Measurement of phthalate acid esters in non-alcoholic malt beverages by MSPE-GC/MS method in Tehran city: chemometrics

Phthalic acid esters (PAEs) are compounds that are used in the bottle as the main plasticizers. Therefore, the possibility of releasing phthalate esters into beverages is very high and there is a concern to consumer health and monitoring organizations. The aim of this research was to assess the phthalic acid esters (di-n-octyl phthalate (DNOP), butyl benzyl phthalate (BBP), dimethyl phthalate (DMP), diethyl phthalate (DEP) and dibutyl phthalate (DBP), bis(di-ethylhexyl) phthalate (DEHP), and total PAEs) in bottled non-alcoholic malt beverages (n = 120) by multi-walled carbon nanotubes were magnetized with iron (MWCNT-Fe3O4) using gaschromatography/mass spectrometry (GC-MS). The results showed that the highest and the lowest levels of total phthalate esters in samples were 9483.93 and 2412.50 ng/L, respectively. The mean of DEHP which has also been found to be carcinogenic in all samples was lower than 5944.73 ng/L. The highest concentration of DEHP in four samples was upper than 8957.87 ng/L. Perceived limit of detection (LOD) ranged from 13 to 30 ng/L and the limit of quantification (LOQ) ranged from 39 to 90 ng/L. Multivariate techniques and heat map visualization were used to assess the correlation among the type and levels of PAEs with the brand, color, product date, pH, sugar, volume, and gas pressure. Therefore, based on heat map and principal component analysis (PCA) results, the DEHP and total PAEs were the closest accessions, indicating that these variables had similar trends. Based on the results, it can be stated that due to the low average of total phthalate esters in non-alcoholic malt beverages, there is no serious health hazard of these compounds for humans.


Introduction
Annually large quantities of phthalic acid esters (PAEs) are used as plasticizers to increase the durability, workability, and flexibility of polymeric materials in polyethylene terephthalate (PET) packaging around the world (Arfaeinia et al. 2020;Harunarashid et al. 2017). Furthermore, these compounds can be present in many products such as inks, paints, cosmetics, and adhesives (Rafiei Nazari et al. 2018). The phthalates are not covalently bound to the polymer matrix composites, thus easily released and transport to the environment and finally can cause contaminate air, soil, water, and food products. However, other parameters may also be considered concerning the potential migration of PET such as the chemical structure of the additive (crystallinity, molecular weight), the plastic polymer type, solubility behaviors, and various storage conditions Jarošová 2006;Kouhpayeh et al. 2017). In recent years, food contamination with PAEs has become a problem of public concern. PAEs have been reported in beverages such as alcoholic and nonalcoholic malt beverages, soft drinks, meat, and meat products, fruits, vegetables, dairy products, cereals, vegetable edible oils, and other food products (Kiani et al. 2018;Liu et al. 2020). Alcoholic and non-alcoholic malt beverages have reached a significant level of consumption (even in muslim countries) in recent decades. In 2007, total world production was up to 1.79 × 10 11 L (an increase of approximately 27 L per person, an increase of 5.4% over 2006) (Ye et al. 2009).
PAEs are also widespread pollutants and appear to be the cause of cancer in humans and animals through defects in fetal, altered hormone levels, uterine damage and reduced sperm production and motility, Sertoli cell damage, and Leydig cell tumors, cryptorchidism. The Environmental Protection Agency of USA (USEPA) has listed PAEs as the important contaminant (Jarošová 2006;Moazzen et al. 2019;Shen et al. 2019). The most important PAEs include bis(diethylhexyl) phthalate (DEHP), di-n-octyl phthalate (DNOP), butyl benzyl phthalate (BBP), dimethyl phthalate (DMP), diethyl phthalate (DEP), and dibutyl phthalate (DBP). DEHP is one of the most important compounds, which causes liver cancer in rodents and humans (Liu et al. 2020;Sinha et al. 2019). DBP, BBP, DEHP, and some phthalate esters metabolites in animals have shown teratogenic effects. The phthalates are involved in the pathogenesis of asthma, allergic symptoms, contribute to airway remodeling and premature puberty in young girls (Harunarashid et al. 2017;Kiani et al. 2018). Concern about the health implications and ubiquitous presence of PAEs has prompted an increasing demand for reliable and fast analytical techniques that allow their quantification and detection at very low levels in alcoholic and nonalcoholic malt beverages (Ye et al. 2009).
Several analytical techniques have been developed to detect PAE in various samples packaged in plastics (especially PET), such as drinking water, beverages, and cosmetics. High-performance liquid chromatography (HPLC), gas chromatography (GC), and GC coupled with mass spectrometry (GC-MS) are commonly used for PAE (Aboul-Enein and Ali 2001, 2002a, 2002b, , Ali et al. 2006Dobaradaran et al. 2020;Moazzen et al. 2019;Nagaraja et al. 2001;Sanagi et al. 2015;Wu et al. 2014;Zawilla et al. 2002). Numerous preparation methods can be used to prepare samples for analysis of these compounds in different matrixes, such as single-drop microextraction (SDME), liquid-liquid extraction (LLE), solvent extraction (SE), and liquid-phase microextraction (LPME). LPME method based on the solidification of a floating organic micro drop (LPME-SFO), accelerated solvent extraction (ASE), stir-bar sorptive extraction (SBSE), solidphase extraction (SPE), and column chromatographic cleanup (CC), and these methods are expensive, insensitive, and timeconsuming Kiani et al. 2019;Roudbari et al. 2020;Shariatifar et al. 2020). Recently, a different type for solid-phase extraction (SPE) that is called MSPE or magnetic solid-phase extraction has been established. Magnetic adsorbents compared with nonmagnetic adsorbents can make the separation process faster and easier without any additional procedures such as filtration or centrifugation. The MSPE method can prevent the timeconsuming column operations seen in SPE (Kouhpayeh et al. 2017;Moazzen et al. 2018;Shen et al. 2019).
Considering the high consumption of non-alcoholic malt beverages in the daily food basket and the common supply of these beverages in plastic bottles, it is necessary to investigate PAE contamination in these beverages. To the best of our knowledge, so far, no study has been conducted using MSPE-GC-MS method to investigate PAE contamination in a variety of non-alcoholic malt beverages in Iran. Therefore, the aim of this study is firstly to fabricate magnetic adsorbents (MWCNTs-Fe 3 O 4 ) for easy and cheap extraction using GC-MS and MSPE method. Secondly, to measure these compounds in non-alcoholic malt beverages and compare them with international standards and other research and finally from the PCA to describe the interrelationships between variables and visualize the data patterns given in this research and the heat map to show the similar or vastly different expression status characteristics values were used.

