5 CI-IAP'I'LR II LITERATURE REVIEW 2.0 DETERMINATION OF OIL QUALITY Hie concept of halal products or foods is currently gaining attention globally because of its recognition as an alternative benchmark for safety, hygiene and quality assurance of what people consume daily (Ambali and Bakar, 2014) . Usually, the ingredient label does not list the origin of the ingredients. Muslims require some protection to ensure that information provided on food labels are true and accurate (Eliasi. 2002). Hidden ingredients from various sources present another serious problem for Muslim consumers (Riaz and (laudry, 2003). Therefore. the increasing demand for clarity in the food industry has enhanced the development of methods for the analysis of food ingredients. One of' the main ingredients in food products or prcparatioils of' cooking ingredients arc fats and oil. According to MAliilah et A. (201 I ), durina the first 50 years of the twcºltºcth century, the use of animal fats in flood was very common. For instance, land (pig fat) was the most used products for domestic Crying as well as raw material in mass production of breads and cakes. Moreover, in fixuf industry, it still serves as an important ingredient in the iörniulation ol'sonic hood products. 6 In determining the quality of oil, the particular oil must be resistance from oxidation. "The resistance of oil to oxidize is known as the oxidative stability (Guillen and Cabo. 2002). It can be expressed as the period of time necessary to attain the critical point of oxidation in terms of Sensorial change or a sudden acceleration of the oxidative process (Silva et al.. 2001). Rancidity has always been associated with changes that lead to the undesirable ilaý our and odor of lipid. Generally, there are two types of rancidity which are oxidative rancidity and hydrolytic rancidity. Both of' the rancidity resulted in reduction of' the oil quality, but diflerent in the mechanism. Since oxidation produces low ºllolecular weieht Off, flavour compounds, thus it is an important indicator of' oil quality and shelf life (Hamilton, 1994). Oxidation of oil also destroys essential fatty acids and produces toxic compounds and oxidized polymers. Oxidation of edible oils is influenced by the reaction energy, fatty acid composition. types of oxygen, and minor compounds such as nmetals, pigments, phospliolipids, free fatty acid (FFA), mono- and diacylglycerols (r\I: AG I)A(I), thermally oxidized compounds and antioxidants (('hoe ct at.. 2005). 2.1 I: ýC 1'OIlti : ýFFE("I'I\'C Tlil: O\IU: ý"I'IO\ OF LDIBLFOIl. ?. l. l Fatty Acid (FA) Comlxositions oi, 011s. The result from Yun and Surh (2012) research shovyed that fatty acid compositions can he the predictor for initial oxidation state or quality of the vegetable oils. I lovvever, the fatty acid composition might not he an appreciable contribution in dctcrminint" the oxidativ'c stability of the ve etable oil when the oil oxidation could be 7 fäcilitated (ic; fry at high temperature or storing for long time after opening the oil's bottle). Unsaturated oil will oxidize more quickly than less unsaturated oil. Soybean, safflower or sunflower oil (iodine values more than 130) stored in the dark had a significantly (P < 0.05) shorter induction period than coconut or palm kernel oil whose iodine value is less than 20 (Tan et al., 2002). Meanwhile, high oleic and high stearic oils from gene silencing of the oilseeds or hydrogenated soybean oil had higher autoxidative stability (Liu et al., 2002). 2.1.2 Storage Temperature and Light Ultraviolet (UV) light as well as visible (Vis) light may accelerate autoxidation process by triggering the hydrogen abstraction that will produce the formation of alkyl radicals (Chug and Min, 20061). For example in sweet fennel oil. leans- anethole had completely oxidised to anisaldelryde after 2 months of storage at room temperature under light (Misharina and Polshkov. 2005). It has been reported that the effect of Tight on oil oxidation become less as the storage temperature increases (Velasco and Dobarganes, 2002). The packaging of the oils is very important because the oxidation occurs in the presence of light. 2.13 0ýv, -, cn Roth concentration and type of' oxygen affect the oxidation of oils. Oxygen concentration has a high impact upon the oxidation. The rate of oxidation will be affected by the concentration of dissolved oxygen and oxygen transmission rate (Kanavouras ct A. 2006). 