Fast and Green Extraction Method Based on HS–SPME/GC–MS to Identify Chemical Markers of X-Ray Irradiated Hen Eggs

Featured Application

HS–SPME/GC–MS method is a rapid, sensitive, selective and green analytical method to determine the radiolytic markers 2–DCB and 2–TCB in irradiated hen eggs, and thus suitable in food safety control programs.

Abstract

Food irradiation is a clean, safe and non-thermal technology applied to destroy pathogenic microorganisms, i.e., Salmonella spp., in hen eggs. Currently, in Europe only the egg white can be irradiated up to 3 kGy, so different control methods are crucial for official inspections to identify illicit treatments. In this work, an analytical method was proposed to determine the radiolytic markers, namely 2–dodecylcyclobutanone (2–DCB) and 2–tetradecylcyclobutanone (2–TCB) in hen egg samples. This method is based on headspace solid phase micro-extraction coupled with gas chromatography/mass spectrometry (HS–SPME/GC–MS). The eggs were treated by an X-ray irradiator at dose levels of 0.5, 1.0 and 3.0 kGy. The preliminary validation showed good selectivity, without matrix interferences in non-irradiated samples. Spiked samples showed linear responses in the range 2.5–25.0 µg kg⁻¹, where 2.5 µg kg⁻¹ was the limit of detection for both analytes. Irradiated samples showed a dose-dependent increase in signal intensity and a constant 2–DCB/2–TCB ratio. The minimum dose level detected was 0.5 kGy for all samples, and the 2–DCB and 2–TCB signals remained stable over one month after irradiation. Not least, white analytical chemistry was used to evaluate the HS–SPME/GC–MS method validation effectiveness, greenness power and economic efficiency, compared to the EN 1785:2003 standard method. The results of this study prove that the HS–SPME/GC–MS method is a reliable green alternative to the official method, which is suitable in food safety control programs.

1. Introduction

Eggs are one of the most nutritionally complete protein sources, providing all essential amino acids in optimal proportions. They are particularly rich in sulfur-containing amino acids, such as methionine and cysteine, which are fundamental for protein synthesis and cellular metabolism [1]. However, the consumption of raw or undercooked eggs is often associated with food safety risks, primarily due to the potential contamination by pathogenic microorganisms, including Salmonella spp. [2]. According to the European Union One Health 2023 Zoonoses Report from the European Food Safety Authority [3], approximately 77,486 cases of foodborne illness were attributed to Salmonella in the European Union, with eggs and egg-based products playing a significant role as the main vehicle of infection.

Microbiological contamination of eggs can occur both at the surface level, with the presence of bacteria on the shell, and inside, due to pathogen penetration through the pores of the cuticle. Conventional sanitation strategies, such as washing or using brushes and water jets, can reduce surface contamination, but they can damage the natural protective barrier of the shell, thus facilitating microorganism’s penetration into the egg [4].

Irradiation could represent a valid alternative for preserving the qualitative characteristics of eggs during storage, offering the advantage of reducing the microbial load without significantly altering the nutritional and functional properties of the egg. Several studies have demonstrated the effectiveness of ionizing radiation treatment in reducing the levels of Salmonella spp., Listeria monocytogenes and Escherichia coli on eggs and egg-derived products. Specifically, treatments with doses ranging from 1 to 3 kGy reduce the microbial load to undetectable levels without changing the organoleptic and nutritional characteristics of the product [5].

Sokovnin et al. (2018) [6] showed that the use of low-energy electron beams reduce the bacterial contamination without altering the physicochemical structure of the eggs. Irradiation with doses of 5 kGy effectively eliminates the microflora without affecting the internal properties of the egg, making this method promising for the poultry industry, especially for treatment of packaged products.

In addition, irradiation can also find technological applications in the bakery and confectionery industry. It is known that irradiation causes protein cleavage, and this change may be the main reason for the improved foaming ability of egg white [7].

Although egg irradiation is permitted in many non-European countries, such as the United States and some Asian countries, it is not yet authorized in Europe. Currently, no European state allows the irradiation of whole eggs, although some countries permit the irradiation of egg whites up to a maximum dose of 3 kGy.

In the European Union, the availability of reliable analytical methods for identifying the treatment is a fundamental pre-requisite for the authorization of irradiation. European legislation [8,9], in fact, states that the products treated with ionizing radiation must be labeled as irradiated, and the label compliance must be verified every year by each Member State through official control plans. It is therefore essential to strengthen the ability to identify irradiated eggs and egg-based products, using all necessary tools available to protect the market from non-compliant irradiated products, and to guarantee the consumer free choice.

