Reliable quantitation of pesticide residues in tobacco, with single quadrupole GC-MS
Introduction
Tobacco is an important cultivated plant in more than 125 countries and more than 4 million hectares of fertile land is used for tobacco cultivation worldwide. Over the past years, the lands for tobacco cultivation in countries such as the United States, Canada and Mexico have decreased in area by half, while in China, Malawi and the United Republic of Tanzania, on the contrary, the area has doubled. Currently, world production of tobacco exceeds 7 million tons per a year.
Tobacco is a sensitive plant, prone to many diseases. So plant care requires a huge amount of chemicals: within a three-month vegetation period it is recommended to carry out up to sixteen treatments with pesticides for the effective protection of the tobacco plant to yield a good harvest. This requires careful monitoring and implementation of safety measures on the application of pesticides, their residues monitoring both in vegetative parts of plants during the growth process and in the final product. So tight control must ensure safety and quality of the final tobacco product.
This work was done to demonstrate an effective method for multi-residue pesticide screening of dried tobacco samples using QuEChERS sample preparation followed by GC-MS detection with highly sensitive mass-spectrometric detector. The pesticide residue LODs in dried tobacco samples for the used method were found to be below MRL specified by regulator [2].
Experimental
Standards and Reagents
All reagents used in this work were HPLC grade. Pesticide standards were commercially obtained from Restek (USA) и Supelco (USA), in particular: Pesticide Mix A-2 (Supelco 46862-U); GC Multiresidue Pesticide Std#1 – OPP (Restek 32563); 8140/8141 OP Pesticide Calibration Mix A (Restek 32277). Besides, Triphenylphosphate Standard (Restek 32281) was used as internal standard.
Extract of a tobacco sample, prepared by QuEChERS sample preparation technique was used in this work. [1]
The Extract was spiked with 45 pesticide mix with 5 concentration levels; 10; 20; 50; 100; 500; 1000 pg/mL (ppb), to build the matrix calibration.
In order to assess effectiveness of proposed QuEChERS sample preparation technique as well as for evaluation of the GC-MS system performance, dry the tobacco sample was also spiked with same 45 pesticide mix calculated in a way to obtain final extracts with concentrations of 50 ppb and 500 ppb respectively.
Table. 1. GC-MSD method parameters.
| Instruments | |
| GC-MS system model | GC-MS |
| inlet | Split/Splitless (liner: Restek 4 mm x 6.3 x 78.5) |
| GC column | Rtx-5MS (30m x 0.25mm; 250um) |
| Method parameters | |
| Injected Sample volume | 1 uL («Sandwitch» mode, 1 air layer – 1 ul) |
| Inlet mode | Splitless; Purge 70 ml/min after 0,7 min |
| Inlet temperature | 250°C |
| Carrier gas flow mode | Constant flow |
| Oven program | 50°C for 1,23 min;
30,0°C/min until 90°C, hold 0,0 min; 9,0°C/min until 310°C, hold 4,0 min; |
| Carrier gas flow and type | (He) 1,2 ml/min |
| GC-MS transfer line temperature | 280°C |
| Detector settings | |
| Ion source | Electron impact (EI) |
| Ion source temperature | 250°C |
| Solvent delay | 4,0 min (240 sec) |
| Data Acquisition mode | SCAN 50 – 550 Da / SIM (see below) |
| Dwell Time | Automatically calculated by iDwell® Time |
The SIM method development involved two steps.
The first step involved GC run of pure pesticide standards mix dissolved in hexane at concentrations of 1 ppm each followed by MS data acquisition in full scan mode to obtain pure mass spectra and enable Q-Tek proprietary data processing algorithm SIM Wizard ® to automatically build SIM method. Such novel algorithm has been described earlier, it involves the procedure for automatic integration of the resulting SCAN-profile experiment, confident identification of target components taking record of their respective retention times based on the chromatographic profile (TIC) and population of SIM method table with ions to be monitored for all selected components. Finally, SIM Wizard ® recommends list of SIM ions for each pesticide target and optimizes time segmented SIM program to achieve best sensitivity at lowest RSD. Notably, the SIM Wizard ® algorithm is fully automatic and takes less than a minute from integration of TIC to optimized time segmented SIM program for the given GC separation.