8 At high temperature, the oxygen concentration on the oxidation of oil increased in the presence of light and metals. It is reported that higher oxygen dependence of oil oxidation at high storage temperature is due to low oxygen solubility in the oil at high temperature (Andersson, 1995). 2.2 METHODS OF DETECTION FOR ADULTERANTS 2.2.1 Diffcrcntial Scanning Calorimctry (DSC) Thermal analysis has been used for both qualitative and quantitative analysis ranging from various liclds such as in pharmaceutical, biological science, minerals, metals and foods. The technique includes differential thermal analysis (DTA), dielectric thermal analysis (DFA), thcrmomechanieal analysis (TMA) and Diflcrential Scanning Calorimetry (DSC). 111 the study done by Marina et al. (2009), they monitored the adulteration of virgin Coconut with palm kernel oil (PKO) and soybean oil (S130). They spiked the virgin coconut oil with PKO and S130 from 2 to 40 ° (w w separately. Using I)SC, they managed to determine the illy acid of all oil. The results show that PKO adulterated oil did not show any adulteration peak but demonstrated a , gradual decrease in the peak height of the major exothermic peak. Differential scanning calorimetry also can provides a unique thermal profiling Ar oil as It can he used to detect lard adulteration in virgin coconut oil (\'C O). l'ronl the experiment done by Mansor et al. (2012), there was one major endothermic peak with a smaller shoulder peak that gradually smoothed out to the major peak as the percentage of lard increased Turin`.: heating 01' the thermocranl 01' the mi. \ture. 9 Meanwhile, during the cooling phase, there was one minor peak and two mayor exothermic peaks which increased as percentage of lard increased and another peak which decreased in size as the percentage of lard increased. Moreover, both melting as vv-ell as cooling profile of VCO is important in determining the presence of lard adulteration. Although TAG and FA analysis by the HPLC and GC- FID are able to detect lard adulteration in VCO with high confidence, they could not provide a qualitative analysis and are limited by the use of chemicals and the requirement for highly trained personnel to operate the systems. DSC may hermits to linger print the primary crystallization of triacylglycerols (003s) molecules and their transitional behaviour. Dahimi ct al. (2014) assessed the cross contain ination caused by lard ranging from 0.5 °ö to S ", b concentration in the mixture containing beef tallow and chicken lat. The result obtained showed that the discrimination of lard from beef tallow and chicken It was very obvious even when the concentration was as low as I 2.2.2 Nuclcar \launctic Rrsonanrc (NNIR) Nuclear magnetic resonance is a technique used to determine the compound structure. It is able to identify the carbon hydrogen network in the compound. NNR has many adýantaýaes although it is less sensitive than III'LC, GC and capillary electrophoresis. NNIR is non-destructive, selective and capable 01' simultaneous detection of a `great number of molecular mass components in a complex mixture. Moreover, the sample preparation is simpler and less time consuming. 10 NMR can be used to detect trans Fatty Acid in olive oil. (Sachi et at.. i)9S). The purity index for virgin We oils is according to the absentees of trans li tty acids, meanwhile refined olive and olive pomace oils contain detectable levels of olcyl, linoleyl and linolenyl trans isomers. High resolution nuclear magnetic resonance (NMR) spectroscopy coupled with mass spectroscopy (MS) has been used to characterize food products and detects possible adulteration in fruit juices, wine and olive oil (()grins et at., 2003). There were also research done by Dais et al. (2007) showing that '"P NN, IR Spectroscopy may be used In determining the quality control and authentication of extra virgin olive oil. The method is based on the derivatization of the labile hydrogens of functional groups, of olive oil compound with the phosphorus reagent 2- chloro- 4,4,5,5-tetramethyldioxaphospholime and the use of the P chemical shifts to identify the phosphitylated compounds. Used trying oil is added to into the commercial qualified vegetable oil by the unscrupulous traders to seek higher protit, therefore the authenticity assurance of the commercial oil remains a challenge in terms of its health and commercial perspectives. Qing Alan- et al. (201 ;) have made a research which focused on using the Low Field Nuclear Magnetic Resonance to discriminate the adulteration of commercial corn, peanut, rapeseed, and soybean oil with two kinds of used frying oil. . 