Analytical Scheme for Simultaneous Determination of Phthalates and Bisphenol A in Honey Samples Based on Dispersive Liquid–Liquid Microextraction Followed by GC-IT/MS. Effect of the Thermal Stress on PAE/BP-A Levels

In this paper, an analytical protocol was developed for the simultaneous determination of phthalates (di-methyl phthalate DMP, di-ethyl phthalate DEP, di-isobutyl phthalate DiBP, di-n-butyl phthalate DBP, bis-(2-ethylhexyl) phthalate DEHP, di-n-octyl phthalate DNOP) and bisphenol A (BPA). The extraction technique used was the ultrasound vortex assisted dispersive liquid–liquid microextraction (UVA-DLLME). The method involves analyte extraction using 75 µL of benzene and subsequent analysis by gas chromatography combined with ion trap mass spectrometry (GC-IT/MS). The method is sensitive, reliable, and reproducible with a limit of detection (LOD) below 13 ng g−1 and limit of quantification (LOQ) below 22 ng g−1 and the intra- and inter-day errors below 7.2 and 9.3, respectively. The method developed and validated was applied to six honey samples (i.e., four single-use commercial ones and two home-made ones. Some phthalates were found in the samples at concentrations below the specific migration limits (SMLs). Furthermore, the commercial samples were subjected to two different thermal stresses (24 h and 48 h at 40 °C) for evidence of the release of plastic from the containers. An increase in the phthalate concentrations was observed, especially during the first phase of the shock, but the levels were still within the limits of the regulations.


Introduction
In recent years, the goal of researchers in the field of analytical chemistry has been both to develop really sensitive analytical protocols, but also, and above all, simplify analyses considering very important factors such as time and cost. For instance, the analysis time is currently considered the limiting process of the analyses. Basically, most people are only interested in purely economic costs without thinking that they are also connected to environmental ones. The main idea is therefore to use miniaturized analytical techniques for reducing the costs whereas, with regard to the time factor, efforts are being made to develop techniques that accelerate the sample treatment. Over the years, it has passed from liquid-liquid extraction (LLE) [1] to a solid-phase extraction (SPE) [2,3] up to developing a new technique, dispersive liquid-liquid microextraction (DLLME), which uses very few quantities of solvents when compared to those of the LLE and which eliminates the problems connected with plastic honeycombs has spread to reduce the risk of melting the wax during the hottest summer seasons, with consequent loss of the crop. The possible presence of PAEs and BPA in honey is also related to the production processes that may involve direct contact with unsuitable plastic (e.g., single-dose plastic packs).
The determination of PAEs and BPA presents considerable difficulties related mainly to their low concentration, which is currently difficult to determine even with very sensitive instruments [25]. To overcome this problem, it is necessary to develop an analytical method that provides for the extraction and pre-concentration procedure. In this paper, ultrasound vortex assisted DLLME (UVA-DLLME) was used as the extraction technique [26], whereas Gas Chromatography coupled with Ion Trap Mass Spectrometry (GC-IT/MS) was used as the analysis technique. The method developed, validated, and experimentally consolidated in the laboratory takes into account all the parameters that influence the extraction of the analytes. For this purpose, the extraction solvent, the ionic strength of the solution, the pH, and the extraction volume were analyzed and studied.
The developed method was applied to six real samples (i.e., four single-use commercial ones available in the Italian market, one home-made honey nectar sample, and one sampled from honey produced in plastic honeycomb). This last sample is interesting because the plastic honeycomb represents a new frontier in honey production, according to the author's knowledge, this measurement is the first analysis for understanding the migration of such compounds from the honeycomb to the honey. Finally, the four commercial samples were subjected to thermal stress (i.e., exposed to 40 • C for 24 and 48 h) for evidencing the effects of the heat on the plastic containers.