The European Committee for Standardization (CEN) standardized several analytical detection methods based on radiation-induced physical, chemical or biological changes in foods, to identify the irradiation treatment [10,11,12,13]. Electron-paramagnetic resonance (ESR) spectrometry is a physical method for detecting ionizing radiation treatment, which is already being used to control products containing hydroxyapatite crystals, such as eggshells. The ESR method has been used not only to indicate irradiated products, but also as dosimetry. In addition, the ESR signal suffers from fading, which indicates a significant error in the measurement of absorbed dose after a long period of time from irradiation treatment [14]. Among the other available analytical techniques, gas chromatography coupled with mass spectrometry is applicable to determine of 2–alkylcyclobutanones (2–ACBs), in particular 2–dodecylcyclobutanone (2–DCB) and 2–tetradecylcyclobutanone (2–TCB), two markers of irradiation in fat-rich food [15]. In this regard, the standard method EN 1785:2003 [15] is based on Soxhlet extraction using hexane, sample clean-up by florisil column, gas chromatography (GC) separation of analyte and detection by mass spectrometry (MS). This method is solvent- and time-consuming, resulting complex for routine analysis and it is limited to the analysis of whole eggs irradiated at 1 kGy and above. Several other approaches have been used or proposed to reduce sample processing time, as well as reducing solvent costs and the amount of waste [16,17]. Recently, an alternative method to detect 2–DCB, based on the static headspace solid phase micro-extraction (HS–SPME) and GC–MS analysis procedure, was proposed for the analysis of irradiated meats and dairy products [18,19]. This approach has proven to be one of the most promising methods for detecting food irradiation. In the present study, conducted as part of a project funded by the Italian Ministry of Health, in progress at the Italian National Reference Laboratory for the treatment of foods and their ingredients with ionizing radiation, egg samples from Italian farms were subjected to irradiation with X-rays at doses ranging from 0.5 to 3 kGy and then analyzed. The analysis was carried out using the HS–SPME/GC–MS technique to determine 2–DCB and 2–TCB as irradiation markers.

This work presents the results obtained by applying the gas chromatographic method, in order to evaluate its reliability and sensitivity for the identification of irradiated eggs. This research involves optimization of sample preparation and preliminary validation studies that include evaluation of the following parameters: instrument linearity, method linearity, selectivity, limit of detection (LOD), minimum detectable dose level (MDL), diagnostic sensitivity and specificity and stability of the signals.

Furthermore, the method proposed in this paper was further evaluated in terms of validation effectiveness, greenness and economic efficiency. The concept of green analytical chemistry (GAC) has evolved over the past two decades, promoting environmentally friendly and safe analytical practices [20]. In 2013, Galuszka et al. [21] formalized 12 principles which include minimizing reagent use, energy consumption, waste generation and operator exposure, while promoting automation and miniaturization. However, ensuring a balance between sustainability, analytical performance and practical utility remains complex. In order to address this, the RGB (red, green and blue) model was introduced, integrating three dimensions: red (analytical performance), green (environmental safety) and blue (practical/economic aspects) [22,23,24]. This framework led to the development of white analytical chemistry (WAC), which proposes 12 new principles across these three pillars (

Table 1. List of 12 WAC principles relating to analytical efficiency, practical and economic criteria [20].

The 12 Principles of White Analytical Chemistry (WAC)
RED PRINCIPLES (analytical performance)	GREEN PRINCIPLES (green chemistry)	BLUE PRINCIPLES (practical side)
R1: Scope of application	G1: Toxicity of reagents (impact and biodegradation)	B1: Cost-efficiency
R2: LOD and LOQ	G2: Amounts of reagents and waste	B2: Time-efficiency
R3: Precision	G3: Consumption of energy and other media	B3: Requirements
R4: Accuracy	G4: Direct impacts (safety, use of animals and GMOs)	B4: Operational simplicity

), as an alternative to the known 12 GAC principles, balancing analytical rigor, eco-safety and practicality [25]. To evaluate the degree of whiteness of the method, the RGB12 algorithm, developed by Nowak et al., 2021 [20], demonstrated the reality of mixing red, blue and green colors to obtain white light. The WAC principles were used to evaluate the HS–SPME/GC–MS method compared to the EN 1785:2003 standard method.

The aim of this research is to contribute to the improvement of food safety controls and consumer protection by providing an effective analytical tool for monitoring irradiation in hen eggs and egg-based products.