2.3 Sensor Electronic Nose Electric nose were designed to mimic the nuunnnaliall clf 1ctorv system vv'ith the advaftaggc 01, repeatable nlcasurement, allowing idcntiticatloll and cla sil cation of 11 aroma mixtures (Davide et al., 1995). The system consists of multi sensor array, an information processing unit, software with digital pattern recognition algorithm as well as reCercnce library database (I latfield ct al., 1994). Hai and \Vang (2006) used electric nose to detect maize oil adulteration in camellia seed oil and sesame oil. However, the PCA result showed that it cannot be used to discriminate the adulteration of camellia seed oil but can be used for adulteration in sesame oil. Moreover, using ANN (artificial neuro network) model, the electric nose cannot predict the percentage of adulteration in camellia seed oil but it is able to do it in sesame oil. In the experiment done by Marina et at. (2010), virgin coconut OR was mixed with refined, bleached and deodorized (R131)) palm kernel olein at different level of adulteration. 'The adulteration peaks were identified using linear regression from the chromatogram prutlle. The electric nose based on the acoustic wave sensor, was used to generate a pattern of volatile compound present in the sample. Principal component analysis (PCA) was used to differentiate between pure and adulterated samples. The PCA provided good differentiation of samples with 74 of the variation accounted by PCI and 17 "o accounted by PC2. Pure samples formed a separate cluster from all of the adulterated samples. ?.?. -} Fluorescencc Spectroscopy Fluorescent spectroscopy has the advantages of having high sensitivity and SCICCII\'11V. 1 hUS, 11 may be used 10 I11\'CS11('111C ICCal 11111C structure uni dynamics in 12 solution as well as under microscopes. It works by measuring, the intensity of photons that being emitted Isom sample after it has absorbed photons. In the paper written by Poulli ct al. (2006), they managed to demonstrate the potential of total synchronous fluorescence (TSyF) spectra to differentiate virgin olive oil from sunflower oil and synchronous fluorescence (SvF) spectra combined with multivariate analysis. 'l'hc spectrum was acquired by varying the excitation wavelength in the region 270 until 720 nm while the wavelength interval was in the region from 20 to 120 nm. As a result, the TSyF contour plots for sunflower in contrast to virgin olive oil, showed fluorescence region in the excitation wavelength which range from 325 until 385 nn]. The potential application of fluorescence spectroscopy in detecting the adulteration of milk fat with vegetable oil and characterizing the samples according to the source of fat has been studied. Ntakatsane ct al. (2013) have adulterated pure butterfat with different vegetable oils at various concentrations which are 0,5,10,15, 20,30 and 40 %. The 2- and 3- dimensional front face fluorescence spectroscopy and gas chromatography were used to obtain the fluorescence spectra and fatty acid profile based on the total concentration of saturated fatty acid and unsaturated fatty acids, and also on the 3 major fluorophores which are tryptophan, tocopherols and riboflavin. The result showed that fluorescence spectroscopy was able to detect up to 5° of adulteration of vegetable oil in the butte lint. 13 12. E Fourier Transform Infrared (FTIR) Spectroscopy FTIR spectroscopy works according to Beer's law which stated that the intensities of' the bands in the spectrum are proportional to the concentration of the corresponding samples (Vlanchos et al., 2006). Over the years in the field of fats and oils, FTIR spectroscopy has received great attention in quantitative analysis because of the 'green analytical chemistry'. The method has been known of the easy sample preparation with reduced or no sample pre- treatment steps (Sherazi et al., 2010). When combined with partial least square (PLS) model, it has been used in the quantitative analysis of lard in mixture with other animal fats (Che Man and Mirghani. 