Experimental Design
The investigated compounds are listed in Table 1, along with the main analytical parameters. Table 1. List of compounds investigated in this study. Abbreviations-CAS numbers, formulas, molecular weights (MW), selected ion monitoring (SIM) peaks, solubilities and log K ow , median lethal doses (LD 50 ), acceptable daily intakes (ADIs). hottest summer seasons, with consequent loss of the crop. The possible presence of PAEs and BPA in honey is also related to the production processes that may involve direct contact with unsuitable plastic (e.g., single-dose plastic packs). The determination of PAEs and BPA presents considerable difficulties related mainly to their low concentration, which is currently difficult to determine even with very sensitive instruments [25]. To overcome this problem, it is necessary to develop an analytical method that provides for the extraction and pre-concentration procedure. In this paper, ultrasound vortex assisted DLLME (UVA-DLLME) was used as the extraction technique [26], whereas Gas Chromatography coupled with Ion Trap Mass Spectrometry (GC-IT/MS) was used as the analysis technique. The method developed, validated, and experimentally consolidated in the laboratory takes into account all the parameters that influence the extraction of the analytes. For this purpose, the extraction solvent, the ionic strength of the solution, the pH, and the extraction volume were analyzed and studied.
The developed method was applied to six real samples (i.e., four single-use commercial ones available in the Italian market, one home-made honey nectar sample, and one sampled from honey produced in plastic honeycomb). This last sample is interesting because the plastic honeycomb represents a new frontier in honey production, according to the author's knowledge, this measurement is the first analysis for understanding the migration of such compounds from the honeycomb to the honey. Finally, the four commercial samples were subjected to thermal stress (i.e., exposed to 40 °C for 24 and 48 h) for evidencing the effects of the heat on the plastic containers.

Experimental Design
The investigated compounds are listed in Table 1, along with the main analytical parameters. hottest summer seasons, with consequent loss of the crop. The possible presence of PAEs and BPA in honey is also related to the production processes that may involve direct contact with unsuitable plastic (e.g., single-dose plastic packs). The determination of PAEs and BPA presents considerable difficulties related mainly to their low concentration, which is currently difficult to determine even with very sensitive instruments [25]. To overcome this problem, it is necessary to develop an analytical method that provides for the extraction and pre-concentration procedure. In this paper, ultrasound vortex assisted DLLME (UVA-DLLME) was used as the extraction technique [26], whereas Gas Chromatography coupled with Ion Trap Mass Spectrometry (GC-IT/MS) was used as the analysis technique. The method developed, validated, and experimentally consolidated in the laboratory takes into account all the parameters that influence the extraction of the analytes. For this purpose, the extraction solvent, the ionic strength of the solution, the pH, and the extraction volume were analyzed and studied.
The developed method was applied to six real samples (i.e., four single-use commercial ones available in the Italian market, one home-made honey nectar sample, and one sampled from honey produced in plastic honeycomb). This last sample is interesting because the plastic honeycomb represents a new frontier in honey production, according to the author's knowledge, this measurement is the first analysis for understanding the migration of such compounds from the honeycomb to the honey. Finally, the four commercial samples were subjected to thermal stress (i.e., exposed to 40 °C for 24 and 48 h) for evidencing the effects of the heat on the plastic containers.

Experimental Design
The investigated compounds are listed in Table 1, along with the main analytical parameters. hottest summer seasons, with consequent loss of the crop. The possible presence of PAEs and BPA in honey is also related to the production processes that may involve direct contact with unsuitable plastic (e.g., single-dose plastic packs). The determination of PAEs and BPA presents considerable difficulties related mainly to their low concentration, which is currently difficult to determine even with very sensitive instruments [25]. To overcome this problem, it is necessary to develop an analytical method that provides for the extraction and pre-concentration procedure. In this paper, ultrasound vortex assisted DLLME (UVA-DLLME) was used as the extraction technique [26], whereas Gas Chromatography coupled with Ion Trap Mass Spectrometry (GC-IT/MS) was used as the analysis technique. The method developed, validated, and experimentally consolidated in the laboratory takes into account all the parameters that influence the extraction of the analytes. For this purpose, the extraction solvent, the ionic strength of the solution, the pH, and the extraction volume were analyzed and studied.
The developed method was applied to six real samples (i.e., four single-use commercial ones available in the Italian market, one home-made honey nectar sample, and one sampled from honey produced in plastic honeycomb). This last sample is interesting because the plastic honeycomb represents a new frontier in honey production, according to the author's knowledge, this measurement is the first analysis for understanding the migration of such compounds from the honeycomb to the honey. Finally, the four commercial samples were subjected to thermal stress (i.e., exposed to 40 °C for 24 and 48 h) for evidencing the effects of the heat on the plastic containers.

