Butter oil (ghee) was made in the dairy pilot plant, Alexandria University starting by using the raw cow milk of the herd of the faculty of agriculture, Alexandria University. Palm oil (shortening, melting point 36˚C - 38˚C) was obtained from a local market and manufactured by Wilmar Company, Singapore. All chemicals were consumed analytical grade.Also, standards were utilized including Cholesterol solution, 10 mg/mL in chloroform (Supelco), β-Sitosterol, 100 μg/mL in chloroform (Supelco), and Fatty acid methyl esters standard mixture, It was produced by Sigma-Aldrich Co LLC, Merck KGaA, Darmstadt, Germany.

The observation data were analyzed using IBM SPSS software package version 20.0. (Armonk, NY: IBM Corp). The used tests were F-test (ANOVA) for normally distributed quantitative variables, to compare between more than two groups, and Post Hoc test (Tukey) for pairwise comparisons. Significance of the obtained results was judged at the 5% level.

The saponification value (SV) was 212.4 in butter oil (BO) and 165.6 in palm oil (PO) (Table 1). The saponification value in cow milk fat should be in the range of 218 - 228 mg KOH/g oil (ES154-P7-2005). The palm oil used in this study is shortening with a melting point of 36˚C - 38˚C. In the present study, there were significant differences between the blenders regarding saponification value (P ≤ 0.001), iodine value (P ≤ 0.001), and refractometer (P ≤ 0.001). On the other hand, the peroxide value was zero in all treatments (Table 1). The saponification value of milk fat was much higher than that of palm oil as observed. The SV decreased as the percentage of palm oil in the mixtures of MF and PO increased. The saponification value is inversely proportional to molecular weight [15]. The saponification value is of little use in the detection of common adulterants in milk fat. However, Kumar et al. [16] noted that this method could be used for detecting the presence of mineral oils, such as liquid paraffin, in milk fat.

The Iodine value (IV) of milk fat was 34.48 g Iodine/100g fat as shown in Table 1. The IV of milk fat should be in the range of 30 - 45 g Iodine/100g fat according to ES:154-7:2005, while the IV of palm oil was 55.57 g Iodine/100g oil. The blenders had an IV ranging between 34.48 to 55.57 g Iodine/100g fat, dependent on the milk fat/palm oil ratio, but the IV is still in normal range (ES:154-7:2005) of MF in the mixture containing the addition of PO in the ratio 20%

*MF: Milk fat, PO: Palm oil, Mean with Common letters are not significant (i.e. Means with Different letters are significant).

(42.07). According to these results of iodine value could be used to detect the adulteration of milk fat with palm oil if the PO is higher than 20% in the blend. These results agreed with Gandhi et al. [17] who studied the iodine value for detecting palm olein and sheep body fat in milk fat products, they found that the uncertainty of IV to detect the percentage of palm oil in milk fat was about 20%. Similar observations were observed by [18] [19] [20], as the IV was out of the normal range by adding PO to butter oil by about 40.0% or more.

The refractive index (RI) is simple, rapid, and widely used as a preliminary screening method for knowing the quality of milk fat. The RI of milk fat should be in the range of 1.4524 - 1.4552 at 40˚C [ES:154-7:2005]. Refractive indexes (RI) were 1.4538 for milk fat and 1.4622 for palm oil. The RI increased in blenders of milk fat with palm oil by increasing the level of palm oil, but the RI was still in the normal range of MF by adding up to 20% of PO, so the RI method could be used to detect the adulteration of palm oil with level more than 20% of adding PO in milk fat. Our results disagreed with Sharma and Singhal [21], who mentioned that a 5% - 20% presence of vegetable fat can be detected in milk using this method, depending upon the nature of the adulterant.

Butyro reading of different ratios of milk fat and palm oil mixtures using a digital Butyro-refractometer ranged from 39.10 in 100% MF to 49.13 in 100% PO (Table 1), while the highest value was in MF:PO 20%:80% ratio (49.37) and the lowest value was in MF:PO 100%:0% (39.10). No significant difference was found between the mixtures regarding to Butyro reader, except for MF: PO 20:80%.

The highest value of the refractometer index recorded by the Butyro-refractometer reader was in the 100% MF (1.4555) and the lowest value was in 100% PO (1.4593). Besides, there was no significant difference (p ≥ 0.05) between all the studied mixtures regarding the refractometer index recorded by the Butyro-refractometer reader, so the refractometer index recorded by the Butyro-refractometer could not be used as indicator of replacing milk fat with palm oil.

The fatty acid profiles of MF analyzed by GC, (Table 2 and Figure 1), revealed that the capric acid (C10:0), lauric acid (C12:0), myristic (C14:0) acid, stearic acid (C18:0) and -linolenic acid (ALA; C18:3 n-3) were found at levels of 3.00%, 3.33%, 12.04%, 13.34%, and 1.23%, respectively. While all these corresponding values were not detected in 100% PO. Furthermore, while 100% MF had the highest levels of saturated fatty acids (SFA) (66.27%) compared to 47.36% in 100% PO, pure milk fat has a lower level of palmitic acid (C16) than pure palm oil. On the other hand, pure palm oil had high levels of oleic acid (C18:1 n-9) and linoleic acid (LA; C18:2 n-6) when compared with pure milk fat. Also, unsaturated fatty acids (USFA) and USFA/SFA were at high values in pure palm oil.