2001), cake formulation (Syahariza et al., 2005) and chocolate products (Che Man ct al., 2000. FTIR has been widely used in researches including vegetable oil. For instance in the authentication of virgin olive oil (Lai et al.. 1995) and extra virgin olive oil (Alain and Hamid, 2007). Rohman et al. (2011) managed to develop a fast technique of FFIR spectroscopy for the detection and quantification of lard in Canola oil, Corn oil, Extra Virgin Olive Oil, Soybean oil, and Sunflower oil. At fingerprints of 1500-1000 cm A they are able to quantiiv and classify the lard in the mixtures with vegetables oil. A minor difference of peak heights between lard and vegetable oils were ObserV'ed at 1117 cm I and 1097 cm corresponding to C-II bending vibration and C-I1 deformation vibrations of fatty acids. Accurding to Rendini et al. (2007), in fats and oils, most of the peaks and shoulders of speCtrlllll indicates specific functional groups. Since the Illaill components of both lard and Vegetable oil are triglycerldes, thus, their spectra look very similar. However. due to the fingerprint technique (no two 14 compounds having the same spectra in terms of amount and intensity of peak) FTIR spectroscopy can be used to extract difference among these oils. Che Man et al. (201 1 a) used FTIR spectroscopy to detect the presence of lard in French fries pre-Cried in palm oil adulterated with lard. The result showed that the spectra of palm oil and lard were different at frequency 3006 cm-' and the frequency region of 1120-1 095em 1. They exhibit a shoulder band at 3006cm 1. It was associated with the stretching vibration of cis olefinic double bonds (Guillen and Cabo, 1997). The results also showed that lard contained linolenic acyl groups twice as much as palm oil which were later reflected in the spectrum for lard because sharper band at frequency 3006cm-1 was observed compared to lard spectrum. Moreover, it was also Ibund that as the ratio of lard in palm oil was increased, the peak area and height will also gradually increased. FTIR spectroscopy can be used to determine the authentication of igella sativa oil when combine With chemomet'ics. Nurrulhidayah et al. (2011 b) prepared the binary mixtures of A'igella satiia oil and grape seed oil in the concentration ranges of 0.5 ° 60 % (v"v). l. venthough both oils FT IR spectra appear quite similar, there were actually, some significant differences either in number of peaks at region 17HOcm-i - 1700 cm1. It was also observed that Nigc'lla saliva oil has two sharp peaks in 1744 cm-i and 1710 cm i while grape seed oil has only one peak at 1744 cm-i. These peak were associated with carbonyl (C=O) stretching vibration (Rohn= mid C he Man. 2009a). 15 2.2.6 Gas Chromatography (GC) Currently, chromatographic techniques such as gas chromatography-mass spectrometry (GS-MS) (Park et al., 2010) gradually become the most important and common techniques to detect adulteration. According to Xie et al. (2013), GC-MS is a common too] for the analysis of separation of fatty acids and sterols. Gas chromatography equipped with flame ionization detector was used in the experiment to detect the adulteration of olive oil with relatively cheap oil such as soybean oil, sunflower oil and canola oil. The iodine value and the refractive index in the two samples of adulterated oil were significantly higher (P < 0.01) when compared to the reference (genuine) olive oil. Using the GC method, it showed that fatty acid (FA) profiles in the two samples exhibited higher amounts of linoleic and linolenic acids but significantly lower amounts for oleic acid (Jafari et al., 2009). Seo et al. (2010) performed a method for the identification and detection of corn oil in adulterated sesame oil. The fatty acids compositions were determine using GC-FID and IRMýIS. The result showed that the content of palmitic, linoleic and linolenic acid increased gradually as the mixture rate of corn oil was increased, while the content of stearic acid and oleic acid decreased. Xie et al. (2013) develop a method of detecting adulteration of camellia seeds oil. They randomly selected camellia seed oil with different compositions of eight soybean oil (particularly the oleic acid content) with different levels of adulteration mixtures at 1.2,3,4,5,10,15,20,35,45 and 50 `%. Later. all samples were esterificd and analyzed using GC-MS. The result showed that oleic acid (('18: 1) and 111101CIC acids (C: 18: 2) were the predominant lütty acids in all oils samples. I ligher oleic acids 16 werc found in camellia seed oil (78.19 %-85.63 "ö) than in soybean oils (23.88 %- 27.69 ", o). As for the linoleic acid, soybean oil contain higher value (50.60 % 52.23 compared to camellia seed oil (6.53 `%- 9.49 %). 