Pesticide residues in cucumbers cultivated in Bangladesh
Jennie Haag Anna Landahl
cultivated in Bangladesh
Jennie Haag Anna Landahl
Abstract
Pesticides are widely used for preventing crop losses due to pest attack. In Bangladesh, food safety and health of farmers are being compromised as a result of poor regulation concerning usage of pesticides in food production. The aim of this study was to identify pesticides applied on cucumber crops in Bangladesh and quantify pesticide levels in these crops. A method for extraction and clean-‐
up was developed to allow the quantification of four pesticides by GC-‐ECD in vegetable samples, specifically cucumber. The accuracy of the method was validated using recovery and its precision by studying the standard deviation and relative standard deviation. Analysis of cucumber samples obtained in the field showed no traces of the target pesticides. The results indicate that different types of chemicals are used on the examined crops. It is also believed that the growth habit of cucumber may affect the exposure to pesticides. To overcome the health hazards, restrictions regarding the types and quantities of chemicals used on the fields need to be implemented. Further studies would benefit from being executed in a controlled environment, and from monitoring the types and amounts of pesticides that are applied.
Populärvetenskaplig sammanfattning
Bangladesh är ett av världens mest tättbefolkade länder och en tredjedel av befolkningen lever i fattigdom. Landets ekonomi och befolkningstäthet ställer höga krav på lantbrukets produktionskapacitet. En stor del av befolkningen försörjer sig på odling och många har problem med att få tillräckligt hög avkastning. För att minimera förluster i skörden till följd av skadedjursangrepp besprutas grödorna med pesticider.
Pesticider kan orsaka negativa hälsoeffekter då de innehåller ämnen som är avsedda att döda oönskade organismer. En del pesticider är också kända för att orsaka cancer, fosterskador, genetiska defekter och allergiska reaktioner. I nuläget finns ingen reglering för pesticidanvändning gällande applikationsmängd och kontinuitet i Bangladesh. Avsaknaden av regleringar och information till lantbrukarna leder ofta till att större kvantiteter än nödvändigt, och ur produktions-‐ och hälsosynpunkt berättigat, används. Detta innebär en hälsorisk för lantbrukare, då de ofta besprutar grödorna utan nödvändig skyddsutrustning. Det föreligger även en risk att människor som konsumerar besprutade livsmedel får i sig kemikalierna via födointaget. Det är därför av intresse att utreda hur höga halter av pesticider som återfinns i livsmedel.
En forskargrupp vid Institutionen för organisk kemi, Dhaka universitet, i Bangladesh har sedan 2003 studerat pesticider. Den studie som beskrivs här är en del av forskargruppens pågående projekt
”Studies of organic pollutants in food and environment” och syftar till att identifiera och kvantifiera giftiga kemikalier i miljö och livsmedel. Vi tog fram och utvärderade en metod för att kvantifiera fyra utvalda bekämpningsmedel i gurkor med gaskromatografi. Totalt analyserades 14 gurkor odlade på 9 olika platser i Bangladesh.
Resultatet av studien påvisade inga halter av de fyra utvalda pesticiderna i de undersökta proverna.
Detta är dock inget bevis på att inga kemikalierester finns i gurkorna eftersom andra pesticider kan ha använts. Studien genomfördes under regnperioden i Bangladesh och den kraftiga nederbörden tros ha påverkat resultaten. Det är också möjligt att gurkornas vertikala växtsätt, under ett täckande lövverk, kan ha en skyddande effekt mot besprutningen. Vidare studier hade gynnats av odling i en kontrollerad miljö där pesticidanvändandet kunde övervakas.
Acknowledgements
We would like to express our gratitude towards everyone that has contributed in making our project feasible. Without your knowledge and guidance it would not have been possible for us to actualize this study.
Thank you,
Henrik Kylin,
Professor at the department of water and environmental studies, Linköping University, for your commitment of being our Swedish supervisor and for devoting your time to our project.Peter Sundin,
Programme director at International Science Programme (ISP), Uppsala University, for guidance and assisting us with relevant contacts.Roger Herbert,
Senior lecturer at Department of Earth Sciences, Program for Air, Water and Landscape Sciences, Uppsala University, for your help with administrative matters.Sida,
through the International Science Programme (ISP) at Uppsala University, for funding our study in Bangladesh via an MFS scholarship.Dr. Nilufar Nahar,
Professor and Research Group Leader, Department of chemistry, University of Dhaka, for being our supervisor and dedicating your time to our project. For welcoming us to your research group and for your great hospitality during our stay in Dhaka.Dr. Mohammad Shoeb,
Professor at Department of Chemistry, University of Dhaka, for answering all of our questions, providing us with valuable information and making our stay memorable.Dr. Md. Iqbal Rouf Mamun,
Professor at Department of Chemistry, University of Dhaka, for your guidance and for always making us feel at home.Md. Nashir Uddin Al Mahmud,
Assistant professor, Government College Bangladesh, for your tireless help in the laboratory.We would like to thank everyone in the laboratory for helping us in our work and making us feel welcome.