Fatty acids are usually analyzed by GC equipped with a flame ionization detector (FID). Fatty acid values in the literature have been reported as a percentage of total fatty acids; thus, individual fatty acid values may vary in different studies [15]. Increasing the levels of palm oil led to gradually decrease of myristic acid and stearic acid (Table 2), while palmatic acid, oleic acid and linoleic acid gradually increased. Furthermore, the levels of capric acid (C10:0), lauric acid (C12:0), myristic acid (C14:0), and stearic acid (C18:0) could be used as an indicator for the detection of the adulteration of milk fat by palm oil. The same results were found by other researchers [16] [22]. In this study, the C14:0 level was used to detect the PO in milk fat as it correlated well with increasing the level of PO and could be used as an indicator of more than 20% of the PO. Due to the variation of breeding, season, and feeding, the low level (10% or lower) of adding PO in milk fat is difficult to detect by GC.

The total content of cholesterol and β-sitosterol in pure milk fat was 252.76 µg/ml and 0.10 µg/ml, respectively, compared to 14.75 µg/ml and 1.73 µg/ml in pure

*MF: Milk fat, PO: Palm oil, C8:0 (Caprylic), C10:0 (Capric), C12:0 (Lauric), C14:0 (Myristic), C16:0 (Palmitic), C18:0 (Stearic), C18:1n9c (Oleic), C18:2n6c (Linoleic) C18:3n3 (Linolenic), SFA: saturated fatty acid, USFA: Unsaturated fatty acids, USFA/SFA: Unsaturated fatty acids/saturated fatty acid.

palm oil, respectively (Table 3). In this method, the lowest level of cholesterol was in pure palm oil, while the lowest β-Sitosterol was in pure milk fat.

The β-sitosterol was found as the major phytosterol in palm oil (3.45 µg/ml) and the same findings were concluded by Abd El-Aziz et al. [19]. They found that cholesterol and β-sitosterol contents were 231.0 g/ml and 8.5 g/ml in milk fat, respectively. As a result, when palm oil was added to the milk fat, it resulted in a progressive increase of β-sitosterol and a gradual dropping of cholesterol when compared to pure milk fat; the changes were proportionate to the addition level. Consequently, Contarini et al. [23] assumed that the determination of β-sitosterol content could be useful to uncover the adulteration of milk fat with palm oil when applied to an admixture containing 5.0% palm oil, based on the evidence for the presence of β-sitosterol content in pure milk fat and the highest limit of β-sitosterol in palm oil.

However, because of the large amounts and great variability of the cholesterol present in milk (204.3 to 382.4 mg/100g), which was reported by other researchers as well, [19] [24] [25], the determination of cholesterol content would be unsuccessful when administrated to a mixture containing a lower concentration of palm oil. Regardless of the cholesterol amount, the higher percentage of cholesterol fraction in milk (252.76 µg/ml) compared with the low level in palm oil (14.75 µg/ml), could be useful in detecting palm oil in milk. According to these results and the higher limit of cholesterol percentage reported by Contarini et al. [23] and Borkovcová et al. [14], the addition of 5.0% palm oil was sufficient to depress the percentage of cholesterol fraction to the lower limit even if the percentage was high initially.

*MF: Milk fat, PO: Palm oil.

As observed in Table 4 and Figure 2, different wavelengths and different levels of percentage of absorbance of each functional group were recorded. These functional groups exhibit the whole composition of milk fat, as well as the bonding behavior and absorption intensity of these functional groups. The primary ingredients in milk are recognized using fingerprint traits such as absorbance peak intensities, locations, wave numbers, and forms of the absorption peaks of the original FTIR spectra.

When comparing the mixed samples to a pure milk fat sample, the spectrum behavior produced in this adulterated sample is almost the same, but the wave numbers are somewhat different. The absorbing and relative intensities of wavenumbers differ slightly and the detectable variations could be recorded regarding CH₃ bending (86% in 100% MF and 92% in 100% PO groups), PO stretching (symmetric) of >PO₂ Polyphosphate bending (79% in 100 MF and 84% in 100% PO groups), phospholipid and methylene group CH2 (90% in 100% MF and 96% in 100% PO groups).

The wave number positions of absorbance peaks, peak intensities, and peak widths are specific to the functional groups of the sample; thus, each sample has a unique “fingerprint” absorbance spectrum [26]. Because of the complexity of milk, which can result in overlapping peaks of several elements, the primary ingredients in milk are recognized using fingerprint traits such as absorbance peak intensities, locations, wave numbers, and forms of the absorption peaks of the original FTIR spectra.

Our findings were in accordance with Rohman et al. [27] who carried out a general investigation through the spectra and showed that the spectra were very similar because the major component in fat and oil is fatty acids. Further investigation revealed some differences in several peaks, especially in the fingerprint region, Windarsih and Irnawati [28], suggested that deep investigation could differentiate between milk fat and palm oil spectra. However, in adulterated samples of milk fat with PO, it is very difficult to differentiate between authentic

*MF: Milk fat, PO: Palm oil.

milk fat spectra and adulterated milk fat spectra. So, FTIR spectra could not be used to detect adulterated milk by replacing milk fat with palm oil, but they could be used to detect some other additives. He et al. [26] suggested that some differences can be seen clearly between the milk and the adulterated samples. For example, there are two distinct auto-peaks at 2924 and 2840 cm⁻¹ in the synchronous spectrum of the milk, while in the adulterant samples, including urea, glucose, and melamine, there are more than two auto-peaks between 3000 and 2700 cm⁻¹, and also, their intensities and positions vary from that of the raw milk.