Reagents and chemicals
Phthalate acid esters (DEHP, BBP, DBP, DEP, DMP, and DNOP) and other chemicals including ethanol, methanol, NaCl, and n-hexane were obtained from Sigma-Aldrich Company (St. Louis, MO, USA). FeCl3.6H2O, benzyl benzoate (internal standard), NaBH4, HCl, NaOH, and 3,5dinitrosalicylic acid (DNSA) were purchased from Merck Company (Germany). The diameter and length of multiwallet carbon nanotube (MWCNTs, Panchkula, India) were 30-60 nm and 5.0-30 mm, respectively. The grade of all the other used chemicals and solvents were analytical reagents. The stock solution of the analyzed contaminant was ready in methanol (100 mg/mL). Afterward, the solutions of phthalate acid esters working standard were prepared by consecutive dilutions of the stock solution (10-10,000 ng/L). To prepare the internal standard, 10 μg of benzyl benzoate was first dissolved in 10 mL of methanol. After preparing the internal standard solution, 100 μL of this solution was added to each sample. On the same day of the study, the quality control was prepared from diluted solutions of the stock standard. The prepared solutions were retained at 4°C and in the dark place, until analysis. In this research, all the laboratory glass dishes (before using) were washed with an Al 2 O 3 solution, and then they were immersed in acetone for 40 min and were washed with n-hexane as well as finally were dried (in the oven) for 5 h at 150°C.

Determination of sugar
In the first, samples with different dilutions were prepared from sucrose stock solution (1000 mg/dL) as standard. Afterward, 10 mL of non-alcoholic malt beverage samples were diluted with distilled water up to 100 mL as samples; 2 mL of each standard solution was added into a test tube and 2 mL of distilled water was inserted into a separate test tube (blank solution). Then, 2 mL of hydrochloric acid (6 M) was added and placed in boiling water bath for 10 min and then 8 mL NaOH (2.5 M) was added to adjusted pH 7.0. Afterward, 2 mL of 3,5-dinitrosalicylic acid (0.05 M) was added and cover the test tube with film and shake well to mix. Then placed in a boiling water bath for 5 min and then placed in ice water for 10 min. The time between 3,5-dinitrosalicylic acid solution addition and measurements should be the same for all test solutions. Finally, the absorbance of the five standards and references at 580 nm was measured using a UV/Vis spectrophotometer (Roig and Thomas 2003).

