Convolvulus arvensis (CA) leaves were collected from the gardens of the Faculty of Agriculture, Cairo University, Egypt. Convolvulus arvensis voucher number (000194 WM 194. 09-,05-,02 - 09) was placed in the herbarium of Orman Botanic Garden, Giza, Egypt. Antioxidant-free sunflower oil was kindly supplied by Arma company, 10^th of Ramadan, Egypt.

Folin-Ciocalteu reagent, Quercetin, P-anisidine, Isooctane, Gallic acid, methyl alcohol, silica gel and TBHQ (Tert-Butylhydroquinone) were purchased from Sigma-Aldrich Co., St. Louis, USA.

The plant leaves were dried at 45 ^oC, milled up to 50 mesh and stored at room temperature. The extraction of polyphenols was performed according to Elsayed et al. (2020) with a slight modification. Aqueous ethanol (80%) at solid-to-​solvent ratios of 1/10, 1/20 and 1/30 (w/v) were used to extract the phenolic compounds using a benchtop lab stirrer (Heidolph, Germany) at maximum speed. Extraction was conducted for 20, 30, 60, 90 and 120 min. The optimum extract with the highest total polyphenols yield was concentrated under vacuum at 45 ^oC using the EYELA Rotary Evaporator (Tokyo Rikakikai Co., LTD, Japan). The concentrated phenolic extract was divided into two parts. One of them was lyophilized (Edward freeze dryer, England) and the other one was microencapsulated.

Microencapsulation of the optimal CA leaf extract was prepared using the mini spray drier BUCHI B-290 (BÜCHI, Flawil, Switzerland) according to the method described by Sukri et al. (2021). Firstly, equal weights of gum Arabic and maltodextrin were dissolved in deionized water with the assistance of vigorous mixing and then the optimal extract was added. When the process was finished the extract represented 20% of the total solid (20g extract/ 80 g encapsulating material). The preceding mixture was magnetically stirred for 5 min before sonication for 20 min using UP200S ultrasound homogenizer (IKA Hielscher GmbH, Berlin, Germany) at 200 W. The spray dryer was operated under the following conditions: mixture flow rate = 8 ml/min, inlet air temperature = 130 ^oC, outlet temperature = 80 ^oC and drying air flow rate = 85% of the suction fan controller.

The total phenolic content of the plant extracts was determined by Folin-Ciocalteu reagent according to the Elsayed et al. (2020) method using UNICO UV-Visible spectrophotometer (UNICO Instruments Co., LTD, U.S.A.) at 750 nm. The results were expressed as mg Gallic acid equivalent per gram plant dry weight (mg GAE/g DW).

The total flavonoid content was determined using the aluminum chloride colorimetric method at 510 nm according to Formagio et al. (2014). The results were expressed as milligrams Quercetin equivalent/gram of dry weight (mg QE/g DW).

An agilent1260 infinity HPLC Series (Agilent, USA) equipped with a Quaternary pump and a Kineted® 5µm EVO C₁₈ column 100 mm x 4.6 mm, (Phenomenex, USA) were used to identify phenolic and flavonoid compounds. The separation was achieved using a ternary linear elution gradient with (A) HPLC grade water 0.2% H₃PO₄ (v/v), (B) methanol and (C) acetonitrile at a flow rate of 0.7 mm/min. The detection wavelength was set at 284 nm. The injection volume was 20 μl and the column temperature was held at 30 °C. The compounds were identified by comparing their retention times with those of authentic standards. Calibration curves were used to calculate the concentrations of the compounds.

According to the permissible concentration of TBHQ in vegetable oils (CODEX, 2019 CXS 210-1999) and previous studies (Wang et al., 2020), lyophilized and microencapsulated leaf extracts were separately added to sunflower oil at 120 mg gallic acid equivalent/kg oil and TBHQ was added at the same level (120 ppm). Sunflower oil without added antioxidants was used as a control. Frying was performed in a deep-frying pan. The surface/oil volume ratio was 0.28 cm^-1. One kilogram of oil was heated to 180 ± 10 ^oC. Fifty g of potato chips were fried for five min every 20 min, which was consecutively repeated for 8 hours per day. This frying process was repeated for four consecutive days with the same oil without replenishment. A 50 ml sample of the oil was collected both in the fresh state and at the end of every frying day. Oil samples were stored at -18 ^oC for analysis. All the frying processes were done twice.

The IR absorption spectra in potassium bromide disks of oil sample were taken in the region (4000- 400 Cm^-1). A FTIR spectrometer (NICOLET 380 FT-IR, Thermo Scientific, made in China) was used. Data were manipulated by a computer equipped with Professional Windows XP Software.

The data are presented as mean values ± SD. The data were subjected to analysis of variance (ANOVA) followed by Tukey’s test using significance level of 0.05 (XLSTAT, Addinsoft, USA).

