Potato flour (homemade), wheat flour, sugar, and egg yolk were purchased from Wal-Mart supermarket (Shanghai, China). RPS (50 ºC) and RPOL (24 ºC) were donated by Palm Oil Research and Technical Service Institute of Malaysian Palm Oil Board (PORTSIM) (Shanghai, China).

The pre-optimized recipe for potato cookies was used: potato flour 200 g, wheat flour 300 g, fats (RPS-RPOL) 300 g, sugar 200 g, egg yolk 150 g. Melted RPS (heated at 55 ºC) and RPOL were well mixed, followed by the addition of sugar and egg. The mixture of potato flour and wheat flour was added last, and the dough was kneaded and sheeted to a uniform thickness of 5 mm. Then, the dough was cooled at 4 °C for 20 min and cut into the size of 30 mm × 30 mm. Baking was carried out in a SM-522 baking oven (Xinmai machinery Co., Ltd., Wuxi, China). After that, the cookies were cooled at room temperature, then sealed and stored at -20 ºC.

The fats in the formula were blended with RPS and RPOL at three different ratios (RPS-RPOL: 100:0, 83:17, and 65:35) (the dough becomes difficult to form and oil flows out when a larger proportion of RPOL is added, hence the ratio of 65:35 was chosen as the maximum). The three levels of baking temperature (160, 180, and 200 ºC) and three levels of baking time (10, 12, and 15 min) were used. Under these usual baking conditions, the cookies were fully cooked but not burnt.

In order to simplify the sensory analysis, the L₉ (3³) orthogonal experimental design was adopted. The factors and levels are shown in Table 1. The 9 products were put into randomly numbered containers. Fifteen semi-trained panel members (all from the Food Science Department, Shanghai University, Shanghai, China) evaluated the color, aroma, taste, texture, and overall acceptability of the products through a 7-point hedonic scale, from 1 (strongly dislike) to 7 (strongly like) (Harianti et al., 2018).

100 g of cookies were finely ground before extraction and the lipids were extracted by 500 mL petroleum ether in a DK-S12 electric-heated thermostatic water bath (Hualian Medical Equipment Co., Ltd., Shanghai, China) at 65 °C for 20 min. The micronutrients were stored in a deep freezer for further analysis after the rotary evaporation of the solvent at 35 °C.

A β-Carotene analysis was performed on UV-1800PC Ultraviolet spectrophotometer (Mapada Instrument Co., Shanghai, China) (Pan et al., 2016). 0.1 g of cookies was accurately weighed, the volume was made constant with n-hexane to 25 mL, and then placed in a colorimetric dish. The absorbance of the sample was measured by UV-1800PC at 446 nm wavelength with a blank as control. The calculation method of carotene was as follows:

Where E represents the sample absorbance, I represents the width of the colorimetric dish, C represents the content of red palm oil in the sample (g/100mL), Co represents the content of β-carotene in red palm oil (mg/kg).

Squalene was analyzed by a LC-20A High Performance Liquid Chromatography (HPLC) coupled with an SPD-M20A UV detector (Shimadzu Co., Japan) (Pan et al., 2016). The unsaponifiables were extracted from 3 g oil samples, then dissolved and diluted to 25 mL with n-hexane. The extracts were filtrated through a 0.22 μm membrane filter and chromatograms were acquired at 204 nm using an injection volume of 25 µL. An Inertsustain C18 column (250 mm × 4 mm, 5 μm; Shimadzu Co., Japan) was equilibrated using acetonitrile/methanol 4:0 (v/v) at a flow rate of 1 mL/min. Squalene was identified and quantified by comparison to the corresponding standards (Aladdin Industrial Co., Shanghai, China) as external standards.

Tocotrienols and tocopherols were analyzed according to a modified procedure of Yui et al. (2016) using the external standard (α-, β-, γ-, and δ-isoforms, Solarbio Co., Beijing, China) method on an LC-20A HPLC equipped with a RF-10AXL Fluorescence detector (Shimadzu Co., Japan). 2 g of oil sample were dissolved in methanol and sonicated for 20 min, centrifuged at 500×g for 10 min. Excitation and emission absorbance were set at 290 nm and 330 nm, respectively. Separation was done on an Inertsustain C18 column (250 mm × 4 mm, 5 μm; Shimadzu Co., Japan) thermoset at 30 °C. The mobile phase comprised a mixture of methanol-water (98:2, v/v) at a flow rate of 1 ml/min (injection volume = 5 μL).

The Schaal oven test was used for sample pre-treatment to accelerate oxidation (Michotte et al., 2011). The samples were placed in the electro-thermal blower drying box (63 ± 1°C) for 40 days to measure the β-carotene content at 3-day intervals. The determination of β-carotene was performed according to the method of Pan et al., (2016).