Finally we would like to show our appreciation to all of our friends, both inside and outside the laboratory, for making our stay in Bangladesh unforgettable!
Uppsala, May 2014.
Jennie Haag & Anna Landahl
Table of Contents
Abstract ... i
Acknowledgements ... iii
1 Introduction ... 1
1.1 Background ... 1
1.2 Objective ... 1
2 Bangladesh ... 2
2.1 Pesticide use ... 2
3 Background ... 4
3.1 Gas chromatography with electron capture detector (GC-‐ECD) ... 4
3.2 Target pesticides ... 5
3.2.1 Cypermethrin ... 5
3.2.2 Diazinon ... 5
3.2.3 Fenvalerate ... 5
3.2.4 Chlorpyrifos ... 6
4 Method ... 7
4.1 Sample collection ... 7
4.2 Laboratory study ... 8
4.2.1 Method development ... 8
4.2.2 Method validation ... 11
5 Results ... 13
5.1 Gas chromatographic analyses ... 13
5.2 Calibration curves and recovery ... 18
6 Discussion ... 27
6.1 Method development ... 27
6.2 Contaminants ... 28
6.3 Method validation ... 28
6.4 Sources of error ... 29
7 Conclusions ... 29
8 List of abbreviations ... 31
9 References ... 32
Appendix 1 ... 34
Appendix 2 ... 35
1 Introduction 1.1 Background
The minor field study presented in this report was a part of an ongoing project performed by the organic environmental chemistry research group at Dhaka University, Bangladesh. The research group has been working on pesticide residue analyses since 2003. During the past ten years the research group has established methods to determine “old” (classical chlorinated insecticides) and
“new” current-‐use pesticides in different sample matrices and trained a number of skilled staff in the field. Currently, the research group is working on a project titled Studies of organic pollutants in food and environment. The objective of this project is to identify and quantify toxic chemicals in different food and environmental samples, educate students and young scientists, play an active role in both regional and international student exchange programmes, disseminate knowledge through conferences and seminars, and to publish scientific papers in peer reviewed journals. The project is mainly funded through the International Programme in the Chemical Sciences (IPICS) at Uppsala University.
1.2 Objective
Food safety is compromised in Bangladesh as a result of too high application of pesticides and the use of unauthorized toxic chemicals in food production. Pesticides used to prevent crop losses include compounds that are known to cause harm to human health and the environment. Many harmful effects are believed to be a direct result from overuse and misuse of toxic chemicals.
The objective of this study is to investigate the occurrence of four target pesticides in cucumber samples collected in Bangladesh and quantify the concentrations of these pesticides in the samples.
A method for extraction and clean-‐up of the samples was developed and validated to establish the quality, i.e., the precision, repeatability and of the accuracy of the method.
2 Bangladesh
Bangladesh is located where the river Padma enters the Bay of Bengal and consists of a river delta (Husain and Tinker, 2013). Bangladesh is one of the most densely populated countries in the world and around a third of the Bangladeshi people live in poverty. The population growth is causing a great pressure on the natural resources of the country, especially on the cultivatable land (IFAD, 2013). In Bangladesh around 84% of the population are, in some meaning, dependent on agriculture for their livelihood (Dasgupta et al., 2004). Despite this, most Bangladeshis struggle to keep the agricultural production at a significant level. Due to the country’s location, Bangladesh depends upon the vagaries of the monsoon (Husain and Tinker, 2013). Around two thirds of Bangladesh’s area is less than 5 meters above sea level. The exposure and vulnerability to floods makes the conditions for farming unsustainable. Approximately 60% of farmers in Bangladesh are functionally landless (FAO, 2013). The major staple crop of Bangladesh is rice which constitutes 71% of the caloric intake of the Bangladeshi people (Dasgupta et al., 2004). Major loss in rice production, caused by floods and pest attacks, forces the price of staple rice to increase rapidly. Also, the price of quality rice seeds is increasing, making it hard for small farmers to increase their production (FAO, 2013).