Experimental Design
The investigated compounds are listed in Table 1, along with the main analytical parameters. hottest summer seasons, with consequent loss of the crop. The possible presence of PAEs and BPA in honey is also related to the production processes that may involve direct contact with unsuitable plastic (e.g., single-dose plastic packs). The determination of PAEs and BPA presents considerable difficulties related mainly to their low concentration, which is currently difficult to determine even with very sensitive instruments [25]. To overcome this problem, it is necessary to develop an analytical method that provides for the extraction and pre-concentration procedure. In this paper, ultrasound vortex assisted DLLME (UVA-DLLME) was used as the extraction technique [26], whereas Gas Chromatography coupled with Ion Trap Mass Spectrometry (GC-IT/MS) was used as the analysis technique. The method developed, validated, and experimentally consolidated in the laboratory takes into account all the parameters that influence the extraction of the analytes. For this purpose, the extraction solvent, the ionic strength of the solution, the pH, and the extraction volume were analyzed and studied.
The developed method was applied to six real samples (i.e., four single-use commercial ones available in the Italian market, one home-made honey nectar sample, and one sampled from honey produced in plastic honeycomb). This last sample is interesting because the plastic honeycomb represents a new frontier in honey production, according to the author's knowledge, this measurement is the first analysis for understanding the migration of such compounds from the honeycomb to the honey. Finally, the four commercial samples were subjected to thermal stress (i.e., exposed to 40 °C for 24 and 48 h) for evidencing the effects of the heat on the plastic containers.

Experimental Design
The investigated compounds are listed in Table 1, along with the main analytical parameters. hottest summer seasons, with consequent loss of the crop. The possible presence of PAEs and BPA in honey is also related to the production processes that may involve direct contact with unsuitable plastic (e.g., single-dose plastic packs). The determination of PAEs and BPA presents considerable difficulties related mainly to their low concentration, which is currently difficult to determine even with very sensitive instruments [25]. To overcome this problem, it is necessary to develop an analytical method that provides for the extraction and pre-concentration procedure. In this paper, ultrasound vortex assisted DLLME (UVA-DLLME) was used as the extraction technique [26], whereas Gas Chromatography coupled with Ion Trap Mass Spectrometry (GC-IT/MS) was used as the analysis technique. The method developed, validated, and experimentally consolidated in the laboratory takes into account all the parameters that influence the extraction of the analytes. For this purpose, the extraction solvent, the ionic strength of the solution, the pH, and the extraction volume were analyzed and studied.
The developed method was applied to six real samples (i.e., four single-use commercial ones available in the Italian market, one home-made honey nectar sample, and one sampled from honey produced in plastic honeycomb). This last sample is interesting because the plastic honeycomb represents a new frontier in honey production, according to the author's knowledge, this measurement is the first analysis for understanding the migration of such compounds from the honeycomb to the honey. Finally, the four commercial samples were subjected to thermal stress (i.e., exposed to 40 °C for 24 and 48 h) for evidencing the effects of the heat on the plastic containers.

Experimental Design
The investigated compounds are listed in Table 1, along with the main analytical parameters. hottest summer seasons, with consequent loss of the crop. The possible presence of PAEs and BPA in honey is also related to the production processes that may involve direct contact with unsuitable plastic (e.g., single-dose plastic packs). The determination of PAEs and BPA presents considerable difficulties related mainly to their low concentration, which is currently difficult to determine even with very sensitive instruments [25]. To overcome this problem, it is necessary to develop an analytical method that provides for the extraction and pre-concentration procedure. In this paper, ultrasound vortex assisted DLLME (UVA-DLLME) was used as the extraction technique [26], whereas Gas Chromatography coupled with Ion Trap Mass Spectrometry (GC-IT/MS) was used as the analysis technique. The method developed, validated, and experimentally consolidated in the laboratory takes into account all the parameters that influence the extraction of the analytes. For this purpose, the extraction solvent, the ionic strength of the solution, the pH, and the extraction volume were analyzed and studied.
The developed method was applied to six real samples (i.e., four single-use commercial ones available in the Italian market, one home-made honey nectar sample, and one sampled from honey produced in plastic honeycomb). This last sample is interesting because the plastic honeycomb represents a new frontier in honey production, according to the author's knowledge, this measurement is the first analysis for understanding the migration of such compounds from the honeycomb to the honey. Finally, the four commercial samples were subjected to thermal stress (i.e., exposed to 40 °C for 24 and 48 h) for evidencing the effects of the heat on the plastic containers.