Worth mentioning, the software “Q-Tek-operator” offers an operator a built-in library of recommended SIM ions for a few hundreds of compounds (Figure 1). The library has been populated by data obtained from a variety of open literature sources, vendors publications as well as from own experimental data collected by Q-Tek application team, that altogether assures high quality of the SIM-library and helps an operator to build a highly sensitive SIM-method. The library of SIM ions (SIM-library) can be edited by the operator, it can be augmented or customized for later use in a specific lab.
Fig. 1. Workflow of quantitation (SIM) method set-up.
SIM-library operating concept.
Each compound in the SIM-library is recorded as an information card (fig. 2). The compound card contains the list of the compound target ions (highlighted ion for the quantitative analysis (main, quantifier) and confirmation ions (QC, qualifiers) ) as well as basic compound information, including the compound molecular weight, name, type, CAS number. The latter is the main link used as a search key to verify if a compound detected in full Scan mode after integration is already present in user SIM-library. Therefore, if the CAS number of the detected compound matches that in SIM-library the SIM Wizard® algorithm will offer to use the ions from SIM-library as a preferred option. Otherwise, selection of target ions will be executed by SIM Wizard® algorithm based on the compound mass-spectrum that has been recorded during data acquisition in Full Scan mode.
Having selection of compounds target ions for the SIM method accomplished, the Maestro-Operator Software is ready to proceed to step 2.
fig. 2. Compound information card in SIM-library
The Step 2 is automatic calculation of SIM method acquisition parameters using iDwell ® Time algorithm for optimal ion scanning time for each target compound in the sample. The iDwell ® Time algorithm splits the chromatogram into time segments, in order to minimize number of monitored ions per a specific time-segment (fig. 3). This approach greatly improves ion statistics resulting in correct peak shape and improved sensitivity and precision of the final method.
fig. 3. Automatic split of TIC by iDwell®Time algorithm into time segments to be used in SIM-method.
Sample Preparation
QuEChERS sample preparation technique was used. The QuEChERS technique is well described in method for pesticide analyses [1] , which is based on acetonitrile sample extraction followed by SPE clean-up of the matrix components from the extract.
To extract the target compounds from tobacco sample the acetonitrile extraction was used at 2.0 g (sample) to 10.0 ml (acetonitrile) ratio. The extract was cleaned with Bondesil PSA sorbent, then the aliquot of tobacco extract was transferred to an autosampler vial and injected to the Q-Tek GC-MS.
Fig. 4. Sample preparation steps by QuEChERS.
Table. 2. Retention time and monitored ions for target pesticide compounds used for data acquisition
| № | Compound name | RT, min | Quantifier Ion | Qualifier Ion 1 | Qualifier Ion 2 |
| 1 | Dichlorvos | 7.7 | 185.0 | 109.0 | – |
| 2 | cis-Mevinphos | 10.2 | 192.0 | 127.0 | 164.0 |
| 3 | Demeton-O | 12.6 | 88.0 | 171.0 | 114.0 |
| 4 | 2,4,5,6-Tetrachloro-m-xylene | 12.7 | 245.9 | 243.9 | 209.0 |
| 5 | Ethoprophos | 12.9 | 200.0 | 158.0 | 139.0 |