2.1 Pesticide use
Pesticide use in Bangladesh has increased rapidly over the past four decades (Figure 1). The farmers use pesticides to increase the crop production and to prevent crop losses due to pest attack (Figure 2; Rahman, 2012). The concerns regarding high pesticide usage are the possibility of pesticide resistance and their harmful effect of human health and environment (Rahman, 2002). Pesticides include compounds that are known to cause cancer, genetic damage, foetal defects, and allergic responses. Many of the harmful effects are believed to be a direct result from overuse and misuse of pesticides. The lack of information to the farmers leads to a higher use of toxic chemicals than recommended. Pesticide poisoning and environmental damages are now common in Bangladesh due to overuse. To prevent the hazardous effects of pesticides, farmers need to be educated about the risks of overuse and the importance of using safety gear (Dasgupta et al., 2005). Despite this, little effort has been made in Bangladesh to develop other methods than pesticides for pest management (Rahman et al., 1994).
Figure 1. Trends in pesticide use (MOA, 2007).
0 5000 10000 15000 20000
1985 1990 1995 2000 2005
Year
Pes(cide consump(on (mt/kl)
distribution and use of pesticides. However, the pesticide use in many developing countries is poorly regulated and therefore violations to the code may occur. DDT is an example of a pesticide that is banned in Bangladesh, but that is still available on the market (Meisner, 2004). Lately an increased concern for the sustainability of pesticide usage has arisen. Through collaboration with international development agencies, the Bangladeshi government promoted Integrated Pest Management (IPM), which is an alternative to conventional pest management. IPM is an ecologically based method to prevent insects from harming the crops. Through judicial use, biological techniques, natural parasites and predators to control pest populations, pesticide use is minimized, reducing the damages to human health and environment (Dasgupta et al., 2004). IPM was first used on rice crops in Bangladesh in 1981 through the FAO Inter-‐Country Programme (ICP). A national IPM policy was launched in Bangladesh in January 2002. Since the chemical pesticides are expensive the IPM can reduce the farmers’ costs, which can increase the profit. The adoption to IPM may however reduce the productivity, which can lead to decreased profits (Rahman, 2012).
Figure 2. Farmer applying pesticides on cucumber crops.
3 Background
In this section the choice and features of the analytical method and a description of the four target pesticides is presented.
3.1 Gas chromatography with electron capture detector (GC-‐ECD)
The gas chromatograph used for analysing the samples was a Shimadzu GC-‐2010 with an auto injector, AOC-‐20i, an RTX-‐5 MS column (30 m x 0.25 mm i.d. x 0.25 µm phase thickness, RESTEK, USA) and an electron capture detector (ECD) (Figure 3). Each injection was of 1 µL in the split-‐less mode, opening the split after 2 min with a split ratio of 20:1. The injector was held at 220 ⁰C and the detector at 290 ⁰C. The temperature programme was 120 ⁰C for 2 min, increasing 10 ⁰C/min to 270
⁰C and held for 1 min, and then 2 ⁰C/min to 290 ⁰C which was held for 3 min. Both the make-‐up and carrier gas was nitrogen, with a column flow of 1 mL/min.
It is important for the system that the injector acts like a sluice. This is established by a self-‐sealing rubber membrane which is pierced by an injection needle. The carrier gas, nitrogen, transports the sample from the injection through the column and to the detector. The column is located in a thermostated oven and contains a stationary and a mobile phase. When the sample extract is transported through the column its components are partitioned differently between the two phases and are thus transported through the column at different speed. The separation depends on the compound properties, column type, the gas flow, and the oven temperature. In general the components with the lowest boiling point are least adsorbed to the stationary phase, but a large difference in polarity between the compounds might also affect the retention. The components are then registered by the detector (Simonsen, 2005).
Figure 3. Schematic of the gas chromatographic system.
The ECD is selective and has low detection limits for compounds with high electron affinity (that
“capture electrons”), for example halogen-‐containing pesticides. The ECD detects compounds by decreasing the ionization level in the detector. A high standing current is produced in the ECD by interaction of a radioactive β-‐emitter and the carrier/make-‐up gas. When an electronegative compound reaches the detector it captures electrons and thereby decreases the current which results in a negative peak. The free electrons have a faster mobility than the formed negative ions which are not captured by the anode. The concentration of the compound is proportional to the degree of the captured ions. The identification of the components is based only on retention time
spectrometric analysis. This enables detection of the analysed compounds at lower concentrations (USGS, 2014).
3.2 Target pesticides
The four target pesticides are commonly used for pest control in Bangladesh.