Experimental Design
The investigated compounds are listed in Table 1, along with the main analytical parameters. hottest summer seasons, with consequent loss of the crop. The possible presence of PAEs and BPA in honey is also related to the production processes that may involve direct contact with unsuitable plastic (e.g., single-dose plastic packs). The determination of PAEs and BPA presents considerable difficulties related mainly to their low concentration, which is currently difficult to determine even with very sensitive instruments [25]. To overcome this problem, it is necessary to develop an analytical method that provides for the extraction and pre-concentration procedure. In this paper, ultrasound vortex assisted DLLME (UVA-DLLME) was used as the extraction technique [26], whereas Gas Chromatography coupled with Ion Trap Mass Spectrometry (GC-IT/MS) was used as the analysis technique. The method developed, validated, and experimentally consolidated in the laboratory takes into account all the parameters that influence the extraction of the analytes. For this purpose, the extraction solvent, the ionic strength of the solution, the pH, and the extraction volume were analyzed and studied.
The developed method was applied to six real samples (i.e., four single-use commercial ones available in the Italian market, one home-made honey nectar sample, and one sampled from honey produced in plastic honeycomb). This last sample is interesting because the plastic honeycomb represents a new frontier in honey production, according to the author's knowledge, this measurement is the first analysis for understanding the migration of such compounds from the honeycomb to the honey. Finally, the four commercial samples were subjected to thermal stress (i.e., exposed to 40 °C for 24 and 48 h) for evidencing the effects of the heat on the plastic containers.

Experimental Design
The investigated compounds are listed in Table 1, along with the main analytical parameters.

Standard Solutions
• Preparation of stock solution for each analyte, 1000 µg g −1 : weigh 10 mg of each PAE/BPA; make up to volume with 10 mL of acetone.
• Preparation of diluted PAE/BPA mix solution, 10 µg g −1 : appropriate dilution of the mother solutions with acetone to set up a PAE/BPA mix solution. weigh 1 mg of phenanthrene; make up to volume, 10 mL, with acetone (100 µg g −1 ); appropriate dilution for obtaining 0.5 and 0.05 µg g −1 I.S. solutions.
• All solutions were stored in darkness vials at 4 • C. Add distilled water up to 10 g.

3.
Solubilize the honey in the solution.

4.
Check pH and adjust at pH 4.
Methods Protoc. 2020, 3, x FOR PEER REVIEW 5 of 13 3. Solubilize the honey in the solution. 4. Check pH and adjust at pH 4. CRITICAL STEP: the pH choice is decisive for the successful procedure. PAEs are better extracted at pH alkaline but they are only recovered at pH 4, BPA is extracted at acid pH whereas at alkaline pH its recovery is very low (between 20-30%): pH 4 allows for the quantitative recovery of all compounds. 5. Add 7.5 µ L of phenanthrene as I.S. 6. Vortex 15 s. 7. Add the extraction solvent, 75 µ L of benzene. 8. Vortex 5 min: formation of the macroemulsion. 9. Ultrasound 6 min.
CRITICAL STEP: This step is fundamental for the formation of the microemulsion. The ultrasound give the power for the microemulsion. 10.
PAUSE STEP: the microemulsion formation is essential for the procedure. If it does not form, the analytical procedure can be stopped because the extraction has not occurred. 11. Add NaCl 10 g L −1 to break the microemulsion. OPTIONAL STEP: the addition of NaCl helps the microemulsion break. 12. Centrifugation at 4000 rpm per 30 min. 13. Withdraw 1 µ L of the supernatant. 14. Inject into gas chromatography equipped with ion trap mass spectrometry (GC-IT/MS).

Thermal Stress Procedure
 24 h at 40 °C, withdrawing 2.5 g of honey and processing like in Section 3.1.  48 h at 40 °C, withdrawing 2.5 g of honey and processing like in Section 3.1.

Results and Discussion
The first applications found excellent results in the pre-concentration of organic analytes from aqueous samples, demonstrating high extraction efficiency [27][28][29]. However, before proceeding with the extraction of the investigated analytes, some fundamental parameters must be optimized for the correct extraction. Among these, the most important are the volume of extraction and dispersive solvent, the extraction time, the pH, the ionic strength, and the amount of sample examined. All optimization procedures were carried out on a blank honey sample spiked with known amounts of PAEs/BPA. A blank sample of honey was collected from a local apiary and stored in darkness 4 °C in PAE-free glass bottles: it was processed in advance for the optimization of extraction conditions and validation of the developed method.
The pH of the aqueous solution is decisive for the quantification of the molecules because at basic pH, following the centrifugation phase, there is the formation of a gel, which by absorbing the analytes adversely affects the recoveries. Furthermore, the natural composition of most honeys allows for an easier pH adjustment at acid values without drastically changing the nature of the matrix [30].
The ultrasound treatment of the solution is a critical step because this allows for the formation of the oil-in-water microemulsion. This occurrence makes the solvent droplets very small, which disperse in the solution, thus increasing the contact surface and recovery.
The addition of NaCl favors the reverse of the phases and therefore the breaking of the emulsion; it also increases the ionic strength of the solution, further promoting the extraction of the molecules by reducing the solubility of the analytes (salting-out effect) [31]. With regard to the extraction of the molecules, studies in this laboratory showed that even without the addition of NaCl, the recoveries were quantitative [32][33][34]. Considering the matrix in question, the solution also had good ionic strength due to the presence of mineral salts, even in the absence of NaCl.
To identify the best extractant solvent, five different solvents were tested (i.e., n-heptane, isooctane, benzene, toluene, and toluene + acetone (1 + 1 v/v). The recoveries are reported in Table 2, benzene showed the best performance in the compound extraction.
CRITICAL STEP: the pH choice is decisive for the successful procedure. PAEs are better extracted at pH alkaline but they are only recovered at pH 4, BPA is extracted at acid pH whereas at alkaline pH its recovery is very low (between 20-30%): pH 4 allows for the quantitative recovery of all compounds. 5.
Vortex 5 min: formation of the macroemulsion.

