Evaluation of three extraction methods for the isolation of PAHs from recycled paperboard materials intended for food contact applications

The present work is part of a comprehensive study aiming to identify and quantify an array of harmful contaminants at very low concentrations in recycled paperboard materials intended for food contact applications. In this initial study, three commonly used extraction techniques were optimized and tested for their efficacy and accuracy in isolating 13 polycyclic aromatic hydrocarbons (PAHs) from three commercially available food packaging materials made from recycled paperboard. The techniques were Soxtec extraction (SE), ultrasound-assisted extraction (UAE), and headspace solid-phase microextraction (HS-SPME), followed by gas chromatography-mass spectrometry (GC/MS) analysis. The recoveries of all three methods were rather satisfactory; 10 out of 13 substances were recovered at rates between 70 and 124% using the SE method, 8 out of 13 substances were recovered at rates between 50 and 134% using the UAE method, and finally, 8 out of 13 substances were recovered at rates between 70 and 105% using the HS-SPME method. Regarding accuracy, the HS-SPME method gave the best results at 125 μg kg−1, as 7 out of 13 concentration values were within acceptable limits, while only five and four concentration values were within limits for the UAE and SE methods, respectively. Furthermore, the results of both the calculated RSD and HorRat values indicated that HS-SPME is more accurate for extracting PAHs than the other two methods.


Introduction
The food industry is undergoing a time of radical transformation in terms of food packaging. The pursuit of a circular economy is gaining ground in public opinion and legislators worldwide, in contrast to the existing linear economic model, which is gradually becoming obsolete. This new economic model favors technologies that endorse the concept of keeping materials and products in circulation at their highest value. To this end, the reuse and recycling of waste, especially packaging waste, are becoming essential tools on the road to a sustainable future. To meet consumer demand for environmentally friendly alternatives and to combat soaring packaging waste, food and packaging companies are shifting toward recycled food packaging [1]. In addition, paper-based materials have been taking off in an attempt by producers to reduce the amount of single-use plastics in their packaging process [2,3]. However, as some chronic diseases are associated with certain chemicals migrating from packaging, it is of great importance to evaluate the safety of these recycled packaging materials upfront.
The primary role of food packaging is to safeguard the enclosed product against all types of contamination, and this task can only be fulfilled if the potential threats are identified and eliminated in time. Regarding chemical contamination, migration seems to be one of the most common culprits [4]. Therefore, to ensure a safe product, more emphasis must be placed on the quality of raw materials and the production process of the surrounding food packaging. Insightful risk management and compliance with Good Manufacturing Practice (GMP) are reasonable steps toward this end [5,6].
Paperboard, an easily recyclable, environmentally friendly material, has many technical advantages and properties that qualify it as an ideal packaging agent. Depending on the pulp's processing, the final chemical composition of the material can vary significantly [7]. According 1 3 to the CEPI guidelines [5], only substances that have been priorly assessed as safe or are included in official risk-free lists are to be used as intentionally added raw materials and substances (IAS). In addition, not intentionally added materials and substances (NIAS), such as contaminants and by-products, must be continually monitored and evaluated during the manufacturing process of paper and board materials intended for food packaging. Studies have found that both IAS and NIAS can be detected as residues in food contact materials (FCMs) and are reportedly migration hazards [8][9][10][11]. When the raw material consists of recycled fibers, the risk of contamination can become higher during the production process of paperboard [12][13][14]. Many previous studies have delved into the potential presence of hazardous contaminants in recycled cellulose and the impact such chemicals can have on human health [12,[15][16][17][18][19][20][21][22]. The substances in question include chemicals mostly present in inks, adhesives, plasticizers, and colorants [14,23,24], such as phthalates, diisopropyl naphthalenes (DIPNs), Bisphenol A, mineral oil hydrocarbons (MOHs), and polycyclic aromatic hydrocarbons (PAHs).
PAHs, a diverse group of organic compounds, are one of the most widespread pollutants with proven adverse effects on various living organisms, including humans [25][26][27][28][29]. However, despite their toxicity, PAHs are inextricably integrated into everyday life due to their presence in staple products such as toys, shoes, as well as foodstuffs. Therefore, it is necessary to identify and quantify these types of chemicals in products that come into contact with foods and can potentially contaminate them. This study aims to optimize and evaluate three standard extraction methods for their efficacy and accuracy in isolating 13 polycyclic aromatic hydrocarbons (PAHs) from 3 commercial food packaging materials made from recycled paperboard. The extraction methods tested were chosen based on their low cost, low environmental impact, easy-to-use procedures, and quick results.