2. Materials and Methods

2.1. Samples and Chemicals

Hen egg samples were purchased from local markets and stored at room temperature. The standards of 2–DCB (purity ≥ 95.0%), 2–TCB (purity ≥ 95.0%) and 2–cyclohexylcyclohexanone (purity ≥ 95.0%) were purchased from Sigma–Aldrich (Buchs, Switzerland). As indicated in the EN, 1785:2003, 2–cyclohexylcyclohexanone was employed as internal standard (IS). Ethanol was purchased from Panreac (Barcelona, Spain), and ultrapure water was obtained from a Milli–Q purification system (18.2 MΩ × cm) (Millipore, Milan, Italy). Stock standard solutions 1000 mg L⁻¹ in ethanol of 2–DCB, 2–TCB and IS were prepared and stored at −20 °C. A working standard solution (5 mg L⁻¹) in water of IS was prepared and used to spike samples at 10 μg L⁻¹. The IS was employed as Quality Control (QC) to assess the system’s suitability and stability, and the repeatability of the data acquisition process. Working standard solutions (1 and 4 mg L⁻¹) in water of 2–DCB and 2–TCB were used to spike 5 mL of water to obtain the solvent-matched calibration (SMC) curve, in the concentration range between 5.0 and 40.0 μg kg⁻¹. The non-irradiated egg samples were spiked at the following four concentration levels: 1.25, 2.50, 12.5 and 25.0 μg kg⁻¹, to obtain matrix-matched calibration (MMC) curve.

2.2. Irradiation Procedure

For the irradiation treatment, shell eggs were irradiated in a food-grade polyethylene plastic bag. The irradiation was performed using a low-energy X-ray irradiator (RS–2400, Radsource Inc., Suwanee, GA, USA) operating at 150 kV and 45 mA at room temperature. The dose levels used for X-ray radiation treatment were 0.5, 1.0 and 3.0 kGy, at a dose rate of approximately 2 kGy h⁻¹. Aliquots of non-irradiated samples were used as control (NI). Non-irradiated shelled and homogenized whole eggs were stored in 50 mL falcon and frozen at −20 °C until analysis. The irradiated samples, with shells and stored at room temperature, were immediately analyzed at time T₀ and after 1 month (time T₁) for stability evaluation. All egg samples were shelled before analysis.

2.3. Sample Preparation

Two g of irradiated and non-irradiated shelled whole egg samples were weighed in a 20 mL glass headspace vial, then 5.0 mL of ultrapure water and 10 μL of 5 mg L⁻¹ IS working solution were added. Appropriate amounts of 2–DCB and 2–TCB working standard solutions were added in non-irradiated samples to obtain four concentration levels (1.25, 2.50, 12.5 and 25.0 μg kg⁻¹). The sealed vials were shaken manually and vigorously for a few seconds, then sonicated for 10 min at room temperature at the automatically set frequency (28–34 kHz) (LIARRE s.r.l., Casalfiumanese, Bologna, Italy) and immediately analyzed.

2.4. Instrument and HS–SPME/GC–MS Method

The non-polar polydimethylsiloxane (PDMS) coated fiber (100 μm film thickness and 23 gauge (Supelco, Bellefonte, PA, USA)) was found to be the most suitable for 2–ACBs extraction [26] and it was therefore used in the present work. The static headspace extraction was carried out with the support of a Triplus RSH autosampler (Thermo-Fisher Scientific, Waltham, MA, USA). The HS–SPME conditions included an enrichment time of 30 min at 80 °C, and an extraction time of 60 min at 80 °C. Before and after analysis, the PDMS fiber was thermally conditioned and washed, heating the fiber at 250 °C for 30 and 20 min, respectively.

The analysis was performed by using a GC–MS system consisting of a TRACE GC Ultra gas chromatograph equipped with a programmed temperature vaporizer (PTV) injector and coupled with a triple quadrupole TSQ Quantum mass spectrometer (Thermo–Fisher Scientific, Waltham, MA, USA). The most suitable column, matching the non-polar fiber, was a 5%-diphenyl-95% polydimethylsiloxane capillary (30 m × 0.25 mm i.d. × 0.25 µm; Restek S.r.l., Cernusco sul Naviglio (MI), Italy). Taking into account the analytes volatility, a lower initial column temperature helps to condense the volatile fraction, improving the chromatographic peak resolution. A proper temperature ramp, starting with a separating phase and ending with a high-temperature cleaning phase, enhances the analyte/matrix interference separation, and avoids carry-over. The temperature program was set as follows: T1 = 50 °C, hold = 4 min; T2 = 230 °C, rate1 = 15 °C/min, hold = 1 min; T3 = 310 °C, rate2 = 16 °C/min, hold = 1 min.

The total chromatographic run time was 23 min. Helium (99.9995% purity, Sapio s.r.l., Monza, Italy) was used as a carrier gas with a constant flow of 1.0 mL min⁻¹. For mass spectrometry detection, the conditions recommended by the EN 1785:2003 for ACBs analysis in irradiated fat-containing foods were adopted, monitoring ions m/z 98 and m/z 112 in SIM mode. These extraction, separation and detection parameters were previously employed in our laboratory for studies involving 2–DCB and 2–TCB extraction from animal and dairy matrices [17,18,19] and were thus applied in this research. The analysis, including extraction, was performed in triplicate. During each analytical sequence, in addition to the IS, a 2–DCB and 2–TCB reference standard solution (2 μg L⁻¹ in 5.0 mL of ultrapure water) was analyzed to confirm the retention time and to assess the stability of the SPME fiber and its ability to extract the analytes. Blank runs were also carried out using 5.0 mL of ultrapure water, to verify the fiber carry-over. During our experiments, the fiber did not deteriorate and/or suffer from carry-over.

