Measurement of polychlorinated biphenyls in different high consumption canned foods, using the QuEChERS/GC-MS method

Highlights • PCBs were investigated in canned foods samples using GC-MS technique.• Health risk assessment was calculated according to per capita Iranian consumption.• In all samples, the amount of PCBs was lower than the standard.• Canned food does not potential health risks to Iranian consumers.

Various sources that cause the accumulation of dioxins and PCBs in the environs include woodland fires, pesticides, oil factories, trucks, dyes and volcanic outbursts (Saktrakulkla et al., 2020).But it can be stated that the main and biggest sources of emission of these compounds are the burning of household waste, waste of municipal, waste of medical, fire in landfills, and agricultural and fires of forest (Stafilov et al., 2011).
Consequently, PCBs evaporate and enter the air, and this pollution can be deposited on plants and soil, and can also contaminate agricultural crops and animal fodder, and then be transferred to livestock (B.D. Škrbić & Marinković, 2019;B. D. Škrbić, Marinković, Antić, & Gegić, 2017).
PCBs and dioxins may enter the body (human) through skin absorption, ingestion, and inhalation.PCBs and dioxin compounds are stable in terms of stability in the environment and biological conditions, and when humans are exposed to these compounds for a long time, they can cause chronic poisoning (Jirdehi, Qajarbeygi, Hosseini, & Babaei, 2014).Through the atmosphere, these compounds can be deposited on the cover of leafy plants and on the soil, which acts as a natural sink (Adekunte, Tiwari, & O'Donnell, 2010;Schecter, Birnbaum, Ryan, & Constable, 2006).
The presence of any PCB congener in any food item has implications for human health because each PCB congener has specific physicochemical properties that affect bioaccumulation, bioavailability, and toxicities.PCBs cause adverse human health effects by disrupting immune, reproductive, nervous, and endocrine systems (van den Berg et al., 2017).The IARC (International Agency for Research on Cancer) and World Health Organization (WHO) have categorized PCBs as 2A class, probable carcinogenic congeners (Comai, 2020).
Canned foods are a staple of many people's diets in today's world.A 2013 survey of more than 1,000 Americans found that more than 60 percent of respondents stated eating canned foods at least once or twice a week.As mentioned before, the consumption of canned food is increasing for several reasons.More precisely, the per capita consumption of canned food in the world is 1.1/can/week or 22.6 kg/person/ year and per capita consumption of canned fish in the world is 2.2 kg/ person/year (Comerford, 2015).In general, canned foods include canned meats such as tuna, canned vegetables such as beans, and canned mixes such as stews.According to the research that was mentioned earlier, polychlorinated biphenyl compounds are present in all sources of canned food like vegetable, animal.
In recent years, there have been several researches in the measurement of fish, or other plant foods in the field of measuring these compounds (Biljana, 2008;B. Škrbić, 2007;B. Škrbić & Đurišić-Mladenović, 2007), but so far, these stable compounds have not been measured in canned food.
Considering the pollution of the food chain with PCBs and the increasing trend of canned food consumption, there is no comprehensive research on the measurement of PCBs in canned foods in Iran.Therefore, for the first time in Iran, the present research was conducted to measure polychlorinated biphenyls in various widely consumed canned goods using the QuEChERS/GC-MS method and to assess the health risk.

Collection of sample
Forty five canned food samples were analyzed.These canned foods were purchased from markets of Tehran, Iran, and transferred to the laboratory (June to July 2022).
At first, the statistics of the most consumed canned food in each vegetable, meat and mixed group were obtained from the Food and Drug Organization of Tehran city.The most consumed canned meat was tuna (fish meat, vegetable oil and

Sample preparation
The canned food samples were ready by a QuEChERS modified method (Yaminifar et al., 2021).First, 5 g of sample was regimented and poured into a 50 mL glass tube, and then 10 μL of PCB29 (internal standard) with a 10 ng/mL concentration and 20 mL of acetonitrile were added and shaken (5 min).
In the next stage, 6 mg of miscellaneous powder of MgSO 4 and NaCl, with 4 to 1 ratio was added to the tube of glass and then it was vigorously shaken by hand and shaker and centrifuged at 5000 rpm for ten minutes.
Afterwards, 10 mL of the overhead solution was poured into a glass vial (15 mL) and the tube (glass) was placed in the freezer at − 20 • C for fifteen minutes.
Finally, it was poured into a tube (glass) and shaken for thirty seconds with a shaker and kept stable for 5 min and 1 μL was poured into the GC-MS apparatus (Rutkowska et al., 2018).