Ultrasound 6 min.
Methods Protoc. 2020, 3, x FOR PEER REVIEW 5 of 13 3. Solubilize the honey in the solution. 4. Check pH and adjust at pH 4. CRITICAL STEP: the pH choice is decisive for the successful procedure. PAEs are better extracted at pH alkaline but they are only recovered at pH 4, BPA is extracted at acid pH whereas at alkaline pH its recovery is very low (between 20-30%): pH 4 allows for the quantitative recovery of all compounds. 5. Add 7.5 µ L of phenanthrene as I.S. 6. Vortex 15 s. 7. Add the extraction solvent, 75 µ L of benzene. 8. Vortex 5 min: formation of the macroemulsion. 9. Ultrasound 6 min.
CRITICAL STEP: This step is fundamental for the formation of the microemulsion. The ultrasound give the power for the microemulsion. 10.
PAUSE STEP: the microemulsion formation is essential for the procedure. If it does not form, the analytical procedure can be stopped because the extraction has not occurred. 11. Add NaCl 10 g L −1 to break the microemulsion. OPTIONAL STEP: the addition of NaCl helps the microemulsion break. 12. Centrifugation at 4000 rpm per 30 min. 13. Withdraw 1 µ L of the supernatant. 14. Inject into gas chromatography equipped with ion trap mass spectrometry (GC-IT/MS).

Thermal Stress Procedure
 24 h at 40 °C, withdrawing 2.5 g of honey and processing like in Section 3.1.  48 h at 40 °C, withdrawing 2.5 g of honey and processing like in Section 3.1.

Results and Discussion
The first applications found excellent results in the pre-concentration of organic analytes from aqueous samples, demonstrating high extraction efficiency [27][28][29]. However, before proceeding with the extraction of the investigated analytes, some fundamental parameters must be optimized for the correct extraction. Among these, the most important are the volume of extraction and dispersive solvent, the extraction time, the pH, the ionic strength, and the amount of sample examined. All optimization procedures were carried out on a blank honey sample spiked with known amounts of PAEs/BPA. A blank sample of honey was collected from a local apiary and stored in darkness 4 °C in PAE-free glass bottles: it was processed in advance for the optimization of extraction conditions and validation of the developed method.
The pH of the aqueous solution is decisive for the quantification of the molecules because at basic pH, following the centrifugation phase, there is the formation of a gel, which by absorbing the analytes adversely affects the recoveries. Furthermore, the natural composition of most honeys allows for an easier pH adjustment at acid values without drastically changing the nature of the matrix [30].
The ultrasound treatment of the solution is a critical step because this allows for the formation of the oil-in-water microemulsion. This occurrence makes the solvent droplets very small, which disperse in the solution, thus increasing the contact surface and recovery.
The addition of NaCl favors the reverse of the phases and therefore the breaking of the emulsion; it also increases the ionic strength of the solution, further promoting the extraction of the molecules by reducing the solubility of the analytes (salting-out effect) [31]. With regard to the extraction of the molecules, studies in this laboratory showed that even without the addition of NaCl, the recoveries CRITICAL STEP: This step is fundamental for the formation of the microemulsion. The ultrasound give the power for the microemulsion.