The data obtained from various experiments were statistically analyzed. All assays were carried out in duplicate or triplicate and the data are presented as mean ± standard deviation (SD). Analysis of variance (ANOVA) was carried out using IBM SPSS 22.0, followed by Duncan’s multiple range test (P < 0.05).

The effects of different RPS-RPOL ratios, baking times and baking temperatures on the color, aroma, taste, texture and overall acceptability of cookies were studied by L9 (3³) orthogonal tests (Table 1), and the results of orthogonal tests are shown in Table 2. According to the R values, the order of importance of the variables on the color, aroma, taste and overall acceptability was as follows: baking temperature (B) > baking time (C) > RPS-RPOL ratio (A), and the optimal condition was A3B1C1 or A3B1C2; that is, the optimal conditions were RPS-RPOL 65:35, baking temperature 160 ºC, baking time 10 min or 12 min. However, the factors influencing the cookie texture were as follows: baking temperature (B) > RPS-RPOL ratio (A) > baking time (C) and the optimal condition was A1B2C3. Different from other sensory qualities, the RPS-RPOL ratio had great influence on the texture of cookies. The optimal RPS-RPOL ratio was 100:0, which means that the high content in RPS is beneficial to the crisp texture of cookies. In general, considering the influence of all sensory properties, the most significant factor affecting the quality of cookies was baking temperature (B), followed by baking time (C), and then RPS-RPOL ratio (A). The better sensory evaluation scores for the cookies were obtained when the baking condition was A3B1C2 (RPS-RPOL ratio 65:35, baking temperature 160 ºC, and baking time 12 min). These conditions caused evenly-colored cookies, with no incomplete cooking, no charring, and a crisp taste.

The contents of β-carotene in RPOL and RPS were 454.735 ± 1.986 mg/kg and 224.401 ± 1.704 mg/kg, respectively. The content in β-carotene in RPOL was 15 times higher than that in carrot and 300 times higher than that in tomato (Radhika et al., 2017). The effects of different RPS-RPOL proportions, baking temperatures and baking times on the content in β-carotene in oil are presented in Table 3. There was a significant effect of different RPS-RPOL ratios and baking conditions on β-carotene content (mg/kg) in the oil of potato cookies. The content in β-carotene in the oil of potato cookies significantly decreased as the baking time and baking temperature increased (p < 0.05). Without adding RPOL, β-carotene was retained the most at 160 ºC for 10 min (81.7% of the oil), and the least at 200 ºC for 15 min (73.9% of the oil). Siti et al. (2018) conducted a deep-frying experiment at 180 ºC, and after 10 times of repeated frying, the contents in β-carotene in red palm oil decreased from 294 to 143 mg/kg, and from 119 to 82 mg/kg in yellow palm oil. It can be seen that the content in β-carotene in red palm oil after thermal processing is still higher than that in ordinary yellow palm oil.

Added red palm oil to cookies can effectively increase the β-carotene content in cookies. El-Hadad et al. (2010) used 40% shortening and 60% RPOL, and the β-carotene content in cookie samples was 188 mg/kg, 14.8 times higher than in 100% shortening cookies. Researchers have used red palm oil as a substitute fat for a variety of functional food processing, in which the β-carotene can be retained better in red palm oil; the retention in cookies was better than that in bread (Sidhu et al., 2004; Marjan et al., 2016). Thus, the consumption of red palm oil can supplement β-carotene. It was reported that the long-term intake of RPOL as a source of carotenoids, can improve regional population problems (Sobhana et al., 2019), improve the biological characteristics of animal plasma glucose, and cholesterol, and prevent diabetes and cardiovascular diseases. Red palm olein improve the levels of catalase and glutathione peroxidase in red blood cells, which is beneficial to antioxidation (Szulczewska-Remi et al., 2019). In addition, the combined effects of carotene, tocopherols, tocotrienols and other natural antioxidants in red palm olein cannot be underestimated (Yanishlieva et al., 1998).

The squalene content in RPOL and RPS were 185.7 mg/kg and 233.3 mg/kg, respectively. The squalene content in RPS was higher than in RPOL, which is the opposite of β-carotene. Kumar and Krishna (2014) carried out the dry fractionation of crude red palm olein, and obtained the squalene contents of crude red palm olein and crude red palm stearin as 360 mg/kg and 16.4 mg/kg, which were inconsistent with the results of this paper, and may be related to the special processing technology of the oil. Another research showed that the processing method of palm oil can affect the micronutrient content in the oil (Mayamol et al., 2007).