Conditions of instrumental analytical
For the measurement of PCB analytes, a GC equipment (model of 6890; Agilent Technologies, Palo Alto, CA, United States) with a mass quadrupole spectrometer (model of 5973) was applied.The GC-MS conditions were: the gas of carrier: helium (99.9 %); the mode of injection: splitless; the volume of injection: 2 µL; the temperature of injector: 290 • C; rate of flow: 1 mL per minute.Ii this study, the type of column was capillary (HP5-MS; thickness of film: 0.25 µL; length of column: thirty meters; internal diameter: 0.25 mm).Also, the temperature program of oven: temperature of initial: 90 • C, isothermal: 2 min, rate of first: five centigrade degree/meters to 280 • C and isothermal: three minutes.Finally, the SIM (selected ion monitoring) mode, was applied for the detector at the 70 eV EI (electron ionization).

Methods quantification
The fundamental parameters for PCBs evaluates were validated: LOQ (limit of quantification), recovery percent, linearity, precision, accuracy and LOD (detection of limit).To the linearity of methods, by adding different levels of standard (0.10-40 ng/mL) on dissimilar days (in triplicate) were prepared 5 samples (in series) and then injected them (in duplicate) into the GC-MS equipment.Additionally, for each triplicate were prepared 2 samples of control (negative), which one of them was without matrix and the other was without sample.For each of the evaluated PCB analytes, the achieved data were applied to draw a curve of calibration.In this study, using the assessment of the RSD (relative standard deviation) percent, the precision was estimated, moreover identified as the CV (coefficient of variation), which was lower than eighteen percent, representing the method effectiveness for repeatability (intra-day) and for intermediate accuracy (inter-day).For repeatability, by the levels with eight nomogram per milliliters of each PCBs analytes, 5 samples (equally) were prepared under analysis.According to the proposed procedure, these samples (all on the same day) were assessed, from phase of extraction until investigation by the GC-MS equipment.For the procedure accuracy, the method of standard addition was applied, including adding different identified quantities of the standards (certified) of each PCB into the matrix (before the sample preparation).For each PCB analyte, by the standard addition (in triplicate) at levels of 20, 10, 1, 0.1, 0.05and 0 ng/mL were prepared 6 samples and the determined quantities were associated with the amounts added.Finally, by the procedure of average blank value were evaluated the LOD and LOQ parameters.
The ILCR from exposure to PCBs congeners (a probable human carcinogen) computed by Eq. (2) (Yijin Kang et al., 2020): Where ILCR is incremental lifetime cancer risk, EDI is estimated chronic daily intake and SF is carcinogenic slope factor of oral intake PCBs by USEPA's (United States Environmental Protection Agency) guidance (2 mg/Kg b.w. per day) (EPA, 2003).Also, the 95th percentile values of the risk were presented in order to distinguish significant risk.In this study, MCS (Crystal Ball v 11.1.2.4.600 software) was applied to calculate the probable carcinogenic risk and non-carcinogenic attributes (EDI and ILCR).This method was introduced by USEPA to analyze uncertainty for risk management.