2.2.7 H cadspacc 1 ieadspace analysis is normally defined as vapor-phase extraction which involve the partitioning of analyes between a non volatile liquid or solid phase and the vapor phase above the liquid or solid. It is also expected that the vapor phase mixture contains less components than the usual complex liquid or solid sample and this mixture is transferred to a GC (Snow and Slack, 2002). There are several techniques that could be describes as GC- headspace which are static (vapor- phase extraction) and dynamic (purge and trap) headspace. IIS-GC consist of two steps. Firstly, the sample will be placed in a vial that has a gas volume above it, Later this vial will be thermostatted at a constant temperature until they reach the equilibrium. Then, an aliquot of the vial's gas phase will be introduced into the carrier gas stream which will carry it into the column to be analyse. A newer technique such as solid phase microextraction (SPNIF) uses traps that "Al help to separate the volatile analytes from the excess of the diluted headspace gas. The dynamic headspace technique is a continuous method of gas extraction and separated the volatile components li-om the matrix by continuous flow of an inert gas above the sample and it is known as purge and trap (Kolb and Litre. 2006). One of' the main advantages of using hcadspacc method is the speed of' the analv, sis since 110 prior sample preparation steps are required. and the simplicity of the 17 measuring process. When coupling a headspacc sampler to a mass spectrometer (I-IS- NIS), it is able to recognise complex mixture of volatile compounds, without the associated with gas sensors (Zubritsky, 2000). Morales et al. (1994) did a research on determining the volatiles in virgin olive oil by using dynamic headspace gas chromatography. It has the advantage of concentrating sample which will make it possible to detect compounds that are present but at low concentration but contribute significantly to the flavour. Tenax TA was used as adsorbent material, thermal desorption and cryofocusing prior to capillary GC to avoid undesirable peak broadening. The result of gas chromatogram of virgin olive oil showed the presence of 100 components, 56 of which were identified in their work. The volatiles identified corresponded to different chemical lämilies such as 7 hydrocarbons, 9 alcohols, 9 aldehydes. 9 ketones, I acid, 12 esters, and 2 furans. Lorenzo et al. (2002) proposed the use of direct coupling of a hcadspacc sampler to a mass spectrometer for the detection of adulterants in olive oil using the application of the linear discriminate analysis technique. They performed the experiment for three different tasks; non adulterated olive oil, adulterated olive oil, olive oil adulterated with sunflower oil/olive oil adulterated with olive-pomace oil, and non adulterated We oil /olive oil adulterated with sunflower oil/olive Al adulterated with olive-pomace oil. The result showed that this method has been very good Iür the three classification tasks addressed, thus it might be used as a screening method. Pena ct al. (2005) develop an analytical method to detect adulteration of virgin olive oils and olive oils Nv, ith hazelnut oil by a hcadshace autosamhler directly coupled 18 to a mass spectrometer. The end result showed that a minimum adulteration level of 7 `Yo and 15 ',, o can be detected in refined and virgin olive oil respectively. Headspace analyzer could also be used for the identification of pork for halal authentication. Nurjuliana et al. (2011) study the aroma profiling and identify the components that contribute to the flavour of pork by employing gas chromatography mass spectrometer with headspace analyzer. The result showed that there were a total of 43 volatile components of porks identified by the GC-MS-HS. The majority of the compounds are well known lipid oxidation products including ketones, aldehydes as well as alcohols. Whereas the most detected compounds were aldehydes and ketones. The result obtained also showed that the volatile profile of pork contained a higher proportion of heptanal, which correlated with the experiment done by Shahidi (1994) that aldehydes are the major components identified in the volatiles of cooked pork. HS-SPN, 1E-GC-MS could be used to differentiate between the volatile compound of sunflower oil and high oleic sunflower oil. Hexanal, l-2 heptanal, E-2 decenal and E, E-2,4-nonadienal were the most suitable compounds in differentiating the two oil varieties from each other (Peterson, 2012). In the study done by Zhao et al. (2013), they did a preliminary investigation to examine and compare the performance of flavours between three types of pure vegetable oils and two types of adulterated oils using I-IS-GC x GC -TOF 1\MIS. They detected the volatile profile of tlu-ce types of pure vegetables oils and two adulterated oils which were sesame oil and peanut oil adulterated with soybean oils at six different levels of adulterations 5 %, 10 %, 30 '%,, 50 °/0,70 °ö and 90 ')o respectively. They identified 30 common volatile components of the flavour among the three types of vegetable oils such as aldehydes. alcohols, ketones, acids, esters and hydrocarbon. 