10.
Methods Protoc. 2020, 3, x FOR PEER REVIEW 5 of 13 3. Solubilize the honey in the solution. 4. Check pH and adjust at pH 4. CRITICAL STEP: the pH choice is decisive for the successful procedure. PAEs are better extracted at pH alkaline but they are only recovered at pH 4, BPA is extracted at acid pH whereas at alkaline pH its recovery is very low (between 20-30%): pH 4 allows for the quantitative recovery of all compounds. 5. Add 7.5 µ L of phenanthrene as I.S. 6. Vortex 15 s. 7. Add the extraction solvent, 75 µ L of benzene. 8. Vortex 5 min: formation of the macroemulsion. 9. Ultrasound 6 min.
CRITICAL STEP: This step is fundamental for the formation of the microemulsion. The ultrasound give the power for the microemulsion. 10.
PAUSE STEP: the microemulsion formation is essential for the procedure. If it does not form, the analytical procedure can be stopped because the extraction has not occurred. 11. Add NaCl 10 g L −1 to break the microemulsion. OPTIONAL STEP: the addition of NaCl helps the microemulsion break. 12. Centrifugation at 4000 rpm per 30 min. 13. Withdraw 1 µ L of the supernatant. 14. Inject into gas chromatography equipped with ion trap mass spectrometry (GC-IT/MS).

Thermal Stress Procedure
 24 h at 40 °C, withdrawing 2.5 g of honey and processing like in Section 3.1.  48 h at 40 °C, withdrawing 2.5 g of honey and processing like in Section 3.1.

Results and Discussion
The first applications found excellent results in the pre-concentration of organic analytes from aqueous samples, demonstrating high extraction efficiency [27][28][29]. However, before proceeding with the extraction of the investigated analytes, some fundamental parameters must be optimized for the correct extraction. Among these, the most important are the volume of extraction and dispersive solvent, the extraction time, the pH, the ionic strength, and the amount of sample examined. All optimization procedures were carried out on a blank honey sample spiked with known amounts of PAEs/BPA. A blank sample of honey was collected from a local apiary and stored in darkness 4 °C in PAE-free glass bottles: it was processed in advance for the optimization of extraction conditions and validation of the developed method.
The pH of the aqueous solution is decisive for the quantification of the molecules because at basic pH, following the centrifugation phase, there is the formation of a gel, which by absorbing the analytes adversely affects the recoveries. Furthermore, the natural composition of most honeys allows for an easier pH adjustment at acid values without drastically changing the nature of the matrix [30].
The ultrasound treatment of the solution is a critical step because this allows for the formation of the oil-in-water microemulsion. This occurrence makes the solvent droplets very small, which disperse in the solution, thus increasing the contact surface and recovery.
The addition of NaCl favors the reverse of the phases and therefore the breaking of the emulsion; it also increases the ionic strength of the solution, further promoting the extraction of the molecules by reducing the solubility of the analytes (salting-out effect) [31]. With regard to the extraction of the molecules, studies in this laboratory showed that even without the addition of NaCl, the recoveries were quantitative [32][33][34]. Considering the matrix in question, the solution also had good ionic strength due to the presence of mineral salts, even in the absence of NaCl.
To identify the best extractant solvent, five different solvents were tested (i.e., n-heptane, isooctane, benzene, toluene, and toluene + acetone (1 + 1 v/v). The recoveries are reported in Table 2, benzene showed the best performance in the compound extraction.
PAUSE STEP: the microemulsion formation is essential for the procedure. If it does not form, the analytical procedure can be stopped because the extraction has not occurred. 11. Add NaCl 10 g L −1 to break the microemulsion. OPTIONAL STEP: the addition of NaCl helps the microemulsion break. 12. Centrifugation at 4000 rpm per 30 min. 13. Withdraw 1 µL of the supernatant. 14. Inject into gas chromatography equipped with ion trap mass spectrometry (GC-IT/MS).

Thermal Stress Procedure
• 24 h at 40 • C, withdrawing 2.5 g of honey and processing like in Section 3.1. • 48 h at 40 • C, withdrawing 2.5 g of honey and processing like in Section 3.1.