Statistical analysis
Statistical Analysis descriptive statistics (minimum and maximum, mean and standard deviation) Pre-test and Post-test mean were taken and compared by SPSS version 20.The Kolmogorov-Smirnov test was used to determine the distribution of the study parameters.The Kruskal-Wallis test was applied to determine the significance between groups.The principal component analysis is a multivariate technique and is widely used in food contamination with the aim to find out the associations between contaminants and food products (Heydarieh et al., 2020;B. D. Škrbić, Marinković, & Spaić, 2020).To further analyze the relationship between the type and quantity of polychlorinated biphenyls in different canned foods, we performed a heat map analysis using Clustvis software (https://biit.cs.ut.ee/clustvis/) (Moradi et al., 2021).1.In this study, LOD, LOQ, recovery and RSD ranged from 0.06 to 0.32, 0.18 to 1.07 ng/g, 97.05 to 102.5 % and 5.66 to 9.50 %, respectively.Compared with our study, Fathabad et al. analyzed PCB compounds by using Soxhlet Extractor with HRGC/HRMS and expressed LOQ for all DL-PCBs ranged between 0.03 and 0.09 pg/g fat, also, in other study by Kiani et al. that used modified QuEChERS/ GC-QqQ-MS/MS method, they expressed the LOQs, LODs, recovery, and RSD for the PCB analytes were 0.180-0.360,0.06-0.12ng/g fat, 97.45-102.63 %, and 6.33-8.86%, respectively, furthermore, in other study by Shahsavari et al. that used modified QuEChERS extraction and GC-QqQ-MS/MS method they stated the LOD, LOQ and recovery for the PCB analytes were varied 0.04-0.16,0.132-0.482ng/g fat, 5.2-9.2 and 95.5-107.2%, respectively, and finally in study of Shariatifar et al. that used modified QuEChERS extraction and GC-QqQ-MS/MS method, they stated the recovery, LOD and LOQ of NDL-PCBs were varied 93.22-109.19%, 0.04 to 0.14 and 0.120 to 0.420 ng/g, respectively (Fathabad et al., 2020;Kiani et al., 2023;Shahsavari et al., 2022;Shariatifar, Maftahi, Tooryan, & Arabameri, 2022).

Assessment of PCBs in canned foods
The expressive statistical analysis of gained values, demonstrates the median, minimum and maximum of each analyte of PCBs (Table 2).The Kruskal Wallis test showed that PCB 18, PCB 28, PCB 44, and PCB 101 the concentration varies significantly in different types of canned foods (p < 0.05) (Table 2).The highest and the lowest amount of NDL-PCBs in canned fish, which is known as tuna, belongs to PCB 52 with amounts of 0.56 and 0.18 ng/g fat, respectively.Continuously, PCB 52 had the maximum amount of NDL-PCBs in comparison with other indicator PCBs, in pasta source (Max.:0.27,Min.: 0.00 ng/g fat), haricot (Max.:0.88,Min.: 0.10 ng/g fat) and lentiform (Max.:0.47,Min.: 0.43 ng/g fat).However in eggplant canned samples, PCB 101 was the compound with the highest amount (Max.:0.0.53,Min.: 0.24 ng/g fat).
According to the mentioned standard in previous study (Commission Regulation (EU) NO1259/2011) the limit of acceptable for PCBs in herbaceous, meaty and mixed canned foods are 40 ng/g (EU, 2011; Yaminifar et al., 2021).
According to the outcomes of present study, the content of all considered polychlorinated biphenyls compounds in canned foods were lower than the limitations of standard.The outcomes displayed the mean levels of total NDL-PCBs in canned foods for PCB 28,52,101,138,153 and 180 was 0.06,0.27,0.36,0.19,0.11,0.14respectively, with a maximum concentration of 0.36 ng/g fat belonging to PCB 101.The PCB with the highest mean was PCB 52 (0.27 ± 0.20 ng/g fat) and the concentration of maximum was 0.88 ± 0.20 ng/g fat.The minimum average were PCB 28 and 138 (not detected).
In this study, a comparison with other studies is presented in Table 3.Generally, the amount of PCBs in present study, in comparison with the others, concluded that environmental contaminations of main sources like air, soil and water with pesticides, oil products, pollutants of trucks, paints and demolition of houses can lead to the agricultural crops and animals pollution.This is more serious when animal farms are near The average level of PCDD/Fs + DL-PCBs in vegetable oils is 0.18 pg/g and the mean level of indicator PCBs in these products was 1.3 ng/g.(Baars et al., 2004).

Yagüe et al. 2005 pesticides and PCBs
Virgin olive oil Only one organophosphate pesticide was identified in one sample, the level of which was 10 ng/g chlorine pesticides were also detected in 5-47 % of the samples in very low concentrations.(Yagüe et al., 2005).The mean concentration of these pollutants was lower than the standard limit.The concentration of human exposure to PCBs in entire products was below 1 ng/g body weight per day, considering a 70 kg in men or 57 kg in women.(Adenugba et al., 2008).total NDL-PCBs was ranged 0.17 to 2.89 ng/g ww (Biljana, 2008).