3.2.1 Cypermethrin
Cypermethrin (Figure 4) is a synthetic pyrethroid insecticide that affects the insects’ central nervous system. It kills insects that come in contact with or ingest the substance. Cypermethrin is excreted quickly from the human body and is unlikely to bioaccumulate. When working with or handling cypermethrin, side effects such as skin burning and tingling, dizziness and itching might be experienced. The US EPA classifies the insecticide as a possible carcinogen (NPIC, 1998).
Figure 4. Structure of cypermethrin (Sigma-‐Aldrich, 2013a).
3.2.2 Diazinon
Diazinon (Figure 5) is used as insecticide, acaricide and nematicide. It is a synthetic chemical that belongs to a group called organophosphates. Diazinon is one of the most widely used insecticides for agricultural pest control. The chemical is fat-‐soluble and can be stored in fat tissues in the human body. When working with diazinon effects from daily exposure can be nausea, dizziness and headache. Studies have shown that long-‐term exposure to the substance can lead to neurological problems (NPIC, 2009a).
Figure 5. Structure of diazinon (Sigma-‐Aldrich, 2013b).
3.2.3 Fenvalerate
Fenvalerate (Figure 6) is a pesticide primarily used as an insecticide. The insecticide is stable to heat and sunlight and is applicable on a wide range of pests. Fenvalerate is considered to be of moderate toxicity to mammals and is harmful to the central nervous system after only a short-‐term or acute exposure (WHO and FAO, 1996). Symptoms experienced by humans when handling the pesticides, even handling according to recommendations, are burning sensation in the skin, cough, dizziness, headache and nausea (PAN, 2010).
Figure 6. Structure of fenvalerate (Sigma-‐Aldrich, 2013c).
3.2.4 Chlorpyrifos
Chlorpyrifos (Figure 7) is an organophosphate pesticide used in agriculture to control insect attacks.
The substance attacks the nerve cells which causes nervous system failure. Chlorpyrifos has a wide usage spectrum and kills insects upon contact. Signs of acute toxicity can be seen directly after exposure to the pesticide. When exposed to high doses humans may experience direct symptoms such as vomiting, abdominal cramps and diarrhoea. Neurological symptoms may also appear as a delayed symptom to the exposure (NPIC, 2009b).
Figure 7. Structure of chlorpyrifos (Sigma-‐Aldrich, 2013d).
4 Method
4.1 Sample collection
In order to get a representative picture of the overall pesticide usage on cucumber crops in Bangladesh, cucumber samples were collected from nine different areas (Figure 8).
Figure 8. Map of Bangladesh illustrating the location of the nine different collection areas (LGED, 2012).
In some cases several samples were gathered from different fields or markets in the same area. A total of 14 cucumber samples were collected and transported to Dhaka University for further analysis (Table 1). The exact application time, identity, amount and concentration of the used pesticide were unknown for most of the gathered samples. It was also uncertain which substances were used on the various fields due to the farmers’ frequent change of pesticides.
Table 1. Sample ID, location and date of collection for the 14 samples
Sample ID Location ID Location Date of Collection
AJ1 1 Ananda Bazar, Dhaka 17/06/2013
AJ 2 2 Shaitbaria, Kaligong 23/06/2013
AJ 3 2 Shaitbaria, Kaligong 23/06/2013
AJ 4 2 Shaitbaria, Kaligong 23/06/2013
AJ 5 3 Isshardichor, Mymensingh 28/06/2013
AJ 6 4 Balla kantha, Gofurgaon, Mymensingh 28/06/2013
AJ 7 5 Saiza Chor, Gofurgaon, Mymensingh 28/06/2013
AJ 8 5 Saiza Chor, Gofurgaon, Mymensingh 29/06/2013
AJ 9 6 Madarinagar, Nandail, Mymensingh 29/06/2013
AJ 10 7 Razabaria, Nandail, Mymensingh 29/06/2013
AJ 11 8 Varella, Comilla 12/07/2013
AJ 12 9 Gobindapure, Comilla 12/07/2013
AJ 13 9 Gobindapure, Comilla 12/07/2013
AJ 14 9 Gobindapure, Comilla 12/07/2013
4.2 Laboratory study
The cucumber samples collected from several fields in different districts of Bangladesh were transported to Dhaka University for analysis. A method for extraction and clean-‐up of the samples were developed and validated to establish the accuracy of the procedure. The extracts were then analysed by GC-‐ECD (Figure 9) to determine the samples interiorly pesticide concentrations.
Figure 9. GC-‐ECD (to the left) and rotary evaporator (to the right) used for analysis.