Reagents
The PAHs mix used in the sample spiking procedure was of high analytical grade. It consisted of the following 13 compounds: Naphthalene (NPH), Acenaphthylene (ANY), Acenaphthene

Sample preparation
Four types of commercial paperboard intended for packaging dry food products were used, all provided by a manufacturing company located in Greece. All samples were without labeling (ink), so removing any external surfaces was unnecessary before the assays [15,30]. Three consisted of 100% recycled material (BR, R, and R KRAFT), and one was of virgin material (V).
More specifically, the materials used in the present study were: R1: "TRIPLEX R KRAFT" commercially labeled paperboard. 100% made of recycled fibers, fully coated triplex white lined chipboard (WLC) with kraft (brown) reverse side. It is popular in western European markets and is used almost exclusively for packaging dry foodstuff. Basis Weight: 440 g/m 2 , thickness: 541 μm.
R3: "TRIPLEX BR" commercially labeled paperboard. 100% made of recycled fibers, fully coated triplex white lined chipboard (WLC) with coated (white) reverse side. It comprises up to 40% of paperboard production, is suitable for food packaging, and is the paperboard of choice for high-specification packaging for cosmetics and pharmaceutical products. Basis weight: 440 g/m 2 , thickness: 541 μm.
Every recycled paperboard sample was initially freed from potentially present substances by undergoing cleansing ultrasound baths in dichloromethane. From the original 1 g L −1 stock solutions in methanol, 10,000 μg L −1 and 1000 μg L −1 dilutions were prepared for each contaminant, and their final concentrations ranged from 20 to 20,000 μg L −1 . The spiking process was optimized through contamination testing to ensure the uniform distribution of the contaminants in all paperboard samples. Finally, standard reference curves were generated for each substance.

Analytical methods
The methods tested in this project were the Soxtec Extraction (SE), Ultrasound-Assisted Extraction (UAE), and Head Space Solid Phase Micro Extraction (HS-SPME), 1 3 followed by a gas chromatography-mass spectrometry (GC/MS) analysis applied so as to identify the method yielding the best analytical results. The extraction conditions were optimized for all processes by comparing the relative recovery rates of integrating the contaminant peaks in the respective chromatograms (yields) to the compound exhibiting the highest recovery.
Finally, the analytical results were compared to the limits proclaimed in the Food Contact Guidelines priorly mentioned [5,31], and it was ascertained that the determined values were within acceptable limits for all commercial paperboards tested.

Soxtec extraction (SE)
The Soxtec extractions of the paperboard samples were performed using a VELP Scientifica Srl (Italy) semi-automatic SER 148 solvent extractor. Prior to the analysis, several factors were examined to ensure optimum conditions. Three types of solvents were tested: ethanol, hexane, and dichloromethane (DCM), and in accordance with previous studies [19,32], DCM showed the best extraction results. After ensuring the appropriate solvent, various volumes of DCM were subsequently examined, and it was found that 50 mL of DCM was the ideal amount for this extraction. Several parameters of the extraction were then studied: extraction temperature (70 °C, 90 °C, 110 °C, and 130 °C), extraction time (20,30,40, 60, 90,120, 150, and 180 min), as well as rinsing time (10, 30, 60, and 90 min). Finally, the optimum method was followed in terms of type of solvent, solvent volume, temperature, extraction time, and rinsing time. Precisely 1.0 g of each contaminated paperboard sample, as well as of the virgin sample, was placed in a 33 mm × 80 mm extraction thimble and was extracted with 50 ml dichloromethane (DCM) at 110 °C for 120 min. After completion, the thimble was raised to rinse position for another 90 min. The sample was then dried completely using synthetic air, redissolved in 1.0 ml of acetonitrile, and finally, 1 μL of the mixture was injected into the GC/MS instrument. All assays were conducted thrice.

Ultrasound-assisted extraction (UAE)
An Elmasonic S10 H ultrasonic bath (Elma GmbH, Singen, Germany) was used for the UAE analysis. The parameters of the liquid-phase extraction were the same as in previous work [19]. Approximately 175 mg of each paperboard sample, cut in square shapes (5 × 5 mm), were extracted in the ultrasonic bath for 25 min at 20 °C using DCM. After the mixture was filtered, it was rinsed twice with 1 ml dichloromethane. The organic phase was then evaporated to dryness with nitrogen, redissolved in 0.175 ml dichloromethane, and finally stored in screw cap vials until further analysis. Triplicate determinations were performed for all samples.