| 6 | Naled | 13.2 | 185.0 | 109.0 | 145.0 |
| 7 | Phorate | 13.6 | 260.0 | 121.0 | 97.0 |
| 8 | α -Lindane | 13.7 | 216.9 | 180.9 | 219.0 |
| 9 | Demeton-S | 14.0 | 88.0 | 171.0 | 114.0 |
| 10 | δ-Lindane | 14.5 | 181.0 | 217.0 | 183.0 |
| 11 | Diazinone | 14.9 | 304.1 | 179.1 | 137.1 |
| 12 | Disulfoton | 15.0 | 274.0 | 142.0 | 88.0 |
| 13 | Isazophos | 15.2 | 257.0 | 161.0 | 119.0 |
| 14 | Chlorpyriphos-methyl | 15.9 | 285.9 | 287.9 | 125.0 |
| 15 | Heptachlor | 16.1 | 273.9 | 271.9 | 100.0 |
| 16 | Fenchlorphos | 16.3 | 284.9 | 286.9 | 125.0 |
| 17 | Fenitrothion | 16.6 | 277.1 | 260.0 | 274.1 |
| 18 | Pirimiphos-methyl | 16.6 | 290.0 | 305.1 | 276.1 |
| 19 | Fenthion | 16.9 | 278.0 | 245.0 | 263.0 |
| 20 | Chlorpyrifos | 17.0 | 313.9 | 199.0 | 315.9 |
| 21 | Trichloronate | 17.3 | 297.0 | 269.0 | 109.0 |
| 22 | Pirimiphos-ethyl | 17.5 | 331.1 | 318.1 | 304.1 |
| 23 | Merphos | 17.7 | 298.1 | 153.0 | 209.0 |
| 24 | Quinalphos | 17.9 | 298.1 | 146.0 | 157.1 |
| 25 | Tetrachlorvinphos | 18.5 | 330.9 | 328.9 | 109.0 |
| 26 | α-Endosulfan | 18.5 | 338.8 | 240.9 | 263.9 |
| 27 | Prothiofos | 18.8 | 309.0 | 311.0 | 239.0 |
| 28 | Tribufos | 18.9 | 169.0 | 285.1 | 202.0 |
| 29 | Dieldrin | 19.1 | 379.9 | 262.9 | 243.0 |
| 30 | Endrin | 19.5 | 279.0 | 243.0 | 245.0 |
| 31 | Fensulfothion | 19.7 | 293.1 | 292.1 | 140.0 |
| 32 | p,p’-DDD | 19.9 | 236.9 | 165.1 | 234.9 |
| 33 | Sulprofos | 20.2 | 322.0 | 139.0 | 140.0 |
| 34 | p,p’-DDT | 20.6 | 236.9 | 165.1 | 234.9 |
| 35 | Triphenyl phosphate (ISTD) | 21.0 | 326.0 | 170.0 | 215.1 |
| 36 | Pyridafenthion | 21.5 | 340.0 | 199.1 | 188.1 |
| 37 | Phosmet | 21.6 | 76.1 | 160.0 | 161.0 |
| 38 | EPN | 21.7 | 185.0 | 157.0 | 169.0 |
| 39 | Methoxychlor | 21.8 | 227.1 | 228.2 | 274.1 |
| 40 | Phosalone | 22.4 | 366.9 | 182.0 | 184.0 |
| 41 | Azinphos-methyl | 22.4 | 160.0 | 132.1 | 77.0 |
| 42 | Pyrazophos | 23.0 | 232.1 | 265.1 | 373.1 |
| 43 | Azinphos-ethyl | 23.1 | 132.1 | 160.0 | 77.0 |
| 44 | Pyraclofos | 23.2 | 360.0 | 139.0 | 194.0 |
| 45 | Coumaphos | 23.8 | 361.9 | 363.9 | 228.1 |
| 46 | Decachlorobiphenyl | 25.1 | 497.8 | 427.8 | 493.8 |
Results and discussion
Linearity
Pesticides quantitation in tobacco samples were carried out on the GC-MS following matrix calibration using 1ppm triphenyl phosphate as internal standard.
Fig. 5. Extracted ion chromatogram of pesticide mix @ 1 ppm in tobacco matrix.
1-Dichlorvos; 2-cis-Mevinphos; 3-Demeton-O; 4-(2,4,5,6)-Tetrachloro-m-xylene; 5-Ethoprophos; 6-Naled; 7-Phorate; 8-α-Lindane; 9-Demeton-S; 10-δ-Lindane; 11-Diazinone; 12-Disulfothon; 13-Isazofos; 14-Chlorpyrifos-methyl; 15-Heptachlor; 16-Fenchlorphos; 17-Fenitrothion; 18-Pirimifos-methyl; 19-Fenthion; 20-Chlorpyrifos; 21-Trichloronate; 22-Pirimiphos-ethyl; 23-Merphos; 24-Quinalphos; 25-Tetrachlorvinphos; 26-α-Endosulfan; 27-Prothiofos; 28-Tribufos; 29-Dieldrin; 30-Endrin; 31-Fensulfothion; 32-p,p’-DDD; 33-Sulprofos; 34-p,p’-DDT; 35-Triphenylphosphate (ISTD); 36-Pyridafenthion; 37-Phosmet; 38-EPN; 39-Methoxychlor; 40-Phosalone; 41-Azinphos-methyl; 42-Pyrazophos; 43-Azinphos-ethyl; 44-Pyraclofos; 45-Coumaphos; 46-Decachlorobiphenyl
For calculating the limit of detection for all 45 pesticides three repeated injections of spiked tobacco calibration samples at each calibration point were done . The LOD calculation was done as per European ACAC directive [3]. Data from the study are presented in table 3.