Gas pressure measurement
The gas pressure of the non-alcoholic malt beverage samples was determined using a digital gas meter at room temperature set at 25°C.

The pH measurement
The pH of the non-alcoholic malt beverage samples was determined using a digital pH meter at room temperature and pressure of 1 atm (Godwill et al. 2015).

Evaluate the properties of the prepared adsorbent
Phase identification of adsorbent was assessed by X-ray diffraction (XRD) (Philips, X'PertPro 2002) and characterization of elemental was assessed by energy-dispersive X-ray (EDX) (PHILIPS, S360), and morphological analysis was assessed by scanning electron microscope (SEM) (PHILIPS, and S360 Mv2300) ).

Sample preparation
In the beginning, non-alcoholic malt beverage samples (n = 120, as duplicated) were bought with five brands of most commonly used from a chain store in Tehran, Iran, and all samples were degassed with a bath of ultrasonic at room temperature for 20 min. Afterward, 10 mg of prepared adsorbent (MWCNT-Fe 3 O 4 ) was activated with solutions of methanol and water. To the 10 mL of each degassed non-alcoholic malt beverage samples, 10 mg of the activated adsorbent, 0.5 g of NaCl, and 100 μL I.S. were added and mixed strongly for 4 min to extract the contaminant compounds. Afterward, adsorbent was gathered with an exterior magnet to the side of the laboratory dish within the 90S and the other compounds in the mixture were discarded. Then, 2 mL of n-hexane was added to the adsorbent and vigorously mixed with the blender for 2 min, to elute contaminant compounds (PAEs) from the adsorbent. Later, by using an external magnet, the adsorbent was collected to the side of the laboratory dish, and the supernatant was moved to a vial. Desorption solvent was dehydrated with a mild stream of N 2 gas at room temperature and was maintained in a cold place such as a refrigerator. Finally, the dehydrated contents of the previous step were dissolved in 1 mL solvent (n-hexane), and 1 μL of the mentioned solution was injected into the GC-MS (Agilent 7890 N, MS (5975)). The optimization of the method was performed according to the 1 factor at a time method (Kiani et al. 2018;Moazzen et al. 2019).

Instrument conditions of GC-MS
The column of chromatographic was DB-5-J & W Scientific (30 m, 250 μm, 0.5 μm). Helium (He) was selected as the gas of carrier at 1 mL/min (ratio of the split of 50:1). Splitless was the mode of injection with an inlet temperature of 290°C. The program of GC-MS temperature was: 80°C retained for 2 min, 80-285°C at 7°C/min retained for 10 min. For the quantitative determination of PAEs compounds, the mode of selective ion monitoring was used. The device output of the retention times (RTs), quantitative and qualitative ions of six phthalate acid esters, and I.S. are presented in Table 1.

Validation of the analytical method
The validation of the method was performed based on the currently established guideline of FDA (U.S. Food and Drug Administration) for industries (Kouhpayeh et al. 2017;Moazzen et al. 2018). The validation for quantitative analysis of compounds in samples was shown by the coefficient of estimation (r 2 ) and linear ranges, the limit of detection (LOD), the limit of quantification (LOQ), precisions, and accuracies. The recoveries were within 94.2 to 104.3% for all the samples (Tables 2 and 3).

Structural relationship of parameters
PCA was carried out using the raw data obtained from nonalcoholic malt beverage samples (n = 120). PCA with the eigenvalues greater than 1.0 created a new measurement scale to the identification of the structural relationship of parameters. Further, a scree plot was generated to reveal the retained factors. Separate PCAs were performed for the type and levels of PAEs (DMP, DEP, DBP, DEHP, DOP, and total phthalate) in each non-alcoholic malt beverages (the brand, color, product date, pH, sugar, volume, and gas pressure). For a better understanding of the most significant contribution to the distribution PAEs levels among the different samples, the PCA was done by the software of SPSS (Arabameri et al. 2019;Heydarieh et al. 2020). Multivariate techniques were used to assess the correlation between the amount and type of PAEs levels with properties samples. Heat map analysis was used to analyze the correlation between samples online at https://biit.cs.ut.ee/clustvis/.