19 heterocyclic compounds and benzene rings. Meanwhile. in soybean oil, the main volatile compound was E-2-penten-l-ol and E, E-2-4-hexadienal while no particular flavour was obtained Boni peanut oil compared with another two oil samples. 2.3 PROFILING ?. 3. l R131) Palm Oil Profiling The colour of refined, bleached and deodorized (R13D) palm oil is normally very 11-ht vvcllovv-. Ilovvever, during processing, various components can affect the colour stability of the finished materials and can be a major quality characteristic. The temperature Mhere the oil began to smoke indicates how well it can tolerate heating and rchcatin,, T. The higher- the smoke point, the better it is. Typical frying, temperature is about 60 "F (ISO "C). During this temperature. unsaturated oils tend to break down or polymerized quickly. Palm oil with a smoke point of437 °F (225 "C). well above normal frying temperatures, makes it an ideal frying oil. Palm oil has relatively high melting point (Fife. 2007). In the research clone by Sarnia et al. (20Il), they analyzed for different phvsicochenical parameters between teXtUrized RBD palm oil and partially hydrogenated RB1) palm oil. As for the result for texturized RBI) palm oil. it were reported that the moisture and volatile matter was 0.04, the iodine value was 49.61, the acid value was 0.14. the saponiOcation value was 200 and the melting point was 38 C. Using G('-Iv1S, Dirinck et at. (1977), were able to examine the volatile trace COiiStitUents isolated from halm oil. The result showed that trans-2-octcnal, n-nonanal, 20 trans-2-dccenal. trans-2-undcenal, b-ionone, cis-2,4-decadienal and trans 2.4- ciccadicnal as important contributors to palm oil. Nor et al. (2007) did a research on the changes of headspace volatile constituents of palm olefin, soybean oil, corn oil and sunflower oil while frying from 2 hours to 40 hours. They found that in palm olein, the 2t, 4t-decadienal content decreased from 15.9 pg g1 (2 hours) to 3.2 pg g-1 after 40 hours of frying while hexanal increased from 113 pg g-' (2 hours) to 33.8 pg g-I after 40 hours of frying. The low number of hexanal content indicated that palm olein was more stable towards degradation at higher temperature because of the lower content of linoleic acid. Palm olcin was expected to produce lower quantity of decadienal as compared to other liquid oils (Boskou et al., 2006). Petersen et al. (2013) evaluate the volatile compounds that could be the marker for the edible oil deterioration during the production of deep fried French tries. They compare the result between the sensory characteristic and the volatile compounds. 32 hours frying were performed on sunflower oil, rapeseed oil, high oleic rapeseed oil, high oleic sunflower oil and palm olein. They found that after 3 hours of deep frying, F, F- 2,4-decadienal and heptanal showed the ability to differentiate between sample of various oxidative state, while E, E, 2,4-heptadienal and F, 2- dccenal showed a positive correlation with well known lipid oxidation parameters. ; Alircza et al. (2010) investigate the effect of frying media and storage time on fatty acids compositions and iodine value (1V) of deep fat fried potato chips. All experiments were conducted at 180 °C for 5 days. The result for fatty acids analysis showed a decreased in linolenic (C183) and linolcic acids (Cl 8: 2). while the value for palmitic acid (C1(ß: 0) increased when the heating time was prolonged for palm 21 olein. Nloreover, therc was a significant difference (p< 0.05) in terms of IV for each Oil during the 5 days of heating. 2.3.2 Lard Profiling Animal fat such as lard and vegetable oils were composed of triacylglycerols (TAG), diacylglycerols (DAG), free fatty acids and other minor components including sterols. caratenoids and fat soluble vitamins (Gunston, 2004). However, according to Andrikopolous et al. (2002), the main classes found in fats and oils are TAGS. Therefore. Rohman ct at. (2012) have done an experiment using high liquid performance chromatography with refractive index detector to differentiate the composition of TAG between lard, and other animal fats such as mutton, chicken 1ät and beef as well as cod liver oil. They found that the main TAG composing lard are palmitoolein (P00), palmitoolcostearin (POS) and palmitooleopalmitin (POP) accounting of'? 