Calibration coefficient value (R2) for 40 of the 45 compounds was better than ≥ 0.998 (table 4), except a few compounds, namely: Demeton-O, Naled, EPN, Merphos and Methoxychlor which showed (R2) values 0.998 > R2 > 0.984.
Results and discussion
According to data received 41 of 45 pesticides can be detected at a level below or equal to 0.1 ppm, while the recoveries of the target pesticides by QuEChERS sample preparation technique vary in range from 40 to 130%. Such a wide variation of the compounds recovery has to deal with certain problems related to the extract clean-up because of the difficulty and complexity of the matrix.
According to open source published data, tobacco contains about 4000 names of chemical elements.
Furthermore, raw tobacco undergoes many different physiological processes during preparation of tobacco products, namely at the stage of natural or artificial ageing. These data are confirmed by research (Lili Li & Jieyu Zhao and others), which noted a drastic change of pigments, lipids, amino acids, polyamines and many secondary metabolites in the composition of tobacco, as well as the accumulation of sterols esters and alkaloids in that production stage of tobacco. Therefore, all these compounds are showing out as numerous peaks during mass-spectrometric detection interfering with the target components, leading to increase or suppression of target chromatographic peaks in the profile, and as a result to the ambiguity of target identification and quantitation.
At the same time it must be taken into consideration the need to reduce the limit of detection of the target compounds, due to the fact that the initial samples are extracted with 5 times increased volume of the solvent, hence, the mass-spectrometric system must confidently detect at concentrations that are 5 times lower due to the dilution.
fig. 6. Extracted ion chromatogram of tobacco matrix spiked with pesticides @ 5 ppb level.
A reason for the high LODs for some compounds can also be their ability to degrade, even at room temperature. For example, both Naled and Trichloronate are known to degrade with formation of Dichlorvos even at room temperature, hence these three pesticides are sometimes analyzed and quantified as Dichlorvos.
Table. 3. Detection limits (LOD), linearity and recoveries for the target pesticides in this work.
| № | Pesticide compound | Linearity, R2 | LOD, ppm | MRL, ppm (ACAC) | Recovery, % | |
| 0.05 ppm | 0.5 ppm | |||||
| 1 | Dichlorvos | 0.999 | 0.029 | 0.1 | 67.3 | 63.3 |
| 2 | cis-Mevinphos | 0.998 | 0.045 | 0.04 | 73.0 | 92.2 |
| 3 | Demeton-O | 0.996 | 0.07 | Not specified | n/a | 105.3 |
| 4 | 2,4,5,6-Tetrachloro-m-xylene | 0.999 | 0.011 | Not specified | 74.7 | 62.3 |
| 5 | Ethoprophos | 0.998 | 0.040 | 0.1 | 93.2 | 113.2 |
| 6 | Naled | 0.986 | 0.143 | 0.1 | n/a | n/a |
| 7 | Phorate | 0.999 | 0.040 | 0.05 | 102.0 | 106.32 |
| 8 | α -Lindane | 0.999 | 0.017 | 0.05 | 77.4 | 88.5 |
| 9 | Demeton-S | 0.998 | 0.05 | 0.1 | 31.6 | 136.7 |
| 10 | δ-Lindane | 0.999 | 0.013 | 0.05 | n/a | n/a |
| 11 | Diazinone | 0.999 | 0.010 | 0.1 | 105.8 | 109.8 |