Assessment of health risk
A health risk assessment was made according to the content of PCBs in different high-consumption canned foods.The predicted health risk of PCBs in the canned foods sold in the market of Iran is presented in Table 4.Among all PCBs, health risk was found to be associated with PCB 18, PCB 52, PCB 101, PCB 141, PCB 138, PCB 180, and PCB 194 in the canned foods sold in the Iranian market.Van den Berg et al., (2017) founded that PCBs posed a risk to health of human even in moderate concentration, the continuous intake could cause health hazards in the long term (van den Berg et al., 2017).The predicted EDI of PCBs in the canned foods sold in the Iranian market is showed in Table 3.This outcomes displayed that the EDI values were below the current tolerable daily intake of PCBs (TDI < 10 ng kg − 1 BW day − 1 ).A higher EDI index was detected in the children.The contributions by concentration to EDIs were 30 and 23 % for lentiform and fish canned, respectively.The level of each congener pattern showed in Fig. 1.In lentiform, the highest quantity of EDIs was PCB 180, and the second was PCB 52.
In this way, the EDI of PCBs in some crop products was assessed to be 172.2ng/kg per day for the population in Serbia (Biljana, 2008).Kang et al. (2020) founded that the PCBs in food samples in China and displayed the EDI of the ΣIndicator-PCBs was 26.47 ng/kg per day, which was upper than our research (Yijin Kang et al., 2020).While, Chung et al. (2018) stated that the lesser and higher bound exposures to indicator PCBs was 0.68 and 1.38 ng/kg day in the food samples, respectively (Chung, Lau, & Chu, 2018).
A MCS method or multiple probability simulation was executed to quantitatively estimate the potential oral cancer risk for children and adults.Estimated carcinogenic and mutagenic risk values of PCBs in canned foods sold in the market of Iran are presented in Fig. 2. The rank order canned samples based on the ILCR index (95 %) in adults was Lentiform (7.05E-8) > canned fish (5.73E-8) > Eggplant (5.38E-8) > Haricot (4.33E-8) > pasta source (2.06E-8); and in children was Lentiform (3.40E-7) > canned fish (2.72E-7) > Eggplant (2.44E-7) > Haricot (2.06E-7) > pasta source (9.83E-8).The risk of cancer PCBs for the goal population was much lesser than the US EPA's acceptable level of 10 − 6 , indicating an acceptable cancer risk in consumer exposure to the PCBs.Regardless, this value is an issue to changes with an increase in exposure period and frequency.Given the potential value of risk assessment for addressing ILCR value, food products and exposure to PCB could be an important role in governing potential cancer risks in humans.

Analysis according to heat map outcomes
Heat maps by intuitive visualization of data and results can be useful to present a very brief view of a specific purpose.The hierarchical clustering analysis dendrogram of PCBs in the canned foods is presented in Fig. 3.As the Euclidean space reduced, the samples shows an upper correlation together.As can be observed, the samples were well divided   between the amount and type of PCBs in canned food.

Conclusion
For the first time, this research examined PCB residues in different canned foods in Tehran.The samples were examined by modified QuEChERS method and GC-MS technique that performed very well in this field.In conclusion, herbal food cans had higher concentrations of PCBs in comparison with canned fish, which this difference can be referred to the source of different types of PCBs and the food ingredients.By comparing the outcomes of this research with international standards, it was found the amount of this pollutant in canned food is  completely less than the existing standard.The heat map and PCA demonstrated that the grouping of our different samples based on their PCBs profile was closely related and distinctly different in the canned samples.The MCS results displayed the cancer risk (ILCR) due to oral PCBs estimated in safe limit (CR > 1E-6).The PCBs exposure in both groups was under the tolerable daily intake (TDI < 10 ng kg − 1 BW day − 1 ).One of the limitations of this project is the lack of study on all the canned foods available in Iran, which was due to the limitation of financial resources.Finally, it is suggested to conduct research on other canned foods produced in Iran as well as imported products in the future. -.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.Comparison of the most and least contribution to overall EDI in canned samples.

Fig. 2 .
Fig. 2. The results of simulation for EDI and ILCR of PCBs detected in canned samples.

Table 1
Characteristics of technique.

Table 2
Assessment of concentration of PCBs among different types of canned foods (ng/g).
In this table, zero means less than LOD.F. Vali Mohammadi et al.

Table 3
Comparison of the present study with other researches.

Table 4
Analysis of uncertainty for the EDI of studied PCBs in canned food samples.