Head space-solid phase micro extraction (HS-SPME)
Paperboard samples were steeped in 1 mL of the 1000 μg L −1 standard solution in methanol for identifying and quantifying the selected substances. Approximately 100 mg of the saturated paperboards were weighed, cut into square pieces (5 × 5 mm), and placed in glass vials of 10 ml volume. Finally, all vials were sealed with Teflon septa.
Before the extraction, several factors were evaluated to ensure optimum conditions. Initially, four different fiber coatings were tested on their ability to adsorb the contaminants in one trial run and their behavior toward the substances in question. Out of the four fibers tested (100 μm Polydimethylsiloxane (PDMS), 85 μm Polyacrylate (PA), 75 μm carboxen/polydimethylsiloxane (CAR/PDMS), 65 μm polydimethylsiloxane/divinylbenzene (PDMS/DVB)), the PDMS/DVB fiber yielded the best analytical results, as is similarly supported in other studies [33], and was consequently chosen for the HS-SPME analysis. After selecting the fiber coating and the fiber's exposure time, the oil bath temperature was examined to achieve maximum recovery rates. Tests were performed in the temperature range of 100-190 °C to determine which one succeeded in extracting the highest rates of contaminants. Comparing the integral values of different temperatures indicated that the highest rates of pollutants were recovered when the oil bath temperature was set at 150 °C. Moving up to higher temperature conditions (170 °C) did not affect the values, indicating that the extraction process had already been accomplished. Finally, another set of experiments was conducted to determine the necessary exposure time of the fiber at 150 °C to achieve maximum recovery of the contaminants from the recycled paperboard. Exposure times 20, 30, 40, 60, and 90 min were tested, and it was found that the best results were obtained at 30 min exposure time, while longer than 60 min exposure resulted in acquiring decreased integral values. Extended exposure of the fiber to heat seems to affect the distribution equilibrium between the analyte in the fiber coating and the headspace.
In short, the extraction of volatile compounds from the headspace was performed using a 65 μm PDMS/DVB fiber (Supelco, USA). The fiber was firstly conditioned according to the manufacturer's recommendations. The vials were then placed in an oil bath at 150 °C for 15 min until equilibrium, and after that, the PDMS/DVB fiber was inserted and exposed to the headspace for precisely 30 min. At the end of the exposure time, the fiber was transferred directly to the GC injection port for the desorption of the substances. All experimental tests were performed in triplicate.

3
A standard solution of 1000 μg L −1 in methanol was prepared in a 10 mL volumetric flask. With appropriate dilutions, 6 aliquots of the contaminants were prepared at concentrations of 30, 60, 125, 250, 500, and 750 μg L −1 . The paperboard samples were then saturated with the solutions, and the HS-SPME analysis was performed using the same conditions above. After each extraction, the fiber was injected into the GC for further analysis. All tests were carried out in triplicate. Finally, the reference curves were plotted according to the results.

GC/MS instrumentation and analysis conditions
A Hewlett-Packard 6890 gas chromatograph was used for all chromatographic analyses, coupled with an MS detector (Hewlett-Packard 5973). The GC fused silica capillary column was an HP-5 (5%-phenyl)-methylpolysiloxane (J&W Scientific, Folsom, USA) nonpolar column (60 m × 0.25 mm i.d., with 1 μm film thickness) and the carrier gas used was ultra-high purity helium, at a flow rate of 0.8 mL/min. The injector temperature was maintained at 290 °C, and the injection volume was 1.0 μL in splitless mode. After the extractions were carried out, the fiber was exposed to the injection port of the GC unit. The oven temperature was held at 70 °C for 3 min, increased to 240 °C at 30 °C/min, and increased again to 270 °C at a rate of 15 °C/min. The acquisition was performed in the MS, operating with electron impact ionization (EI, 200 eV) and 2.92 scans/s in a 29-550 (m/z) mass range, while the transfer line temperature was held at 220 °C.
The identification of the compounds was achieved using the Wiley 7, NIST 2005 mass spectral library and was further supported by the comparison of their linear retention indices (Kovats indices) to those of the reference standards and published data [34,35].

Statistics
All statistical data analyses were performed using SAS 9.4 software (SAS Institute Inc., Cary, NC, USA).

Research involving human participants and animals
This article does not contain any studies with human or animal subjects.

Recoveries
According to the updated SANTE Directive (Document N o SANTE/12682/2019) [36], the recoveries for the substances studied should range from 70 to 120% in all contaminated samples with relative standard deviations of ≤ 20%. The limits of detection (LODs) for the standard solutions of the 13 contaminants were determined at a signal-to-noise ratio of 3. The LODs and the limits of quantification (LOQs) for the concentrations tested ranged from 5 to 500 μg kg −1 and from 100 to 1000 μg kg −1 , respectively.