| 12 | Disulfoton | 0.999 | 0.015 | 0.1 | 99.2 | 102.4 |
| 13 | Isazophos | 0.999 | 0.017 | Not specified | 105.3 | 94.72 |
| 14 | Chlorpyriphos-methyl | 0.997 | 0.082 | 0.2 | 98.1 | 91.5 |
| 15 | Heptachlor | 0.999 | 0.013 | 0.02 | 113.1 | 90.0 |
| 16 | Fenchlorphos | 0.997 | 0.059 | Not specified | 88.7 | 104.5 |
| 17 | Fenitrothion | 0.999 | 0.048 | 0.1 | 81.8 | 106.1 |
| 18 | Pirimiphos-methyl | 0.998 | 0.064 | 0.1 | 93.2 | 108.0 |
| 19 | Fenthion | 0.997 | 0.067 | 0.1 | n/a | 111.58 |
| 20 | Chlorpyrifos | 0.998 | 0.132 | 0.5 | 96.6 | |
| 21 | Trichloronate | 0.998 | 0.052 | Not specified | 101.3 | 112.42 |
| 22 | Pirimiphos-ethyl | 0.997 | 0.073 | Not specified | 110.0 | 65.8 |
| 23 | Merphos | 0.978 | 0.181 | Not specified | n/a | n/a |
| 24 | Quinalphos | 0.999 | 0.037 | Not specified | 113.7 | 90.05 |
| 25 | Tetrachlorvinphos | 0.997 | 0.057 | Not specified | 83.8 | 63.6 |
| 26 | α-Endosulfan | 0.999 | 0.015 | 1.0 | n/a | 102.3 |
| 27 | Prothiofos | 0.999 | 0.027 | Not specified | 94.7 | 102.83 |
| 28 | Tribufos | 0.998 | 0.047 | Not specified | 108.9 | 96.9 |
| 29 | Dieldrin | 0.999 | 0.034 | 0.02 | n/a | 94.2 |
| 30 | Endrin | 0.999 | 0.04 | 0.05 | n/a | n/a |
| 31 | Fensulfothion | 0.998 | 0.058 | Not specified | 60.9 | 74.2 |
| 32 | p,p’-DDD | 0.998 | 0.061 | 0.2 | 78.5 | 73.8 |
| 33 | Sulprofos | 0.998 | 0.039 | Not specified | 94.7 | 108.8 |
| 34 | p,p’-DDT | 0.999 | 0.026 | 0.2 | 145.3 | 97.9 |
| 35 | Pyridafenthion | 0.998 | 0.065 | Not specified | 56.3 | 58.5 |
| 36 | Phosmet | 0.999 | 0.036 | Not specified | n/a | n/a |
| 37 | EPN | 0.992 | 0.135 | Not specified | n/a | n/a |
| 38 | Methoxychlor | 0.984 | 0.077 | 0.05 | 107.1 | 89.6 |
| 39 | Phosalone | 0.999 | 0.022 | 0.1 | 66.2 | 53.7 |
| 40 | Azinphos-methyl | 0.999 | 0.073 | 0.3 | n/a | n/a |
| 41 | Pyrazophos | 0.999 | 0.026 | Not specified | 86.5 | 57.8 |
| 42 | Azinphos-ethyl | 0.998 | 0.050 | 0.1 | 49.1 | 45.4 |
| 43 | Pyraclofos | 0.999 | 0.029 | Not specified | 52.4 | 41.5 |
| 44 | Coumaphos | 0.999 | 0.025 | Not specified | 61.8 | 45.3 |
| 45 | Decachlorobiphenyl | 0.999 | 0.029 | Not specified | 45.8 | 40.0 |
Conclusion
Chromato-mass-spectrometry system is capable of detecting pesticide residues in very complex tobacco matrix with a high degree of confidence.
Sophisticated algorithm of automated set-up and fine-tune of an optimal SIM-method running behind easy to use SIM-Wizard tool, complemented with SIM- library data for selection of optimal SIM ions allows an analyst to reach the levels of detection of pesticides following recommendation by ACAC directive. [2]
References
- Anastassiades M., Lehotay S.J, Stajnbaher D, and F.J. Schenck //The Journal of AOAC International 86. P. 412–431(2003).
- CORESTA Guide No. 1 – The concept and implementation of CPA guidance residue levels, July 2013. Agro-Chemical Advisory Committee (ACAC) of the Cooperation Center for Scientific Research Relative to Tobacco (CORESTA), Paris, France.
Guidance Document on the Estimation of LOD and LOQ for Measurements in the Field of Contaminants in Feed and Food, 2016