2.5. Assessment of 2–DCB and 2–TCB

The presence of 2–DCB and 2–TCB, as chemical markers of irradiation treatment, was determined following the criteria reported in the standard method EN 1785:2003 and in our previous works [17,18,19]. Briefly, the sample was classified as irradiated if all the three following conditions were verified: chromatographic peak at a retention time comparable (±0.5%) with the 2–DCB and 2–TCB retention time of reference standard solution; signal-to-noise (S/N) ratios of this peak extracted at m/z 98 and 112 greater than 3; percentage ratio areas (RA %) values of the peak at m/z 98 and 112 (RA % = A112/A98 × 100) within the range 22.2–25.0% for 2–DCB and within the range 23.8–26.3% for 2–TCB.

2.6. Evaluation of Method Analytical Performances

According to the ISO/IEC 17025:2017 [27], any analytical method must be validated before being applied in official laboratories. In this study, preliminary validation experiments were performed to verify the following: instrument linearity, method linearity, selectivity, LOD, MDL, diagnostic sensitivity and specificity and stability. These parameters were evaluated to verify if this method is fit for purpose. The linearity of this analytical procedure for 2–DCB and 2–TCB was evaluated within the concentration range of SMC, to verify the direct proportionality of instrumental response. Moreover, linearity was also evaluated within the concentration range of MMC, and the matrix effect was determined by means of samples irradiated at doses of 0.5, 1.0 and to 3.0 kGy. Method selectivity, that is, its ability to identify specifically 2–DCB and 2–TCB in the presence of interferences in the sample matrix, was checked by examining NI samples. In order to evaluate the LOD of the method, non-irradiated samples were spiked with 2–DCB and 2–TCB standard solutions to obtain concentrations of 1.25, 2.50, 12.5 and 25.0 μg kg⁻¹. In our research, in addition to LOD, we also evaluated the MDL which represents the minimum irradiation treatment dose that can be detected by the HS–SPME/GC–MS method. This parameter was evaluated by examining the irradiated samples. The diagnostic sensitivity was defined as the fraction of true positive samples (irradiated and not-irradiated samples) identified on the total number of positive samples. The diagnostic specificity was the fraction of true negative samples (irradiated and not-irradiated samples) identified on the total number of negative samples [28]. For the evaluation of diagnostic sensitivity and diagnostic specificity, 9 non-irradiated samples and 9 irradiated samples were used. The accuracy was defined as the fraction of correctly classified samples out of the total number of samples; more specifically, as the percentage of samples correctly identified as true positives and true negatives, i.e., irradiated and non-irradiated samples. The stability of the radiation-induced ACBs signal was studied by analyzing 0.5–1.0–3.0 kGy irradiated eggs immediately (T₀) and after 1 month (T₁) from treatment. The most important validation parameters are resumed in

Table 2. List of parameters and respective number of analyzed samples for validation procedure.

Parameters	2–DCB and 2–TCB (Concentration or Dose)	Number of Samples
Linearity	SMC: 5.0–10.0–20.0–40.0 (μg kg−1) MMC: 1.25–2.50–12.5–25.0 (μg kg−1)	12 12
Matrix effect	Irradiated samples: 0.5–1.0–3.0 (kGy)	9
Selectivity	Non-irradiated samples	9
LOD	Spiked samples: 1.25–2.50–12.5–25.0 (μg kg−1)	12
MDL	Irradiated samples: 0.5–1.0–3.0 (kGy)	9
Diagnostic sensitivity and diagnostic specificity	Non-irradiated samples Irradiated samples: 0.5–1.0–3.0 (kGy)	9 9
Stability	Irradiated samples T0: 0.5–1.0–3.0 (kGy) Irradiated samples T1: 0.5–1.0–3.0 (kGy)	9 9

.

Free software R version 4.1.1 (R Development Core Team, Vienna, Austria, 2020) was used for statistical analysis.

2.7. WAC Evaluation: Red, Green and Blue Models (RGB12)

To assess the compliance of the HS–SPME/GC–MS method with WAC principles, as compared to the EN 1785:2003 standard method, the RGB 12 algorithm was used. Nowak et al. [20] developed the Excel model of the RGB 12 algorithm and described the correct filling of the gray columns in relation to the red, green and blue principles. Briefly, the gray columns can be filled with the appropriate score from 0 to 100, where 0 was the worst result for a particular principle, and 100 the optimal. The criteria related to the use of animals and Genetically Modified Organisms (GMOs) must be filled using 1 if they are used at any stage of the procedure, and 0 if not. All scores should be assigned as objectively and reasonably as possible, taking into account a gradation of scores. For example, a score of 50 means that a given result appears to be unsatisfactory or tolerable; a score of 75 could be assigned in case of an insufficient result for all expected situations; and a score of 100 is given for a method that is appropriate regarding its target applications.