SE
The SE method recovery rates of 11 out of 13 substances for all concentrations ranged between ~ 70 and ~ 124% (Table 1). NPH was not recovered at all, and ANA was recovered at a meager percentage.

UAE
The UAE method recovery rates of 8 out of 13 substances for all concentrations ranged between ~ 50 and ~ 134% ( Table 2).
NPH and B[b]FLAN could not be recovered at any concentrations, while FL, BAN, and CH were recovered at overall rates below 30%.

HS-SPME
The HS-SPME method recovery rates of 8 out of 13 substances for all concentrations ranged between ~ 70 and ~ 105% (Table 3). Contrary to the two methods examined, NPH was recovered at all concentrations; however, its recovery rates were relatively low (~ 35%). B[k]FLAN was  Table 3 Paper recovery rates (RRs) and standard deviation (SD) of the 13 contaminants at three concentration levels (125 μg kg −1 , 300 μg kg −1 , and 500 μg kg −1 ) using the HS-SPME extraction method

Accuracy of methods
The accuracy of the methods was expressed as approximate trueness since Certified Reference Materials (CRMs) were not utilized in this study. The acceptable fluctuation range of concentrations was from − 20 to + 10%, as their values exceeded 10 μg kg −1 . The trueness of the methods was evaluated for a concentration of 125 μg kg −1 . The results of these tests are presented in Table 4.
According to Table 4, HS-SMPE exhibited the highest accuracy at 125 μg kg −1 as 7 out of the 13 contaminants are within limits while concerning the SE and UAE methods, only 4 and 5 of them are, respectively within limits.

Precision of methods
The Horwitz Ratio (HorRat) equation was utilized to assess the precision of the methods [37]. For the calculation of the HorRat value, the relative standard deviation (RSD) at a concentration of 125 μg kg −1 was calculated. The main criteria for the acceptance of a method are: • HorRat ≤ 0.5: excellent precision • HorRat ≥ 2: unacceptable precision For intermediate values, the method is accepted with appropriate documentation.
RSD is a standardized measure of the dispersion of a probability distribution which is often given as a percentage (%RSD) and is expressed as the ratio of the standard deviation to the mean. High RSD rates indicate resulting values spread out from the mean, whereas lower rates signify more precise data measurements. The experimental %RSD values and the HorRat values for all 3 methods tested are given in Table 5.
The estimated RSD and HorRat values attest to the fact that the HS-SPME is a more precise method than the other two.

ANOVA analysis
The analysis of variance (ANOVA) was performed on 20 spiked samples to assess the differences between the average values for each of the 10 out of 13 studied compounds,

Correction factors
Correction factors were determined to further optimize accuracy after concluding that the HS-SPME method is the most reliable of the 3 for the respective analysis. The correction factors were calculated by estimating the average value of 20 contaminated samples at a concentration of 125 μg kg −1 ( Table 6).

Applicability of the method
The method's applicability for the simultaneous determination of all 13 studied contaminants was verified by the extraction of the compounds in three commercial paperboards (Table 7) intended for use as packaging materials in direct contact with foodstuff (R1, R2, R3). Quantification was performed using the equations derived from the calibration curves.
According to the results, the R1 paperboard sample was devoid of all studied contaminants, followed by sample R3 with 2 found contaminants, and lastly, sample R2 with 3.

Conclusions
The results revealed that all the extraction methods tested performed well in terms of precision; all showed an overall repeatability relative standard deviation (RSD) of below 2.5%. Additionally, the HorRat values of all methods were within the limits of performance acceptability, ranging from 0.04 to about 0.1. However, considering the methods' accuracy and the ANOVA analysis, the HS-SPME was the technique that stood out and altogether demonstrated the best analytical results.
The developed HS-SMPE analysis, coupled with the GC/MS method, is not only simple and rapid but also sensitive enough to identify contaminants, such as PAHS, in recycled paperboard at low concentrations. Furthermore, HS-SMPE flaunts various environmental benefits; it is solvent-free, energy-efficient, and reusable, and it produces zero waste. The implementation of this analysis in the line of food packaging production could ensure that potential risks regarding the recovered materials are tackled in time before they come to form the final product, and without simultaneously putting a financial burden on producers.

3
This way consumers, as well as producers, would feel less apprehensive about the use of recycled materials in food packaging and they could fully embrace the benefits of the circular economy.