Results and Discussion
The first applications found excellent results in the pre-concentration of organic analytes from aqueous samples, demonstrating high extraction efficiency [27][28][29]. However, before proceeding with the extraction of the investigated analytes, some fundamental parameters must be optimized for the correct extraction. Among these, the most important are the volume of extraction and dispersive solvent, the extraction time, the pH, the ionic strength, and the amount of sample examined. All optimization procedures were carried out on a blank honey sample spiked with known amounts of PAEs/BPA. A blank sample of honey was collected from a local apiary and stored in darkness 4 • C in PAE-free glass bottles: it was processed in advance for the optimization of extraction conditions and validation of the developed method.
The pH of the aqueous solution is decisive for the quantification of the molecules because at basic pH, following the centrifugation phase, there is the formation of a gel, which by absorbing the analytes adversely affects the recoveries. Furthermore, the natural composition of most honeys allows for an easier pH adjustment at acid values without drastically changing the nature of the matrix [30].
The ultrasound treatment of the solution is a critical step because this allows for the formation of the oil-in-water microemulsion. This occurrence makes the solvent droplets very small, which disperse in the solution, thus increasing the contact surface and recovery.
The addition of NaCl favors the reverse of the phases and therefore the breaking of the emulsion; it also increases the ionic strength of the solution, further promoting the extraction of the molecules by reducing the solubility of the analytes (salting-out effect) [31]. With regard to the extraction of the molecules, studies in this laboratory showed that even without the addition of NaCl, the recoveries were quantitative [32][33][34]. Considering the matrix in question, the solution also had good ionic strength due to the presence of mineral salts, even in the absence of NaCl.
To identify the best extractant solvent, five different solvents were tested (i.e., n-heptane, iso-octane, benzene, toluene, and toluene + acetone (1 + 1 v/v). The recoveries are reported in Table 2, benzene showed the best performance in the compound extraction.
The choice of benzene as the extraction solvent deserves a reason because it is well-known for its dangerousness (i.e., potential occupational carcinogen, flammability. The International Agency for Research on Cancer, IARC, rated benzene as "known to be carcinogenic to humans", Group 1). Although it is more hazardous than the other solvents tested, it demonstrated almost quantitative recoveries for all the molecules under study, especially BPA, and not only for some of them such as has occurred for toluene, iso-octane, and heptane. In addition, the use of benzene has allowed competitive or even better Relative Standard Deviations (RSDs) to be obtained than the other papers present in the literature. The risks for the operator can be reduced by adopting both suitable personal protective equipment (PPE) and collective protection equipment (CPE), and by carrying out the critical operations under a fume hood. Furthermore, in the proposed method, only a very small amount of benzene (75 µL) is needed. This last issue (i.e., working with small volumes) also involves easy disposal management for environmental protection.
Afterward, the volume of the extractant solvent and the best pH of the solution were investigated. Six different volumes (25,50,75,100,150, and 200 µL) were considered and the relative recoveries were compared: Table 3 shows that the best recoveries were obtained using 75 µL of benzene whereas Table 4 shows that pH 4 allowed for excellent recoveries to be obtained for all of the analyzed compounds.    Optimal times and rotation times for vortexing (5 min), ultrasound (6 min), and centrifugation (30 min at 4000 rpm) were established on the basis of previous studies carried out in this laboratory [21,32].
The analytical protocol was validated in terms of linearity range, correlation coefficients, reproducibility, intra-and inter-day errors and recoveries, and by performing the entire procedure on honey samples (i.e., 2.5 g of the honey sample, addition of 7.5 µL of I.S. 50 pg µL −1 , 75 µL volume of benzene, pH 4 of the solution, NaCl 10 g L −1 , vortex 5 min, ultrasounds 6 min, and centrifugation 30 min at 4000 rpm). The proposed method does not have a clean-up, as recently shown in a previous paper where acaricides were determined in honey samples: recoveries can be considered quantitative [35]. Figure 1 shows typical chromatograms of (a) the standard mixture solution (each PAE and BPA at 50 ng g −1 ), (b) the honey sample, and (c) the sample honey spiked with the standard mixture solution. All samples were subjected to the whole procedure: as it can be seen, no peak overlapping was present, the interferences did not affect the qualitative and quantitative analysis, and the peaks were well shaped and clear. Table 5 reports all the analytical data. Table 5. Regression equation, correlation coefficient (R 2 ) in the range from 50 to 5000 ng g −1 , limit of detection (LOD) and limit of quantification (LOQ) (ng g −1 ), intra-and inter-day precision (as relative standard deviation, RSD), and recoveries at two different concentrations (low, 50 ng g −1 , and high, 500 ng g −1 , fortification) of each PAE and BPA investigated in this study. The selected ion monitoring (SIM) m/z of the typical fragment ion for each compound is reported in Table 1 Looking overall at the analytical data, the method is robust. In fact, the regression equation and the correlation coefficients (seven-point calibration curve) were good for all compounds in the concentration range analyzed. LODs and LOQs were directly