3. Results

3.1. Optimization of Sample Preparation

Our previous experience regarding the analysis of 2–DCB and 2–TCB by HS–SPME/GC–MS method involved solid matrices dispersed in water. During HS–SPME extraction, samples were incubated for a total of 90 min (30 min headspace enrichment + 60 min fiber exposure) at 80 °C. Regarding egg samples, the possible coagulation of proteins and lipids at extraction temperatures posed a challenge to obtain a satisfactory extraction of analytes. At the extracting temperature (80 °C), egg proteins and lipids coagulate completely (egg white at ~65–70 °C and yolk at ~70–75 °C) [29], making the now-solid matrix a barrier to analyte volatilization and extraction. The optimization aimed to identify the appropriate sample amount suitable for maintaining the egg/water mixture in a liquid or semi-liquid state. Tests using 5 g and 2 g of homogenized whole egg (irradiated and non-irradiated) demonstrated that 5 g of samples irradiated at 0.5 kGy did not produce a distinguishable signal, compared to the non-irradiated sample, while using 2 g produced the best results in terms of analytes area response. In order to obtain an identification of ionizing treatment as low as possible, the analysis was conducted using 2 g of sample.

3.2. Analytical Performances of HS–SPME/GC–MS Method

As described in Section 2.6, several parameters were considered: instrument linearity, method linearity, selectivity, LOD, MDL, diagnostic sensitivity and specificity and stability. A linear response for both analytes was observed in SMC at 5.0–10.0–20.0–40.0 μg kg⁻¹. Non-irradiated samples were spiked with 2–DCB and 2–TCB standard solutions to obtain concentrations of 1.25, 2.50, 12.5 and 25.0 μg kg⁻¹, and both analytes resulted detectable from 2.50 μg kg⁻¹, which was the lower level of detection (LOD), while concentration of 1.25 μg kg⁻¹ did not produce a detectable signal. All calibration curves showed a good linearity (R² ≥ 0.999) and an error of mean (σM), calculated as σ/√3, near or below 0.30 (Figure 1).

Method selectivity was verified using non-irradiated eggs, where no peaks related to 2–DCB or 2–TCB were found, and no matrix interferences were observed (Figure 2), highlighting that the HS–SPME/GC–MS method was selective for 2–DCB and 2–TCB identification.

MDL was evaluated on low-dose irradiated eggs and 0.5 kGy was found to be suitable for identifying the treatment for all samples by means of 2–DCB and 2–TCB detection. In a previous work, the EN 1785:2003 method was applied to processed poultry meat products, verifying that a dose of 0.5 kGy corresponds to 3.0 μg kg⁻¹ of 2–DCB [17]. In the present work, it was verified that irradiated eggs at 0.5 kGy corresponded to 11.0 μg kg⁻¹ of 2–DCB and 3.2 μg kg⁻¹ of 2–TCB, as mean values of three replicates, calculated on a matrix-matched calibration curve. These results reflect the content of the precursor of 2–DCB and 2–TCB, i.e., palmitic acid (27.3%) and stearic acid (17.4%), in whole eggs [30]. Regarding stability, the two analytes were correctly detected both at low dose levels and especially at 1.0 and 3.0 kGy, the doses commercially used for food treatments. As reported in

Table 3. Normalized value of 2–DCB and 2–TCB (A_98–ACB/A_98–IS), at different doses and the storage time.

Treatment	Time Elapsed from Irradiation	Normalized Value of 2–DCB (NDCB)	Mean ± St. dev of NDCB	Normalized Value of 2–TCB (NTCB)	Mean ± St. dev of NTCB	(NDCB)/(NTCB)
0.5 kGy	Immediately after irradiation	0.094	0.122 ± 0.024	0.013	0.013 ± 0.001	7.1
		0.137		0.013		10.1
		0.135		0.012		11.0
1.0 kGy		0.216	0.262 ± 0.045	0.027	0.029 ± 0.003	8.1
		0.265		0.032		8.4
		0.306		0.028		10.9
3.0 kGy		1.061	1.073 ± 0.029	0.091	0.099 ± 0.011	11.6
		1.051		0.094		11.1
		1.106		0.112		9.9
0.5 kGy	One month after irradiation	0.157	0.167 ± 0.009	0.014	0.017 ± 0.004	10.9
		0.170		0.021		8.0
		0.174		0.015		11.7
1.0 kGy		0.256	0.290 ± 0.029	0.025	0.028 ± 0.003	10.3
		0.307		0.031		10.0
		0.307		0.029		10.4
3.0 kGy		1.218	1.099 ± 0.136	0.104	0.100 ± 0.009	11.7
		0.950		0.089		10.7
		1.129		0.106		10.6
Mean						10.1
St. dev						1.4
CV%						13.4