4.2.1 Method development
To achieve a high quality chromatogram with a baseline with sufficiently low noise, different procedures were tested. The procedures were improved until an adequate level of accuracy was accomplished. The reasons for developing the procedures are explained in the Results section.
4.2.1.1 Procedure 1
The extraction (Figure 10) was initiated by washing the cucumber samples with sufficient water and homogenized with a kitchen blender. Out of the cucumber mash, three replicas (10 g each) were transferred to 50-‐ml Teflon tubes and the additional mash was stored in the freezer. The 10 g of mash was mixed with 20 ml of ethyl acetate and then shaken vigorously for one minute and then vortexed for one minute. To separate the water from the sample 6 g of anhydrous magnesium sulphate (MgSO4) and 1.5 g sodium chloride (NaCl) were added to the tubes. The samples were shaken and vortexed as above and then centrifuged for 5-‐7 min at 4000 rpm. A glass pipette was used to pipette 10 ml of the extract into a round bottom flask (RB-‐flask). The extract was then evaporated using a rotary evaporator (Figure 9) until it was completely dry. To make sure there was no remaining water in the sample 3-‐4 ml n-‐hexane was added and the sample was again evaporated until dryness.
Figure 10. Flow scheme of the extraction of Procedure 1.
The clean-‐up method (Figure 11) used for the first procedure was executed using a combined florisil–
alumina column. The column was made by adding 5g florisil, 5 g alumina and 0.5 g charcoal to a 100 ml flask. To dissolve the solid material 20 ml n-‐hexane were added to the RB-‐flask and then poured into the column. The column was rinsed with 50 ml n-‐hexane until equilibrium was reached and then 5 g anhydrous sodium sulphate (Na2SO4) was added. To dissolve the dried extract, 2-‐3 ml n-‐hexane was added and the RB-‐flask was put into an ultrasonic bath for 30 seconds. The extract was then added to the column. To elute the column 20 ml n-‐hexane were added and when the n-‐hexane surface was 2 cm above the packing 80 ml of Dichloromethane (DCM) was added to the column. The DCM extract were collected into RB-‐flasks and evaporated until dryness. To make sure that the sample was completely dry 3-‐4 ml n-‐hexane were added and the extract was evaporated again. The extract was then dissolved with 4 ml n-‐hexane and 2 ml of the mix were transferred to a vial, using a glass pipette.
Figure 11. Flow scheme of clean-‐up of Procedure 1.
4.2.1.2 Procedure 2
The extraction was executed as described in procedure 1 (4.2.1.1). The clean-‐up (Figure 12) was performed by adding 150 mg Primary Secondary Amine (PSA) and 750 mg Anhydrous MgSO4 to test tubes. The dried extract were dissolved using 5 ml n-‐hexane and 3 g Na2SO4 were then added to the RB-‐flasks. Using a glass pipette, 2 ml of the extract were transferred to test tubes and the samples were vortexed for one minute and centrifuged for five minutes. The extract was rinsed through a 0.45 μm filter and transferred to vials.
Homogenize washed cucumber with kitchen blender
Add 10 g of cucumber mash in 50 ml teflon
tubes
Add 20 ml ethyl acetate. Shake vigorously for 1 min.
Vortex for 1 min
Add 6 g anhydrous MgSO4 and 1.5 g NaCl.
Shake vigorously for 1 min. Vortex for 1 min.
Centrifuge the teflon tubes for 5-‐7 min
Pipese 10 ml of the extract into RB-‐flask.
Evaporate untl dryness
Add 3-‐4 ml n-‐Hexane to the extract.
Evaporate to dryness
Add 5 g Florisil, 5 g Alumina and 0.5 g Charcoal into a 100 ml
flask
Add 50 ml n-‐Hexane to flask and pack the column. Add 5 g of sodium sulphate
Dissolve the dried extract with 2-‐3 ml n-‐
Hexane. Place in a ultrasonic bath for 30
sec
Apply the extract to the
column Elute the column with
20 ml n-‐Hexane
Add 80 ml of Dichloromethane
(DCM).
Figure 12. Flow scheme of clean-‐up of Procedure 2.
4.2.1.3 Procedure 3
Procedure 3 (Figure 13) was performed as described in the extraction from procedure 1 up to the centrifugation step. After being centrifuged, 5 ml of the extract were transferred to test tubes and 3 g Na2SO4 were added. The tubes were then vortexed for one minute. Clean test tubes were filled with 150 mg PSA and 750 mg anhydrous MgSO4 and 2 ml of the extract were transferred to the clean tubes. The sample was vortexed for one minute, centrifuged for five minutes, then rinsed through a 0.45 μm filter and transferred to vials.