1.55 ± 0.05,14.08 ± 0.04 and 5.10 f 0.04% respectively. Dahimi et al. (2013) invcstigatc the use of Gas Chromatography with Flame ionization Detector ((-jC-FID) coupled with chemomctrics techniques to differentiate lard at very low concentrations in beef and chicken has. The results obtained showed that lard contains higher fatty acid (FA) of linolcic acid (C IS: 2cis) and low palmitic acid (C 16: 0) but the result was opposite for beef tallow and chicken fats. The majority Of the compounds are well known lipid oxidation products including, ketones, aldehydes as well as alcohols. AVhereas the most detected compounds were aldehydes and ketones. It was reported that almost all the aldehydes present in pork such as heptanal and nonanal are oxidation products of oleic acid and 22 linolcic acids which were the most abundant unsaturated fatty acids of pork. (Meinert, et al.. 2007). The result obtained also showed that the volatile profile of pork contained a higher proportion of heptanal, which correlated with the experiment done by Shahidi (1997) that aldehydes are the major components identified in the volatiles of cooked pork. Xu et al. (2012) identified a total of 44 volatile compounds in oxidised lard hich 13 comprised from aldehydes, 6 acids, 9 alcohols, 6 ketones, 6 hydrocarbons, 3 esters and I furan. Moreover, they found that aldehydes were the largest group of volatile compounds based on peak areas. Among these aldehydes, (E)- 2-decenal, 2- undecenal, (E) - ocetenal, nonanal. octanal, (E, E)- 2,4- decadienal and (E)- 2- heptenal were predominant in oxidised lard. However, Um et al. (1992) identified that (E)- 2-nonenal. (E)-2-decenal and 2- undecenal were present in the heated beef at high concentrations. When they compare between two breeds of lard (ENL and TCL), hexanal was one Of the main volatile oxidation products. 1lexanal is a typical oxidation volatile from linoleic acid which acts as source of fatty aroma. (Stahnke, 1994). Moreover, it is also commonly been monitored as a measure for lipid oxidation in foods. Meanwhile, decomposition of linolcic acid hydroperoxide will produce (E, E)- 2,4- decadienal (Torres et al., 2005) and volatile oxidation of linolcic acid is (E)- 2- heptenal ( Lee ct al., 2007). Nonanal, (E)- 2- decenal and 2- undecenal has positive relation with oleic acid (Torres et al., 2005). I-Heptanal and octanal are also originated from oleic acid (Machiel, 2004). At low concentration, these alkanals are important since they contribute to pleasant fruity aroma. However, at high concentration, they will produce sharp and pungent attributes (Paleari et al., 2006). Jelen at al. (2006) 23 reported that the fatty acid composition in oil will greatly influenced the production of volatile composition compounds. Xu et al. (2012) detected two most abundant alcohols in lard which are 1- octanol and 1-octen- 3- ol. 1-octen- 3- of is generated from linoleic acid and the contributor to the off- flavour. 2- pentyl furan was the only furan detected in the two lard sources and is a product from linoleic acid. Fatty acid methyl esters (FAME) profiles may be used to discriminate lard from other animal fats. In the study done Indrasti et al. (2010), when gas chromatography hyphenated with time- of- flight mass spectrometry (GC-TOF-MS) combine with two different microbore columns (SLB-5 ms and DB-wax), the differentiation of lard from other animal fats by three FAMES constituents are methyl trans-9,12,1 5-octadecatrienoate (C 18: 3 n3t), methyl 1 1,14,17- eicosatrienoate (C20: 3 n)t) and methyl 11,14-cicosadicnoate (C20: 2n6) which are not present in other animal or plant fats. 2.4 PRINCIPLE COMPONENTS ANALYSIS (PCA) 2.4.1 Introduction Principle components analysis Used a statistical technique to transform an original data set of variables into a smaller set on 1111Correlated variables. It has the ability to represent most o1 the information in the original data set. The technique has hccll used to wide area such as biology, medicine, Chemistry, meteorology geology and social SclellCe. Principal components analysis is S111111a1- to Other IllUltlV-arlate procedures like discriminant analysis and canonical correlation analysis. Both ofthelll 24 involve linear combinations of correlated variables which variables wei