, the normalized area values of 2–DCB and 2–TCB (A_98–ACB/A_98–IS) at different doses were stable and also depended on the storage time, as demonstrated by the CV% value of 13.3%, indicating a low dispersion around the mean value. The ability to detect analytes even after a long time makes the method robust and applicable throughout the average shelf-life of eggs (typically 3–5 weeks). A two-way ANOVA was conducted to verify the dependence of the ratio between the areas of normalized 2–DCB and 2–TCB ((N_DCB)/(N_TCB)) on dose, time and the interaction of the two variables. The test showed that, at a significance level of 0.05, there is no dependence of this ratio upon the indicated variables. Specifically, the estimated coefficients for the dependence on time, dose and the interaction between the two have a probability of being equal to zero of 0.28, 0.12 and 0.57, respectively. Overall, the p-value of the model (probability that all coefficients are equal to 0) was 0.24.

The signal stability of 2–DCB and 2–TCB in irradiated samples is in agreement with what was recently reported [31] about irradiated dairy products; irradiation can stabilize the profile of volatile compounds after treatment and during storage time.

Finally, regarding diagnostic sensitivity and specificity, the proposed method was accurate, with 100% of correctly identified samples, as in the EN 1785:2003, at the MDL and beyond. As demonstrated by Baykalir et al. (2020) [32], saturated fatty acid levels, in particular palmitic acid and stearic acid, were similar in eggs produced by both organic and conventional production systems. Furthermore, the production of 2–DCB and 2–TCB as irradiation markers was only correlated with the irradiation dose applied and the amount of precursor fatty acids. Thus, it can be concluded that multi-batch and multi-source egg samples do not affect the applicability of the HS–SPME/GC–MS method.

3.3. Red, Green and Blue Evaluation of HS–SPME/GC–MS Method: Comparison with the Standard Method EN 1785:2003

The proposed HS–SPME/GC–MS method presents significant advantages with respect to the standard method EN 1785:2003, with respect to the RGB principles, as described in Table 1. This section reports an in-depth analysis of the analytical, chemical and practical aspects of the methods being compared.

3.3.1. Red Principles

• R1: Scope of application
  The R1 principle was excellently satisfied for both methods as they identify the two irradiation markers, 2–DCB and 2–TCB. Furthermore, the two compared methods are applicable to different matrices [15,17,18,19] and therefore, we assigned score of 100 to both.
• R2: LOD and LOQ
  LOD and LOQ were estimated more specifically in terms of the lower level of 2–DCB and 2–TCB in spiked and/or irradiated samples. In the EN 1785:2003, the detection of irradiated liquid whole egg has been validated for doses of approximately 1 kGy and above. The standardized method has been extensively studied and tested, and the literature data support its validity. Regarding the identification of radiation treatment, the LOQ is often expressed as the minimum irradiation dose that can be demonstrated. Kim and others [33], using a method close to the EN 1785:2003, reported low levels (14 μg kg⁻¹) of 2–DCB in irradiated dried shrimp. In a previous work, the EN 1785:2003 method was applied to processed poultry meat products and it was proven that a dose of 0.5 kGy corresponds to 3.0 μg kg⁻¹ of 2–DCB [17]. Based on these considerations, the analytical performances of the two methods were found to be comparable only for 2–DCB. For these reasons, we scored LOD and LOQ with 75 points because the EN 1785:2003 does not provide sufficient results. Conversely, the proposed HS–SPME/GC–MS method demonstrated a lower LOD of 2.50 μg kg⁻¹, and a minimum detectable dose level of 0.5 kGy for both irradiation markers. Thus, a score of 100 was assigned.
• R3: Precision
  Precision is expressed as repeatability and reproducibility of the results. The standard method was validated by inter-laboratory blind trials, so both repeatability and reproducibility were satisfied [34]. In the proposed new method, repeatability was only carried out during intra-laboratory validation. For these reasons, we assigned a value of 100 to the EN 1785:2003 and 75 to the HS–SPME/GC–MS method.
• R4: Accuracy
  The R4 was evaluated in terms of relative error and recovery. The recovery was evaluated by Campaniello et al. (2025) [17] for the standard method applied to chicken meat, resulting in 67.5% for spiked samples at 40 μg kg⁻¹. The EN 1785:2003 does not provide relative errors and recovery. Regarding the HS–SPME/GC–MS method, recovery was not calculated, since it is a qualitative method, while error of mean (σM), calculated with spiked samples analyzed for MMC, was near or below 0.30. Since both methods missed a parameter, a score of 75 was assigned for both.