Figure 13. Flow scheme of Procedure 3.
4.2.1.4 Procedure 4
The extraction of the final procedure (Figure 14) follows in procedure 1 up to the centrifugation step.
After the centrifugation, 10 ml of the extract were passed through 20 g Na2SO4 into a RB-‐flask. The
Na2SO4 were then rinsed with 10 ml ethyl acetate and the sample was evaporated until dryness. To
ensure that the sample was completely dry, 3-‐4 ml n-‐hexane were added and the evaporation was repeated.
Add 150 mg PSA and 750 mg Anhydrous MgSO4 to test tubes
Elute the dried extract with 5 ml n-‐
hexane
Add 3 g Na2SO4 to the RB-‐flasks
Pipese 2 ml of the extract to the test
tubes
Vortex the samples for 1 min and centrifuge for 5 min
Rinse the extract through a 0.45 μm filter and add to vial
Homogenize washed cucumber with kitchen blender
Add 10 g of cucmber mash in 50 ml teflon
tubes
Add 20 ml ethyl acetate. Shake vigorously for 1 min.
Vortex for 1 min
Add 6 g anhydrous MgSO4 and 1.5 g NaCl.
Shake vigorously for 1 min. Vortex for 1 min
Centrifuge the teflon tubes for 5-‐7 min
Pipese 5 ml of the extract to test tubes
and add 3 g Na2SO4
Vortex the test tubes for 1 min
Add 150 mg PSA and 750 mg Anhydrous MgSO4 to clean test
tubes
Pipese 2 ml of the extract to the test
tubes
Vortex the samples for 1 min and centrifuge for 5 min
Rinse the extract through a 0.45 µm filter and add to vial
Figure 14. Flow scheme of extraction of Procedure 4.
The clean-‐up (Figure 15) were made by first adding 150 mg PSA and 750 mg Anhydrous MgSO4 to test tubes. The dried extract was dissolved with 5 ml n-‐hexane and 3 g Na2SO4 were added to the RB-‐
flasks. From the dissolved extract, 2 ml was transferred to the test tubes which was vortexed for one minute and centrifuged for five minutes. The extract was rinsed through a 0.45 μm filter and transferred to vials.
Figure 15. Flow scheme of clean-‐up of Procedure 4.
4.2.2 Method validation
The purpose of method validation is to evaluate the quality of the data produced with the method. In this study the validation characteristics are linearity, repeatability, accuracy, precision and limit of detection (LOD). The linearity and Limit of Detection (LOD) were determined by a calibration curve and the precision by calculations of the standard deviation for the replicas. The accuracy of the analytical method was determined by recovery.
4.2.2.1 Calibration curve
The concentration of pesticides in the cucumber samples were determined by creating a calibration curve that shows the relation between the pesticide concentration and the detector response. A mixture with concentration 1 μg/g of the four target pesticides diazinon, chlopryrifos, cypermethrin and fenvalerate (DCCF) was diluted to eight different concentrations using n-‐hexane (Table 2) to
Homogenize washed cucumber with kitchen
blender
Add 10 g of cucumber mash in 50 ml teflon
tubes
Add 20 ml ethyl acetate. Shake vigorously for 1 min.
Vortex for 1 min
Add 6 g anhydrous
MgSO4 and 1.5 g NaCl.
Shake vigorously for 1 min. Vortex for 1 min
Centrifuge the teflon
tubes for 5-‐7 min Add 20 g Na2SO4 to a funnel
Pipese 10 ml of the extract to through the funnel in to a RB-‐
flask
Rinse with 10 ml ethyl acetate
Evaporate extract untl dryness
Add 3-‐4 ml n-‐hexane and evaporate untl
dryness
Repeat the previous step
Add 150 mg PSA and 750 mg Anhydrous MgSO4 to test tubes
Elute the dried extract with 5 ml n-‐
hexane
Add 3 g Na2SO4 to the RB-‐flasks
Pipese 2 ml of the extract to the test
tubes
Vortex the samples for 1 min and centrifuge for 5 min
Rinse the extract through a 0.45 μm filter and add to vial
create standard solutions. The standard solutions were then analysed by a GC-‐ECD and the concentrations were plotted against the areas in the chromatograms. From this a linear equation was obtained. The LOD for each pesticide was determined as the lowest amount detected in the GC-‐ECD.