determined in the matrices investigated according to the International Conference on Harmonization: Validation of Analytical Procedure [36] (i.e., an analyte that produces a chromatographic peak equal to three times (LOD) or seven times (LOQ) the standard deviation of the baseline noise). These are sufficient for determining such compounds in honey, according to the regulations 10/2010 and 213/2018 [37,38], which define the specific migration limit (SML) of each PAE/BPA from the plastic into the food (i.e., 60 mg kg −1 for DMP, DEP, DiBP, and DNOP, 1.5 mg kg −1 for DEHP, 0.3 mg kg −1 for DBP and 0.05 mg kg −1 for BPA). The intra-and inter-day precision (as relative standard deviation, RSD) were below 7.2 and 9.3, respectively, meaning that the analytical procedure is accurate. Finally, to complete the protocol validation, the recoveries were studied at two different PAE/BPA concentrations (i.e., low (50 ng g −1 ) and high (500 ng g −1 ) fortification). The recoveries ranged between 71-97 % with a RSD below 9 and 76-100 % with a RSD below 6 for low and high fortification, respectively. This was further confirmation about the robustness of the developed analytical protocol for determining PAEs and BPA in the honey matrix.
Methods Protoc. 2020, 3, x FOR PEER REVIEW 8 of 13 The intra-and inter-day precision (as relative standard deviation, RSD) were below 7.2 and 9.3, respectively, meaning that the analytical procedure is accurate. Finally, to complete the protocol validation, the recoveries were studied at two different PAE/BPA concentrations (i.e., low (50 ng g -1 ) and high (500 ng g −1 ) fortification). The recoveries ranged between 71-97 % with a RSD below 9 and 76-100 % with a RSD below 6 for low and high fortification, respectively. This was further confirmation about the robustness of the developed analytical protocol for determining PAEs and BPA in the honey matrix.  To highlight the strengths of this paper, a comparison with previous studies [39][40][41][42][43][44][45][46][47][48] performed on such compounds in similar honey matrices is reported in Table 6 in terms of recovery, LOD, LOQ, and RSD whereas three other papers [47][48][49] determined the PAE/BPA concentrations in the honey matrix without describing the analytical parameters. First, the papers showing only analysis are very few in number and only one [six deal with the simultaneous determination of PAEs and BPA. This confirms the importance and difficulties of this issue. As can be seen, the protocol developed in this study allows for the simultaneous investigation of these dangerous compounds at ng g −1 levels. The only other paper allowing this simultaneous determination showed higher recoveries (up to 120%) and LODs/LOQs (up to 303 ng g −1 and 1013 ng g −1 ).  The procedure was applied to six different honey samples (i.e., four single-use commercial ones and two home-made ones). All samples were stored at 4 • C in darkness. As can be seen in Table 7, the values of the analytes found in the real samples had very low concentrations, in the order of ng g −1 and were within the legal limits (as SMLs) established by the European Union. It is quite interesting because in Sample F, the honey was withdrawn from a plastic honeycomb. The use of these kinds of beehives are widely expanding for different reasons, mainly to save the crop or to avoid the wax melting during the hottest periods of the year. Between the two home-made honeys, which were collected in two different areas but not so far from them, the PAE profile seems to be the same, except with a slight enhancement in the BPA level. Table 7. PAE and BPA levels (µg g −1 ) found in the six home-made and commercial honey samples analyzed by the proposed protocol. Finally, the four commercial samples were subjected to thermal stress: the hypothesis is that high temperatures and irradiation as well as exposure to the Sun's ultraviolet rays, can break down plastics and cause problems (see Table 8). The main consideration with regard to the DMP is that it seems to have disappeared from all samples (i.e., levels < LOD), whereas the other PAE concentrations remained almost constant (only DEHP increased in Samples B, D, and E). Particular attention should be pointed out for BPA: its level was worryingly increased in Sample D (more than six times) and in Sample B (more than 3-times). Furthermore, after the thermal stress, Sample B showed a BPA content of 0.54 µg g −1 , which was higher than the current legal limit [38]. This can be explained because the package was produced and distributed before the entry into force of the aforementioned EC Regulation. Analyses performed after 48 h did not show any significant analytical increases when compared to concentrations obtained after 24 h. This suggests that the significant release of these molecules occurs in the first moments of thermal shock.

Conclusions
In this paper, an experimental analytical method was developed for the determination of PAEs and BPA from plastic containers inside which honey is contained. The study showed a good robustness of the method, simplicity of execution by the operator, and relatively short analysis times. In addition, the good reproducibility, sensitivity, the use of small quantities of solvents, the short extraction time, and the use of instrument now present in most chemical analysis laboratories, make it easily applicable everywhere. In addition, the analysis of real samples has shown the presence of small quantities of phthalates all below the legal limits currently known.
Finally, the samples subjected to thermal stress for 24 and 48 h showed a slight release only in the first hours. Subsequently, the concentration of the analyzed analytes remained constant. The plastics must be kept out of sunlight and should also avoid high temperatures: our recommendation is to avoid plastic materials not labeled as "Food Safe" or "BPA" free.