3.3.2. Green Principles

• G1: Toxicity of reagents
  The RGB 12 algorithm determines the toxicity of reagents, i.e., their impact and biodegradation, by means of the total number of pictograms of reagents employed. For both methods tested, only the reagents were evaluated, excluding the 2–DCB and 2–TCB standards. The Soxhlet extraction which is very long (6 h) and uses a flammable solvent (hexane) [15] represents a major criticism of the EN 1785:2003 method. A further criticism concerns the clean-up via the Florisil^® column, which requires hexane and 1% diethyl ether in hexane to clean-up and elute the 2–DCB. Considering these two steps, a total of six pictograms were recognized (score: 20), against none for the HS–SPME/GC–MS method (score: 100).
• G2: Amount of reagents and waste
  As described in the previous section, the EN 1785:2003 method shows an extensive use of solvents, which results in resource consumption and waste generation because during extraction and purification, approximately 500 mL of solvents were used and discarded. The HS–SPME/GC–MS method proposed, using only water for analyte extraction, reduces the use of solvents and the production of waste. The scores for the G2 parameter were the same as G1.
• G3: Consumption of energy and other media
  This point considers the energy expenditure of the analytical method. We only evaluated the sample preparation, since the GC–MS step was identical for both methods. The HS–SPME/GC–MS method involves a headspace extraction that includes an enrichment time of 30 min, an extraction time of 60 min and fiber conditioning before and after each cycle of 30 and 20 min, respectively. Although this automated procedure takes more than two hours, it is less than the six hours required for Soxhlet extraction and the time needed for sample purification, without considering the additional time required for the solvent evaporation steps. For these reasons, we assigned a value of 50 to the EN 1785:2003 and 75 to the HS–SPME/GC–MS method.
• G4: Direct impacts
  The principle “direct impacts” were determinate in terms of safety of users, use of animals and GMOs. As clarified in Section 2.7, the criteria related to the use of animals and GMOs were filled with zero because they were not used at any stage of the procedure. Even in this case, as occurred in the evaluation of the G1 and G2 parameters, the professional risk is linked to the use of solvents employed exclusively in the EN 1785:2003 standard, for which we attributed a score of 20 to the standardized method and a score of 100 to the HS–SPME/GC–MS method.

3.3.3. Blue Principles

• B1: Cost-efficiency
  When investigating the application of different methods compared to the standard method EN 1785:2003, consideration of cost and availability of instrumentation is needed. Using alternative extraction procedures, such as automated HS–SPME, would be faster, but require a significant starting investment for equipment purchase (15,000.00 € cost presumed for the purchase of SPME autosampler). The Soxhlet method has a relatively low capital cost, estimated at around 1000.00 € for the instrumentation, without considering the cost of solvents used for each analysis, and the economic gap with sophisticated instruments cannot be evenly compensated by the costs of solvents (around 60.00 € for 500 mL). For the B1 parameter, the EN 1785:2003 received a score of 100, while 75 was assigned to that proposed in this study.
• B2: Time-efficiency
  The time-efficiency of a method is related to its speed of analysis. The longer extraction times required by the standardized method have already been commented in section “G3: Energy consumption and other media”. This characteristic, in addition to influencing energy consumption, also affects the number of samples that can be analyzed in a single batch. The HS–SPME/GC–MS method has also been proposed to reduce sample processing time. In fact, using the EN 1785:2003 method, it is possible to analyze an average of three samples per day (eight hours for each sample), while the method proposed in this work processes approximately ten samples per day (two and a half hours per sample), in a completely automatic way. For these reasons, we assigned a value of 25 to the EN 1785:2003 and 100 to the HS–SPME/GC–MS method.
• B3: Requirements
  Analytical methods should be characterized by the minimal practical requirements, including the amount of sample used, access to advanced equipment, personnel qualifications and laboratory infrastructure [20]. In our case, the amount of sample used in the two methods, although different, is irrelevant, since only a few grams are used. The difference is evident with regard to the use of sophisticated technologies and the employment of qualified personnel. The crucial point is once again the sample preparation which, in the case of the EN 1785:2003 method, requires simple instrumentation and unskilled personnel. In contrast, the HS–SPME extraction requires more expensive instrumentation (autosampler for headspace extraction) and adequately trained personnel. The B3 parameter was expressed as a mean value, resulting in 100 for the EN 1785:2003 and 87.5 for HS–SPME/GC–MS method.
• B4: Operational simplicity
  The overall evaluation of analytical methods should take into account their impact on the level of usability, resulting from the sum of miniaturization, integration, automation (online methods, artificial intelligence technologies) and portability (on-site measurements). Taking into account that the methods compared in this section are not portable, we evaluated this parameter as zero for both. The EN 1785:2003 cannot be considered miniaturizable, although Stewart et al. [35] demonstrated that the florisil cartridge with solid phase extraction (SPE) was a reliable method for 2–DCB purification in irradiated chicken meat, liquid whole eggs, minced beef and mango. Furthermore, the standard method cannot be defined as integrated and automated because the operator’s assistance is constantly required during the analytical flow. On the contrary, regarding the proposed method, the HS–SPME, together with the use of the autosampler, can be considered a miniaturization and automation of the extraction and purification steps, reducing manual labor and improving the speed and reproducibility of sample preparation and analysis. As with the B3, the B4 parameter was also expressed as mean value, resulting in 16.7 for the EN 1785:2003 and 66.7 for the HS–SPME/GC–MS method. The results were resumed in Figure 3.