The yield of phenolic compounds was proven to be highly dependent on extraction process parameters. Thus, the effect of plant/solvent ratio and extraction time on the yield of phenolic compounds was investigated. Results in Figure 1(a) illustrate that decreasing plant/solvent ratio caused a significant (p < 0.05) increase in the yield of total polyphenol contents (TPC) at various extraction times. Extending extraction time from 20 to 90 min at different plant/solvent ratios (except plant/solvent ratio of 1/10) caused a significant (p < 0.05) increase in the yield of TPC, whereas further increase in extraction time had an insignificant (p > 0.05) effect on the yield of TPC. The highest TPC yield was 53.15 ±1.58 mg GAE/g DW, which was obtained at a plant/solvent ratio of 1/30 and extraction time of 90 min. These results agree with those reported by Ćujić et al. (2016). They reported that decreasing plant/solvent ratio from 1/10 to 1/80 caused a significant increase in the TPC from 53 to 83 mg GAE/g DW, respectively. They also reported that increasing the extraction time caused an increase in TPC. In a similar trend, Chirinos et al. (2007) reported that the TPC of mashua tubers (Tropaeolum tuberosum) was increased by increasing extraction time from 0 to 60 min, and a further increase in extraction time from 60 to 120 min had no significant effect on the yield of TPC.

The total flavonoid content (TFC) of CA leaf ethanolic extracts obtained under various extraction conditions are presented in Figure 1(b). What stands out in Figure 1(b) is that the highest yields of TFC were achieved at a plant/solvent ratio of 1/20 for various extraction times. Furthermore, at this plant/solvent ratio (1/20) extending the extraction time to higher than 60 min had no significant (p > 0.05) effect on the yield of TFC. In contrast to the TPC extraction pattern, decreasing plant/solvent ratio to 1/30 significantly (p < 0.05) reduced the yield of TFC at various extraction times. Moreover, it could be noted that for all plant/solvent ratios, except (1/20), extending extraction time from 20 to 120 min significantly (p < 0.05) increased the yield of TFC. The highest TFC yield was 18.72 ± 0.12 mg QE/g DW, which was obtained at a plant/solvent ratio of 1/20 and extraction time of 120 min. Similar results were reported by Chirinos et al. (2007), who studied the effect of different extraction times (from zero to 140 minutes) on the yield of TFC. They reported that the TFC yield increased significantly when the extraction time was increased from zero to 60 min, and then it became stable during additional extraction times.

The results indicated that the TPC of the LyCAE was 55.40 ± 0.43 mg GAE/g; whereas it was 9.30 ± 0.22 mg GAE/g for the EnCAE.

The phenolic compounds in the optimal ethanolic extract with the highest yield of polyphenols from CA leaves are listed in Table 1. The predominant phenolic compounds were benzoic acid and rosmarinic acid (164.39 and 134.58 µg/g DW, respectively), while quercetin and rutin (173.25 and 155.55 µg/g DW, respectively) were dominant flavonoids.

Elzaawely and Tawata (2012) showed that benzoic acid, syringic acid, p-hydroxy benzoic acid, vanillin and ferulic acid were the major identified phenolic acids in the CA leaf acidic ethyl acetate extract. This difference in the phenolic composition of the extracts could be attributed to plant species and solvent type.

The FTIR spectra of fresh oil and various frying oil samples at the end of the frying process were measured in the range of 4000 - 400 cm^-1, as shown in Figure 4. IR spectral bands shown at 3470, 1164, and 968 cm^-1 are assigned to a hydroperoxide group, a carbonyl group, and trans double bonds, respectively, (Guillen and Cabo, 1997 ; Hammad et al., 2021), which were used to assess the effect of the frying process on the quality of oil at the end of the frying process. Considering spectral band of 3470, the control sample showed the highest absorbance value at the end of the frying process; whereas samples containing various antioxidants showed absorbance values closed to those of fresh oil (before frying process).

In addition, control samples at the end of frying process showed the highest absorbance value at 1164 cm^-1, which indicates the increase in the formation of the aldehyde functional group in the absence of various antioxidants. The thermal oxidation of oil generates hydroxyl groups which are represented in peaks in the range of 3000 cm⁻¹to 3600 cm⁻¹, especially after multiple frying cycles. Ahmad et al. (2022) reported that the C-O and carbonyl (C = O) bonds of aliphatic esters are responsible for the spectrum peaks near 1150 cm^-1 and 1750 cm^-1, respectively. On the other hand, frying oil containing EnCAE exhibited the lowest absorbance value followed by LyCAE and TBHQ in the same order. At IR spectral band of 968 cm^-1, frying oil containing EnCAE exhibited the lowest absorbance value followed by TBHQ and LyCAE in the same order, which indicates the inhibiting effect of these antioxidants on the formation of trans double bonds. The absorbance of oils at 968 cm^-1 reflects the trans double bond C-H (Guillen and Cabo, 1997). Moreover, multiple frying at various intervals may make oxygen more availableonce the frying oil cools, which could lead to similar peaks in the edible oils (Ahmad et al., 2022).