**2. Results**

### *2.1. Method Validation*

The patulin contamination in fruits and fruit-based products is a worldwide problem, and effective control of PAT strongly depends on reliable analytical methods. The validation of the analytical method for PAT included determination of linearity, recovery, precision (repeatability and reproducibility), and sensitivity (limit of detection, LOD, and limit of quantification, LOQ). Linearity was checked by injection into HPLC-UV of PAT standards in the range from 5 to 100 μg/<sup>L</sup> (Table 1), the correlation coefficient obtained was 0.9916. The average retention time of PAT was 6.383 ± 0.05 min with good coefficient of variation (0.75%). Recovery experiments were done by spiking negative samples of mango and orange at PAT concentrations of 10, 50, and 100 μg/kg. After 1 h, the spiked samples were processed and analyzed by HPLC. The average recovery was 97.4%, with good values for repeatability (relative standard deviation, RSDr less than 5%) and reproducibility (RSDR less than 15%). The limit of detection (LOD) and quantification (LOQ) were determined by signal-to-noise ratio and were 5 μg/kg and 15 μg/kg, respectively. In conclusion, the analytical method used allowed for accurate quantitative determination of patulin in mango and orange samples and fulfilled performance requirements of Commission Regulation (EC) No. 401/2006 [30].

**Table 1.** Linearity of standard working solutions of patulin (PAT). Values are mean of triplicate analysis.


The analytical method used is based on AOAC method 995.10, which was successfully validated through collaborative studies for patulin determination in apple products [31]. In detail, the method consists of four steps, including liquid–liquid extraction with ethyl acetate, sodium carbonate clean-up, sodium sulfate drying, and LC-UV determination. Na2CO3 neutralization is used to lower interference from the phenolic compounds in fruit matrices, such as the 5-hydroxymethylfurfural (5-HMF) [32]. The main shortcoming of the method is the presence of interfering matrix components that might affect chromatographic separation. To better remove interferences for patulin determination, a series of representative random samples were additionally subjected to a second purification step using multifunctional clean-up columns MFC 228. The chromatographic separation was improved, though no significant differences were observed in recovery percentage and patulin concentration. Figure 2 represents HPLC chromatograms of natural occurrence of patulin in mango and orange samples.

**Figure 2.** HPLC chromatograms of natural occurrence of patulin in mango sample (**A**) and orange sample (**B**).

#### *2.2. Occurrence of Patulin in Fruits and Derived Products*

Results of PAT occurrence in 133 samples of mango fruits and derived products are shown in Table 2. A total of 70 samples of mango fruits along with 63 samples of mango-based products (juices, pulp, and jam) were randomly collected from di fferent sites of Punjab, Pakistan. From the data, it is evident that 82 samples were found PAT-contaminated with an incidence level (61.7%) and a total mean concentration of 110.9 μg/kg. The percentage of positive samples of Faisalabad, Sheikhupura, Multan, Shorkot, and Rawalpindi was 50%, 33.3%, 53.3%, 71.4%, and 40%, respectively, whereas the percentage of contamination in mango juice, pulp, and jam was 75%, 87.5%, and 60%, respectively. The average PAT levels in mango fruits were 348 μg/kg, 42.6 μg/kg, 87.5 μg/kg, 254.2 μg/kg, and 14.7 μg/kg in samples collected from Faisalabad, Sheikhupura, Multan, Shorkot, and Rawalpindi, respectively. The average PAT levels in mango juices, pulp, and jam of di fferent brands were 24.3 μg/kg, 82.3 μg/kg, and 5.0 μg/kg, respectively. Although the incidence of patulin was very similar between samples of mango fruit and derived products, the concentration was higher in the first (186.6 μg/kg) compared with the second (26.9 μg/kg). Among the mango fruit samples, it is noteworthy a sample from a Faisalabad's local market with an extremely high PAT content (6415 μg/kg), as well as another sample from Shorkot with a very high PAT content of 2030 μg/kg. In the present study, healthy mango fruits were less contaminated with PAT in comparison with decayed ones.


**Table 2.** Incidence of patulin (μg/kg) in mango fruits and derived products.

Table 3 reports the PAT incidence and concentration in 141 samples of orange fruits and derived products, comprising 77 samples of orange fruits along with 64 samples of orange-based products (juices, pulp, and jam), which were randomly collected from di fferent sites of Punjab, Pakistan. A total of 74 samples were found positive for PAT (52.5%) with a total mean concentration of 6.3 μg/kg, much lower than that of 110.9 μg/kg found in mango. The incidences in orange fruits were 60%, 88.9%, 28.6%, 53.9%, and 18.2% in samples taken from Faisalabad, Sargodha, Layyah, Toba Tek Singh, and Sahiwal, respectively, while 71.4%, 60%, and 21% was assessed in orange juices, pulp, and jams, respectively. The average concentration of PAT in orange fruits was 7.6 μg/kg, 8.1 μg/kg, 8.7 μg/kg, 5.1 μg/kg, and 1.6 μg/kg in samples from Faisalabad, Sargodha, Layyah, Toba Tek Singh, and Sahiwal, respectively. Additionally, orange juices, pulp, and jams contained 8.3 μg/kg, 6.5 μg/kg, and 1.1 μg/kg, respectively. Both incidence and levels of PAT were similar in orange fruits and derived products.


**Table 3.** Incidence of patulin (μg/kg) in orange fruits and derived products.

The percentage of mango samples that exceeded the maximum level of PAT (50 μg/kg) was 29 out of 133 (21.8%) (Table 2), while only 1 out 141 (0.7%) orange samples surpassed the maximum level (Table 3). The maximum percentage of violative mango samples came from Shorkot (57.1%), a city situated southwest of Faisalabad, while northern Rawalpindi, whose elevation above sea level is 508 m, showed the least noncompliant samples (10%). Similarly, exceeding orange samples only came from Layyah, a city located southwest of Faisalabad.
