izparevanja topila z rotavaporjem uporabili Kuderna-
Danish koncentratorje. Vpeljane spremembe so pripomogle k izbolj{anju izkoristkov za vzorce z visoko vsebnostjo
ma{~ob, ki so bili za vse analizirane pesticide nad 60 %. Natan~nost postopka, izra`ena kot relativni standardni odmik,
je bila 10 %. Metoda je primerna za dolo~evanje OCP v mesnih izdelkih z visoko vsebnostjo ma{~ob. Meje zaznav za
OCP so bile v obmo~ju od 0.1 ppb za lindan ter 2 ppb za α-endosulfan.

Arh et al.: Modification of Method for the Determination of ...