Table 2. Dilution of the pesticide mixture DCCF to produce listed standard concentrations Standard Concentration
(µg/g) Concentration DCCF
(µg/g) Volume DCCF
(ml) Volume n-‐Hexane
(ml)
0.1 1 1 9
0.05 0.1 5 5
0.025 0.05 5 5
0.01 0.025 4 6
0.005 0.01 5 5
0.0025 0.005 5 5
0.001 0.0025 4 6
0.0005 0.001 5 5
4.2.2.2 Recovery
Recovery is a measure of how much of an analyte that is lost during the clean-‐up procedure. A cucumber sample, which was analysed and found to be a blank matrix, was spiked with a mixture of DCCF at two different concentrations. Three replicas were made for each of the two spiking levels, High Recovery (HR) and Low Recovery (LR). The HR was spiked with a concentration corresponding to 0.1 µg/g and the LR with 0.05 µg/g wet mass. Extraction and clean-‐up was executed according to Procedure 4 (4.2.1.4) and the samples were analysed with GC-‐ECD. The recovery was calculated using equation 1. The average value, standard deviation (SD) and relative standard deviation (RSD) were calculated using Excel. The standard deviation and the relative standard deviation indicate the methods precision, the correspondence between the separate measurements.
𝑅𝑒𝑐𝑜𝑣𝑒𝑟𝑦 (%) = !!! ∙!∙!∙!""∙!
!∙!∙! Equation 1
A = Area of the chromatogram m = intercept in calibration curve e = ethyl acetate in extract (ml) n = n-‐hexane in extract (ml) s = spiking level (μg/g)
k = gradient of calibration curve g = matrix (g)
p = pipetted amount (ml)
5 Results
5.1 Gas chromatographic analyses
The samples collected from Ananda Bazar in Dhaka were extracted and cleaned-‐up using procedure 1 (4.2.1.1) and analysed by GC-‐ECD. The chromatogram showed a unsatisfying baseline and a unidentified peak after 20 minutes (Figure 16).
Figure 16. Chromatogram for sample AJ2 using procedure 1.
The unidentified peak at ca. 20 minutes was believed to be a result of high water concent in the anlysed sample. A new method was devoloped where the florisil-‐alumina column were excluded and replaced by another extraction method using PSA (Procedure 2, 4.2.1.2). When analysing samples using this method it was seen that the baseline was still not good enough and the peak was still present (Figure 17).
Figure 17. Chromatogram for sample AJ5 using procedure 2.
Since the matrix had a high water content, a second analysis was done adding both PSA and C18 to the extract. The result was disappointing, no change were shown in the chromatogram (Figure 18).
Figure 18. Chromatogram for sample AJ5 using procedure 2 adding both PSA and C18.
The same procedure (procedure 2) was redone adding 5 g of Na2SO4 instead of 3 g used in previous analysis to reduce the amount of water in the sample. The baseline did not improve and the unidentified peak was not reduced by these measures (Figure 19).
Figure 19. Chromatogram for sample AJ4 using procedure 2 adding 5 g Na2SO4.
The possibility of the solvent being the cause of the unwanted peak and the poor baseline were evaluated by changing the solvent ethyl acetate to acetoonitrile. Samples were extracted and analysed in the same way as previously but with acetonitrile as solvent (Figure 20).
Figure 20. Chromatogram for sample AJ9 using procedure 2 with acetonitrile as solvent.
Since the peak and the poor baseline did not improve when switching solvent it was believed that the problem might be in the added magnesium sulfate. To evaluate this, three reagent blanks using n-‐
hexane (Figure 21), acetonitrile (Figure 22) and ethyl acetate (Figure 23) were analysed in the GC-‐
ECD.
Figure 21. Chromatogram for reagent blank (n-‐hexane).
Figure 22. Chromatogram for reagent blank (acetonitrile).
Figure 23. Chromatogram for reagent blank (ethyl acetate).
The chromatograms were still showing an unsatisfying baseline and an unidentified peak, therefore conclusions could be made that neither the MgSO4 nor the solvents were the source for the uneven baseline and high peak. To rule out the possibility of contamination from the rotary evaporator, used in all procedures to dry the extract, an additional procedure were developed (4.2.1.3). Samples were extracted and cleaned-‐up without the interference of the rotary evaporator. Results were improved but the baseline was still not satisfactory (Figure 24).
Figure 24. Chromatogram for sample AJ4 using procedure 3.
A last and final procedure (procedure 4) were developed by adding steps to procedure 2 to further reduce the water content in the sample. Samples analysed with this method displayed an adequately good baseline, but the unidentified peak was still present (Figure 25)
Figure 25. Chromatogram for sample AJ11 using the final procedure.