There is a significant (p < 0.05) influence of different RPS-RPOL ratios, baking times and baking temperatures on squalene content (mg/kg) in the oil of potato cookies (Table 4). The squalene content (mg/kg) in the oil of potato cookies was found to vary significantly under different baking conditions and showed a decreasing trend (p < 0.05). The content in squalene descended with the elevation of baking temperature and extended time and the decreased trend observed in the amount of squalene under different baking conditions was the same as that of β-carotene. Without the addition of RPOL, most squalene was retained at 160 ºC for 10 min, accounting for 95.7% of the oil content. At 200 ºC for 15 min, the retention rate of squalene was the lowest at 58.7%. With the addition of RPOL, the amount of squalene in the oil decreased accordingly. Applying red palm oil as a squalene source may increase the amount of squalene for humans as squalene in food can be well absorbed (the absorption amount reaches 60-85%). Therefore, food with red palm oil as the source of squalene is expected to improve the content in human squalene after intake. Squalene has strong antioxidant effects, which can quench free radical oxygen, prevent lipid peroxidation, promote anti-aging and improve immunity. It can also protect normal cells, inhibit the growth of tumor cells and reduce the risk of breast cancer, colon cancer, pancreatic cancer and other cancers. Squalene can promote metabolism of the body, and protect the heart, liver, kidney and other organs (Reddy and Couvreur, 2009).

The data pertaining to tocotrienol and tocopherol contents (ppm) in the oil of potato cookies is presented in Table 4. The total amounts of tocopherols and tocotrienols in RPOL and RPS were 664.3 mg/kg and 254.7 mg/kg, respectively. In addition, different RPS-RPOL ratios and different baking conditions showed significant differences in tocotrienol and tocopherol contents (ppm) in the oil of potato cookies (p < 0.05). The content in tocotrienols was higher than that of tocopherols, and the β/γ- isomer was higher than of other isomers. The most serious losses were presented in β/γ-tocotrienol, in agreement with the report for baking, deep-frying, and other thermal processing (Rossi et al., 2007).

The retention of tocotrienols and tocopherols was under the influence of process parameters. It was obvious that the δ-, β/γ-tocotrienol, δ-, β/γ-tocopherol and total tocotrienol, tocopherol tendencies resembled β-carotene. The loss in vitamin E was minimum at a low temperature for a short time (160 ºC, 10 min). The thermal degradation of vitamin E in oils may have been aggravated by the increases in temperature and time, as clarified by Hamid et al. (2014).

What’s more, the results from sample baking under the same conditions but with different RPS-RPOL proportions showed that the content in vitamin E in cookies could be increased by adding RPOL, which contains higher vitamin E content. RPOL and RPS are not only rich in natural antioxidant vitamin E, but have also been applied in the production of sugar-snap cookies and bread with good sensory characteristics and consumer acceptability (Al-Saqer et al., 2004). The presence of tocotrienols and tocopherols retards oil autoxidation greatly and protects fatty foods from off-flavors. In addition, these compounds possess gene regulatory functions (α-tocopherol), anti-inflammatory, antitumor activities (γ-tocopherol), neuroprotective properties, preventive effect on cholesterol biosynthesis, and anticancer effects (Kamaleldin and Appelqvist, 1996). In short, considerable retention of tocotrienols and tocopherols in food is beneficial to both food storage and human health.

To characterize the oxidative stability of the products, the change in trends of β-carotene content during the oven test are presented in Figure 1. Generally, combined with the ordinates of Figure 1a, 1b, 1c, samples with a higher ratio of RPOL showed superior properties in the oven test since the β-carotene concentration in RPOL was higher than that in RPS. Besides, the β-carotene contents increased, then decreased and became steady at last with the prolongation of oxidation time, indicating that the oxidation of oil proceeded continuously. The concentration of β-carotene increased from day 0 to day 3, decreased slowly from day 3 to day 12, decreased rapidly from day 12 to day 20, and remained unchanged after day 33. Lee (1986) reported a slight increase followed by a decrease in β-carotene content during the storage of fresh carrots. This phenomenon may relate to processing methods, the influence of lipoproteins on the separation of carotenoids, and the distribution of different carotenoids during storage. However, it is difficult to ascribe the reason for the increase in β-carotene to only one of them. Research showed that carotene was almost completely lost under high-temperature accelerated oxidation, but tended to be stable in this study because the egg lecithin in cookie formula had a protective effect on carotene at 60 ºC (Yanishlieva et al., 1998).

When RPS-RPOL was 100:0 (Figure 1a), the β-carotene content in the samples at 200 ºC for 15 min decreased rapidly from day 3, and it was significantly lower than that of the samples under other conditions and after dozens of days. Such a phenomenon revealed that the samples under this condition (200 ºC, 15 min) were the most unstable. In addition, as shown in Figures 1b and 1c (RPS-RPOL were 83:17, 65:35), β-carotene concentration in samples baked at 200 ºC (10 min, 12 min, 15 min) and 180 ºC (15 min) were significantly lower than that of other samples, however, there was no visible difference in the samples at 160 ºC. The results illustrated that temperature was the dominant factor that affected oxidative stability negatively at temperatures above 200 ºC. In addition, long baking time played a more decisive role when temperature was less than or equal to 180 ºC.