Statistical analysis
The outcomes were statistically analyzed by SPSS version 18 (SPSS Inc, Chicago, IL, USA) for Windows. Data analysis was done by the test of Kolmogorov-Smirnov, and tests of Kruskal-Wallis. The non-parametric Spearman correlation coefficient was used to study the relation between the different parameters. Statistical significance was a p-value of < 0.05.

Analytical validation method
To draw the calibration curves were prepared five different concentration levels of phthalate acid esters (DEHP, DBP, DEP, DNOP, BBP, and DMP) at the range of 10 to 12,000 ng/L (including 10, 100, 1000, and 12,000 ng/L). The correlation coefficient ranged from 0.9979 to 0.9997 and the detections limit (LODs) and quantifications limit (LOQs) for the target analytes were 13 to 30 ng/L and 39 to 90 ng/L, respectively ( Table 2). The recovery values of the 6 PAEs were 94.2 to 104.3% with the RSDs less than 7.6%. To control the quality, three levels (50, 500, and 5000 ng/L) from a mix of PAEs were prepared and were analyzed duplicate in several days. The inter-and intraday precision measured for 3 consecutive days in triplicate analyzes and they were lower than 7.8% and 8%, respectively (Table 3). The selectivity of the method was examined by analyzing 120 non-alcoholic malt beverages.

Characterization of MWCNT-Fe 3 O 4
The SEM image, EDX, and X-ray analysis of the magnetic adsorbents are presented in Figs. 1, 2, and 3. The graphs show that the placement of magnetic particles on the surface of multi-wallet carbon nanotubes is comparatively uniform and after filling with magnetic particles, the surface of the multiwallet carbon nanotubes became rougher (Fig. 1). Besides, there is no significant modification detected in the surface of construction of the magnetic adsorbents after the extraction procedure. A growth in adsorbent diameter is apparent. Also, we concluded that the adsorbent exhibited a chain-like morphology without apparent collection. This conducts to a  high ability of adsorption of the adsorbent. These results are similar to study of the Moazzen et al. (2018). The chemical compounds of the prepared adsorbent were assessed by using EDX analysis. The spectrum of EDX displayed oxygen (O), iron (Fe), and carbon (C). The atomic C, Fe, and O ratio (64.6, 22.7, and 12.7, respectively) as the principal elements in the structure of the prepared adsorbent confirmed the quantitative representation of the existence of Fe 3 O 4 nanoparticles on the MWCNT surface (Fig. 2).
By XRD analysis, the construction of multi-wallet carbon nanotubes and MWCNT-Fe 3 O 4 composites were more confirmed. The patterns of XRD of adsorbent are exhibited in Fig.  3. The strong diffraction peaks at 2θ = 31/77°and 2θ = 45/52°w ere shown MWCNTs and Fe 3 O 4 , respectively. The get XRD outcomes displayed that the Fe 3 O 4 were efficaciously coated on the surface and texture of multi-wallet carbon nanotubes using a co-precipitation technique.