Since the unidentified peak was not interfering with any of the target analytes it was determined that future analyses would be carried out by using the final procedure and the “ghost peak” should be ignored. The obtained chromatograms for the total 14 samples are represented in Appendix 2. None of the four target pesticides were detected in any of the analyzed samples.
After the laboratory part of the study was completed new information was obtained about the poor baseline and the large peak that are present in most of the chromatograms. The undulating baseline at number 1 in Figure 27 (between retention times approximately 12-‐24 min) is believed to be short
chained chlorinated paraffins. The large peak at number 2 is thought to be dioctylphthalate and the cluster of peaks at number 3 probably consists of dinonyl-‐ or didecylphthalates. Possible sources of these compounds are presented in the Discussion.
Figure 26. Chromatogram for high recovery illustrating the contaminants that causes a poor baseline (no. 1 and 3) and a
high peak (no. 2).
5.2 Calibration curves and recovery
By performing recovery experiments the accuracy of the final procedure were validated. Standard solutions were analyzed and chromatograms for the concentrations 0.1 µg/g (Figure 27), 0.05 µg/g (Figure 28), 0.025 µg/g (Figure 29), 0.01 µg/g (Figure 30), 0.005 µg/g (Figure 31) and 0.0025 µg/g (Figure 32) were obtained.
Figure 27. Chromatogram of DCCF 0.1 µg/g for calibration curve.
Figure 28. Chromatogram of DCCF 0.05 µg/g for calibration curve.
Figure 29. Chromatogram of DCCF 0.025 µg/g for calibration curve.
Figure 30. Chromatogram of DCCF 0.01 µg/g for calibration curve.
Figure 31. Chromatogram of DCCF 0.005 µg/g for calibration curve.
Figure 32. Chromatogram of DCCF 0.0025 µg/g for calibration curve.
With the known pesticide concentrations and received areas from the chromatograms, calibration curves were made for the four target pesticides (Figure 33). The coefficient of determination is satisfying for all of the regressions and is valid to use in the further calculations of the recovery.
Figure 33. Calibration curves for diazinon, chlorpyrifos, cypermethrin and fenvelarate.
The samples for the different recovery levels were anlyzed in the GC-‐ECD. Chromatograms for each replica of the HR level (Figure 34, Figure 35 and Figure 36) and LR level (Figure 37, Figure 38 and Figure 39) were obtained.
Figure 34. Chromatogram for recovery level HR, replica number one.
y = 501664x + 3222,1 R² = 0,99278
0 50000 100000
0 0,05 0,1
Area
Concentra(on (μg/g)
Diazinon
y = 1 397 835x + 3 359 R² = 1
0 100000
0 0,05 0,1
Area
Concentra(on (μg/g)
Cypermethrin
y = 2 919 385,96x + 36 234,70 R² = 0,98
0 200000 400000
0 0,05 0,1
Area
Concentra(on (μg/g)
Chlorpyrifos
y = 1 145 546,98x + 4 715,53 R² = 1,00
0 100000 200000
0 0,05 0,1
Area
Concentra(on (μg/g)
Fenvalerate
Figure 35. Chromatogram for recovery level HR, replica number two.
Figure 36. Chromatogram for recovery level HR, replica number three.
Figure 37. Chromatogram for recovery level LR, replica number one.
Figure 38. Chromatogram for recovery level LR, replica number two.
Figure 39. Chromatogram for recovery level LR, replica number three.
By using the received areas and the linear equation from the calibration curves the recovery percentage were calculated by using Equation 1 (4.2.2.2). From this the standard deviation and relative standard deviation were calculated (Table 3). The obtained recovery values were evaluated and the procedure could be validated.
Table 3. Obtained results for area, recovery, SD and RSD for the four target pesticides
Sample Area Recovery Average SD RSD
Diazinon
LR1 29000 100
LR2 29000 100 100 3.0 2.9
LR3 30000 110
HR1 47000 88
HR2 50000 94 92 3.4 3.7
HR3 50000 94
Chlorpyrifos
LR1 180000 100
LR2 190000 110 110 8.7 8.1
LR3 210000 120
HR1 340000 100
HR2 350000 110 110 2.9 2.7
HR3 350000 110
Cypermethrin
LR1 54000 73
LR2 59000 80 77 3.4 4.4
LR3 58000 78
HR1 150000 100
HR2 160000 110 99 16 16
HR3 120000 81
Fenvalerate
LR1 60000 97
LR2 58000 93 94 2.4 2.6
LR3 57000 92
HR1 140000 110
HR2 150000 120 120 5.1 4.3
HR3 140000 120