4. Discussion

The proposed method HS–SPME/GC–MS aims to identify ionizing radiation treatment in eggs by means of 2–DCB and 2–TCB identification. This method, previously validated for animal and dairy matrices in our laboratory, was optimized and adapted to a hen eggs matrix. The preliminary validation showed good selectivity, without matrix interferences in non-irradiated samples. Spiked samples showed linear responses in the range 2.50–25.0 µg kg⁻¹, where 2.50 µg kg⁻¹ is the LOD for both analytes.

Irradiated samples showed a dose-dependent increase in signal intensity and a constant 2–DCB/2–TCB ratio. Validation at doses above 1 kGy covers the majority of food-processing applications. However, there is a need for validated methods that perform with a lower level of detection in order to detect irradiation of ingredients or aged samples, and to improve confidence regarding the absence of irradiation of non-irradiated labeled foods. To satisfy this request, MDL was evaluated on low-dose irradiated eggs and 0.5 kGy was found to be suitable for identifying the treatment for all samples by means of 2–DCB and 2–TCB detection. Furthermore, radiation-induced ACB signals remained stable for one month in eggs irradiated at three dose levels, i.e., 0.5–1.0–3.0 kGy.

Not least, white analytical chemistry was used to evaluate the HS–SPME/GC–MS method validation effectiveness, greenness power and economic efficiency, compared to the EN 1785:2003 standard method. The red, green and blue (RGB12) models, in fact, provide a holistic view, taking into account the analytical (red), green (green) and practical (blue) aspects that influence the quality of the method. The HS–SPME/GC–MS method showed its prevalence in greenness and blueness aspects. The high values of 93.8 for green and 82.3 for blue (Figure 3) were mainly related to the sample preparation by headspace solid phase micro-extraction. In fact, HS–SPME does not require toxic solvents, does not produce waste and is safe for the operator, and these characteristics are reflected in the ecological aspect. The blueness was influenced by positive aspects, such as reducing sample preparation and processing time, miniaturization and automation and negative features, i.e., expensive instrumentation and trained operator. The parameter called “whiteness” (the white bar shown in Figure 4) is the result of the overall conformity of the method and represents the measure of how the method fits well with the intended application.

The WAC approach reports the evaluation of “whiteness” parameter in gray scale, as a percentage (Figure 3) and as withe bar plot (Figure 4): the HS–SPME/GC–MS method had a score of 87.8 and lightest gray and the EN 1785:2003 standard method showed darkest gray with 62.9 points. These results are effective in demonstrating the better performances of the automated HS–SPME/GC–MS method over the standard EN 1785:2003 method.

5. Conclusions

This work demonstrated the feasibility and the whiteness of HS–SPME/GC–MS for identifying irradiation treatment of eggs. In particular, the proposed method showed good appropriateness of validation parameters, such as instrument linearity and method linearity (R² > 0.999), selectivity, LOD (2.5 µg kg⁻¹), MDL (0.5 kGy), diagnostic sensitivity and specificity and stability. Moreover, the WAC approach highlighted the better performances of HS–SPME/GC–MS method, compared to the EN 1785:2003 standard method. The method described in this paper finds its application in laboratories to improve food safety control and consumer protection, by monitoring the irradiation of egg products. Considering the official role of the Italian National Reference Laboratory for the treatment of foods and their ingredients with ionizing radiation in the surveillance of irradiated foods, continuous updating of the applied methods is essential because other matrices may require optimization of the analytical procedure. Furthermore, the HS–SMPE/GC–MS method must always be tested and validated on different matrices to ensure the official nature of controls on irradiated foods. It is important to underline that at the moment, the main advantage of this method is its applicability to animal matrices of different natures (currently: meat products, cheeses and dairy products, eggs) and in the future we foresee its extension to other matrices (exotic fruit, spices and novel foods) and also to other irradiation markers, such as unsaturated hydrocarbons.