*5.1. Sampling*

A total of 274 samples including 70 mango fruits, 40 mango juice, 8 mango pulp, 15 mango jam, 77 orange fruits, 35 orange juice, 10 orange pulp, and 19 orange jam were purchased in di fferent areas of Central Punjab (Pakistan) from January to December 2018. The mango (*Mangifera indica*) and orange (*Citrus sinensis*) fruits were collected in supermarkets and local markets from Faisalabad, and from cities north (Rawalpindi, Sargodha, and Sheikhupura) or south (Multan, Shorkot, Layyah, Toba Tek Singh, and Sahiwal) Faisalabad. The mango- and orange-based products (juice, pulp, and jam) were from commercial brands of nationwide distribution purchased from supermarkets, local markets, and general stores in Faisalabad district. All samples were stored in their original packages at 4 ºC until they were analyzed. The samples were opened and thoroughly homogenized.

#### *5.2. Chemicals and Reagents*

Acetonitrile (HPLC grade), glacial acetic acid, sodium chloride, and sodium carbonate were purchased from Merck. Ethyl acetate, methanol, sodium sulfate anhydrous (analytical grade), 5-hydroximethyl furfural (5-HMF), and patulin (5 mg of solid crystalline) were purchased from

Sigma-Aldrich (Saint-Louis, MO, USA). Stock solution of patulin 1 mg /mL was prepared in acetonitrile and stored at −4 ◦C. Necessary volumes of stock solution were taken for working solutions (5, 10, 30, 50, 70, 100 μg/L) in 0.1% acetic acid. PriboFast multifunctional cleanup columns MFC 228 (Pribolab Pte. Ltd., Singapore) were applied for the optimization of the purification process.

#### *5.3. Sample Preparation and Extraction*

The purchased fruit samples were chosen free from debris, washed with water, shade-dried, and cut into small pieces with sharp knife. A sample of about 500 g was homogenized using high-speed blender (Braun Blender Mix 2000, Marktheidenfel, Germany). Samples of products (juice, pulp, and jam) were taken directly from the original packages. Analytical method for PAT was based on AOAC method 995.10 with little modifications [31]. Homogenized samples (25 g or 25 mL in triplicate) were extracted twice with 50 mL ethyl acetate along with 2 g sodium chloride in 250 mL Erlenmeyer flask (Pyrex, Germany) by shaking at high speed in horizontal shaker (Gunther and Co, Bremen, Germany) for one hour. The upper organic layers were combined and cleaned up with sodium carbonate solution (1.5% Na2CO3). Cleaned extract was rapidly dehydrated with 10 g anhydrous sodium sulfate (Na2SO4) and filtered with Whatman No. 1 filter paper. For alternative second-time clean-up, 9 mL cleaned extract was passed through a multifunctional column MFC 228, and 4 mL purified extract was collected. The extracts were evaporated to dryness under a stream of nitrogen. The dry residue was immediately dissolved in 1000 μL of 0.1% acetic acid and passed through syringe filter (0.22 μm, Millipore, Darmstadt, Germany). The samples were analyzed for PAT by reverse-phase HPLC equipped with UV-Vis detector (SPD-10AS, Shimadzu, Japan) in isocratic mode. The injection volume was 20 μL, and total run time was 10 min for PAT analysis.

#### *5.4. Apparatus and HPLC Conditions*

HPLC comprised a delivery pump (LS-10AS), system controller (SCL-10A), column oven (CTO-10A), UV-Vis detector (SPD-10AS), and Communication Bus Module (CBM-101), Shimadzu, Japan. Separations were performed in a Discovery HS C18 silica-based column (250 × 4.6 mm, 5 μm particle size; Supelco, Bellefonte, PA, USA), maintained at 30 ◦C with a flow rate of 1.5 mL/min in isocratic mode. Mobile phase was a mixture of acetonitrile:water (10:90, *v*/*v*) filtrated with 0.45μm filter (Nylon 66 membranes filter, Supelco, Bellefonte, PA, USA). Cleaned sample extracts and standards were injected using a Rheodyne injector (20 μL loop) in reverse-phase system, and wavelength of detection was 276 nm. Chromatograms were received with CLASS LC-10 Acquisition software. Qualitative and quantitative determinations of PAT were made comparing the retention time and peak area of reference standards. For method validation, a test was performed to confirm the peak separation of PAT and its principal interference 5-HMF. A solution containing 2 μg/mL of 5-HMF and 2 μg/mL of PAT was prepared and injected into HPLC (Figure 3).

**Figure 3.** Chromatogram of 5-HMF (5-hydroximethyl furfural) and patulin standards.