Evaluation of phthalate esters in non-alcoholic malt beverages bottled in PET bottle
Our results showed ( Table 4) the mean of total PAEs was 6657.28 ± 1600.9 with the range of 2412.5 to 9483.93 ng/L. The highest mean of PAEs was DEHP (5944.73 ± 2518.14 ng/ L) that was lower than the EPA and WHO-EU standard level and lowest mean of PAEs was BBP (non detected). The mean concentration of these compounds were DEHP > DBP > DEP > DnOP > DMP > BBP. There was a significant difference between the groups in terms of phthalate esters (p < 0.05). According to Table 4, the mean of all compounds in all the samples were less than the standard defined by the EPA (6000 ng/L) and WHO-EU (8000 ng/L) in drinking water, but the highest of DEHP in the sample was higher than the EPA and WHO-EU standards (8957.87 ng/L). The mean value and range of pH in samples were 3.3 (3.18-3.53) and sugar was 8.7 (per  100g) (4.4-11.9) and gas pressure was 1140 (830-1398) millimeters of mercury (mmHg). There was a significant difference between the groups in terms of sugar content and gas pressure (p < 0.05). Besides the update on phthalate acid esters in nonalcoholic malt beverages, one of the main objectives was to assess the relations between the evaluated variables. The correlation values obtained by multivariate analysis (Spearman correlation) between the variables in all sampeles are showed in Table 5. Significant negative correlations between the phthalate acid esters related variables were obtained, being higher between DEP with gas pressure (r = −.781), volume (r = −.351), sugar (r = −.331), and DBP with gas pressure (r = −.765) (p < 0.05). Significant positive correlation was found between DMP and date in all samples (p < 0.05). However, reduce the sugar of non-alcoholic malt beverages could imply an undesired impact on physicochemical (optical, rheological properties, total soluble sugars) and sensory properties (good consumer acceptance) (Hajieghrary and Homayouni-Rad 2016). In similar studies by Moazzen et al. (2018), the highest concentration of PAEs in carbonated soft drinks in PET bottle was DEHP, that it was higher than the standard level (STL) in the 4 samples (14,008, 9301, 9201, and 6766 ng/L) and other PAEs compounds were lower than the STL, which was somewhat similar to our study . Also, Xu et al. (2020) showed 3 PAEs compounds (DBP, DMP, and DEP) were measured in ten common brands of bottled water (made of PET) in Beijing (China). The total concentration of PAEs ranged from 0.101 to 0.709 ng/ L, and DBP contained the highest contents (0.511 ng/L), which was lower than comparing our study (Xu et al. 2020). Besides, Vincenzo Russo et al. (2014)  (1900 ng/L) are present only in 1 beer sample that was higher than comparing our study (Russo et al. 2014).
The important physical and chemical parameters that can influence the migration of PAEs into food products are composite food samples, pH, volume, storage time, and temperature (Arfaeinia et al. 2020;EFSA Panel on Food Contact Materials et al. 2019;Rafiei Nazari et al. 2018). They reported that migration of PAEs in beverages bottled increased significantly with decrease pH and volume. Nazari et al. (2018) with determining the migration modeling of PAEs from nonalcoholic beer bottles showed that storage duration increased and temperature resulted in an increase in migration level ranging from 600 to 2900 ng/L, which was lower than our study (Rafiei Nazari et al. 2018). Also, Carnol et al. (2017) identified 6 PAEs of Luxembourgish beer stored in various containers (aluminum, glass, and can bottle), and total PAEs were found in all samples at levels of 61,560 ng /L, which was upper than the present study (Carnol et al. 2017). Balderas-Hernández et al. (2020) identified PAEs in tequila beverages (alcoholic bevereage) and showed that 22% of samples (65 samples) lacked PAE. DINP (1,640,000-3,430,000 ng/L), BBP (50,000-2,910,000 ng/L), and DEP (130,000-270,000 ng/L) were found in 5 (1.69%), 37 (12.54%), and 11 (3.73%) samples, respectively. However, these levels were not higher than the maximum standard level of PAEs for alcoholic beverages. DEHP (30,000-4,640,000 ng/L) and DBP (10,000-2,200,000 ng/L) were found in 224 (75.93%) and 96 (32.54%) samples, from them, just 15 (5.08%) and 10 (3.39%) samples, respectively, exceeded the maximum standard levels for alcoholic drinks. Bis(2-ethylhexyl) phthalate was the most repetitious PAEs detected in tequila and detected concentrations of DEHP were two times higher in ultra-aged tequilas compared to those in white tequilas (Balderas-Hernández et al. 2020).
The data comparison of this research with other articles shows differences that can be due to reasons such as alcoholic or non-alcoholic beverage, use of plastic or other containers, contamination of raw materials or secondary contaminants during the production process, carbonated or non-carbonated beverage and also the amount of gas in the carbonated  PAEs in the range of 140-1100 ng/L in some beers (alcoholic), 1 PAE in the range of 1200-1500 ng/L in 3 grape juices, and 6 PAEs in several cider samples, in the range of 300-2100 ng/L, which was less than the present study (Rodríguez-Ramos et al. 2020). March and Cerdà (2015) displayed that DEHP and DBP in low-alcohol beer were 400 and 2200 ng/L, respectively (March and Cerdà 2015). The DEHP was lower than our study and DBP was higher in our study. Ye et al. (2009) measured the PAEs in alcoholic beer bottled in China and indicated concentration of total PAEs was 6220 to 7760 ng/L and concentration of DEHP was 4270 to 5240 ng/L that was lower than our study (Ye et al. 2009). Cinelli et al. (2013) measured the PAEs in alcoholic winebottled in Italy and indicated concentration of DEHP was 1150 to 2660 ng/L that was lower than our study (Cinelli et al. 2013). Wang et al. (2017) showed DEHP in alcoholic beverages (liquor) was ranged from 618,200 to 1,089,000 ng/L, which was higher than our study (Wang et al. 2017). Pang et al. (2020) presented DEHP in alcoholic carbonated beverage and beer was not found that was lower than our study (Pang et al. 2020).Various conditions such as the concentration of compounds in the beverage, beverage pH, shelf life, storage temperature (30-60°C), gas pressure, and sun exposure may affect the PAE concentration of the beverage in PET packaging (Bhunia et al. 2013;Giuliani et al. 2020;Xu et al. 2020). From the results of phthalic acid esters (PAEs) concentration in this study, in comparison with other studies, it can be concluded that water pollution (used in the preparation of malt bevareage), plastic materials such as tanks, piping, and other elements during the production process, and factors such as storage temperature, pH, gas pressure, shelf life, sunlight exposure, and concentration of packaged food are effective on phthalate release. In addition, the amount of plasticizer used in the packaging and use of inappropriate and non-standard raw materials in bottle making can also affect its release. In general, the average concentration of these compounds in non-alcoholic malt beverage samples in Tehran was lower than the mentioned standards, so they probably do not pose a risk to human health.

Structural relationship of parameters
Multivariate techniques and heat map visualization were applied to evaluate the correlation between the type and levels of PAEs with the brand, color, product date, pH, sugar, volume, and gas pressure. Consequently, based on heat map and PCA results, the bis(2-ethyl hexyl) phthalate (DEHP) and total PAEs were the closest accessions, indicating that the DEHP made up most of the total PAE. Heat map clustered the 40 non-alcoholic malt beverages using correlation distance and average linkage, reflecting similarities and relationships among the type and levels of PAEs samples. The heat map grouped samples into two major clusters and two sub-clusters (Fig. 4). The first cluster includes DEHP and total PAEs, and second cluster contains two sub-groups with the brand, color, product date, pH, sugar, volume, gas pressure, DMP, DEP, DBP, BBP, and DNOP. In consideration of the different levels of PAEs in samples, these quantities can be expressed individually into several subgroups based on their levels and obtained a linked contribution to simple distinction in nonalcoholic malt beverages. The DEHP and total PAEs were the closest accessions, indicating that these variables had similar trends. Quantitative results obtained for the type and levels of PAEs were used to PCA to investigate the most significant contribution among non-alcoholic malt beverages. The compounds included samples brand, color, product date, pH, sugar, volume, gas pressure, and levels of PAEs (DEHP, DMP, DNOP, BBP, DBP, and DEP). The dependence relations between different PAEs can be found in Fig. 4, the graph subset illustrates a visual representation of the relations among different PAEs. The nearest neighbor among the properties, the greater significant relationship existed among the dependent variable. As shown in Fig.  5, the first five principal components accounted for 77.86% of the data variance in all samples, and their contribution rates were 26.80%, 15.87%, 14.27%, 11.46%, and 9.43%, respectively. The DEHP and total PAEs were the closest accessions, indicating which these variables had similar trends. The components can be used as the correlation of identical variables. The total phthalate, DEP, DBP, and BEHP had a high positive correlation with PC1, while had a negative correlation with sugar content and gas pressure. The results illustrated in Table 6 where the sugar, pH, and the brand had a positive correlation with PC2, while had a negative correlation with gas pressure and DOP. The grouping of the item in PCA plots illustrated potential migration, which is important for their potential to a health hazard of humans.

Conclusion
In first magnetic adsorbent was ready using a simple, sensitive, and cost-effective method for measuring 6 PAEs (DEHP, DMP, BBP, DNOP, DEP, and DBP) from non-alcoholic malt beverages. Moreover, SEM, EDX, and XRD analyses were used to describe the generated MWCNT-Fe 3 O 4 . Analysis of non-alcoholic malt beverage samples evaluated in Tehran and showed that the concentration of none of the PAEs released from the bottles was upper than the standard levels (6000 ng/L by USEPA). The correlation analysis focuses on the numerical relationship among types and levels of PAEs in all samples. Comparing relationships between PAEs levels from different non-alcoholic malt beverages showed that the bis(2-ethyl hexyl) phthalate (DEHP) and total PAEs were the closest accessions. Furthermore, the results of heat map visualization indicated that the relationship between PAEs levels under different samples was properly diagnosed. Based on the results, it can be stated that due to the low average of total phthalate esters in non-alcoholic malt beverages, it may not pose a serious threat to human health. However, these levels (DEHP) were in 4 (3.33%) samples higher than the maximum standard level of PAEs for non-alcoholic malt beverages.