Authentication of butter from lard adulteration using high-resolution of nuclear magnetic resonance spectroscopy and high-performance liquid chromatography

ABSTRACT

Food authentication is an interesting issue for all parties in the food industry, including the fats and oils industry. Some unethical players try to blend high-quality foods, such as butter, with lower ones like lard, therefore, the analytical methods capable of analyzing the adulteration practices must be developed. This study used proton nuclear magnetic resonance spectroscopy in combination with high-performance liquid chromatography for the authentication of butter from lard adulteration. The identification of triacylglycerol composition of lard as a chemical marker for halal authentication is analyzed using high-performance liquid chromatography and high resolution nuclear magnetic resonance spectroscopy. The suitability of proton nuclear magnetic resonance provides a high-performance approach for determination butter adulterated with lard in their entirety of all proton bearing components. Peaks in the region of 2.60–2.84 ppm show special characteristics only present in lard. Only lard has its own unique characteristics which only polyunsaturated fatty acids would give signals 7 at δ 2.63, that corresponded to the chemical shift of the double-allylic methylene protons. In the same way, the intensity of signal at 2.63 ppm, due to methylenic protons in a position α to two double bonds, that is to say, due to the linoleic group. Furthermore, we also correlate some signals between ¹H and ¹³C-NMR spectra for the confirmation of signals.

KEYWORDS:

• Triacylglycerol composition
• Lard
• Butter
• NMR spectroscopy
• Halal authentication

Introduction

Food authentication has received much attention from consumers, producers, and policy makers. The authenticity of food mainly focused on two domains: food safety and quality. However, this field becomes more complex due to fraud practices in manufacturing, such as adulteration and mislabeling. One of the main foods that play a role in feeding the population are dairy products. This dairy product is essential for everyone, especially for certain groups of consumers (women, children, and the elderly).^[1] The most expensive commodities derived from dairy products is milk fat. For this economical reason, the detection of foreign fat in milk fat is scrutinized issues. There are several ways of adulteration for milk fat, such as homogenization of skimmed milk with less expensive foreign fat and by direct incorporation of foreign fat in butter or butter fat. These frauds were developed in the 1980s^[2,3–5] and were standardized.^[6]

Butter is one of the milk fat products that is economically important to the dairy industry, high in nutrition, and functional to customers. Due to high standard of quality of the products, the authenticity of butter is of great importance for commercial and health reasons. Butter is the only one original product that exclusively made from cow milk. It is made by churning fresh or fermented cream or milk. It is high in lipid cholesterols and saturated fat, which consist of phospholipids, tocopherols, and carotenoids.^[7] Butter is one of the edible animal fats used as a finished product for cooking and as an ingredient in various alimentary preparations.

Recently, lard is commonly used as an edible animal fats raw material in food products. The presence of lard in food products can be viewed from two perspectives, economy and religion. On an economic perspective, some food products make good blends with lard to produce cost-effective products such as margarines, shortenings, and other oil-based foods.^[8] Therefore, the food industry likes to mix lard with butter to lessen the production cost.^[9] From a religious point-of-view, religions such as Islam and Judaism forbid their followers to consume and use food that contains lard or any pig-derived ingredient.^[10] For these reasons, analytical methods are urgently needed for rapid and reliable techniques to detect lard in food products.

Several research groups have developed the analytical methods to detect the presence of lard, mostly based chromatographic methods^[11] and differential scanning calorimetry;^[12] however, most of them are time and resource consuming.^[13,14–18] These disadvantages create an urgency to find simpler and effective methods to identify and quantify adulterated butter. In this study, we proposed proton nuclear magnetic resonance (¹H-NMR) spectroscopy and high-performance liquid chromatography (HPLC) for analysis of butter from lard adulteration.

¹H-NMR spectroscopy is a versatile spectroscopic technique for studying opaque heterogeneous samples that has already proven to have a number of useful applications in dairy research.^{[19] 1}H-NMR is an important tool in the study of food, lipids, oils, and fats, which has gained growing applications, particularly because of the great amount of information. It allows discrimination between oils of different composition in a very short period of time.^[20] Since the NMR spectra obtained are too complex to be analyzed visually, the classification of spectra in this study was carried out by coupled with chemometric analysis.^[21] Qualitative investigations based on the combination of NMR spectra and chemometrics of principal component analysis have been commonly applied in the fields of metabolic profiling,^[22] food technology,^[23,24] and recently, for the classification of feverfew preparations.^[25] The present study highlights the application of ¹H-NMR spectra combined with HPLC for identify the triacylglycerol (TAG) composition of lard as a chemical marker for Halal authentication.

Materials and methods

Sample preparation

A pure lard sample was extracted by rendering the adipose tissue from various parts of slaughtered pigs, which were obtained from several local markets in Selangor, Malaysia. Sampling was done in triplicates and consisted of three pig samples. Adipose tissues were heated using a microwave according to the method proposed by De Pedro et al.^[26] The extracted lard was filtered through double-folded muslin cloth, and anhydrous sodium sulfate was added to the extract to remove residual moisture. The extract was then filtered through Whatman No. 2 filter paper and stored at 4°C before further analysis.

Butter sample preparation

Fresh milk from the Universiti Putra Malaysia (UPM), Serdang dairy farm field was used as a raw material. The heavy cream was taken out from the milk by using a cream separator. The heavy cream then was shaken in the mixer with a medium-high speed until the color of cream changed to a pale yellow. The pale yellow buttermilk was collected and kept at –20ºC before further usage.

NMR measurement

A 10 mg sample of each sample was dissolved in 700 µL of deuterated methanol containing 0.5% tetramethysilane (TMS) and mixed using a vortex mixer. The sample was then sonicated for 1 h in a sonicator bath and left to solubilize for a further 12 h at room temperature. This was followed by centrifuging for 5 min to remove insoluble components, after that, the clear supernatant was transferred into NMR tubes. ¹H-NMR spectra were recorded at 26ºC on a 500 MHz Varian INOVA NMR spectrometer operating at a proton NMR frequency of 499.887 MHz. Each ¹H-NMR spectrum acquired over a spectrum width of 20 ppm, consisted of 64 scans requiring 15 min acquisition time. The resulting spectra were manually phased and base-line-corrected, using Chenomx software (v. 5.1, Alberta, Canada). The spectra were referenced to TMS at δ 0.00 ppm.

Bucketing of ¹H-NMR spectra

¹H-NMR spectra were automatically reduced to ASCII files using Chenomx software (v. 5.1, Alberta, Canada). Spectral intensities were binned by equal width (δ 0.04) corresponding to the region of δ 0.50–10.00. The region of δ 4.70–4.96 (water) and δ 3.28–3.33 (residual methanol) were excised from the analysis.

Analysis of TAG compositional by HPLC

The composition of TAG was analyzed using reverse-phase HPLC (Waters, Milford, MA) coupled with refractive index detector (Waters, Milford, MA). The oil samples were diluted in acetone (1:9 v/v) and directly injected into HPLC instrument. The column used in this study was a LiChroCART 100-RP-18 (12.5 cm × 4 mm i.d.; thickness 5 µm, Merck, Darmstadt). The mobile phase consisting of acetone: acetonitrile (63.5:36.5 v/v) and the flow rate was 1 mL/min at 30°C. For each analysis, 10 µL of sample solutions was injected into HPLC system. TAG peaks were analyzed by the Empower software (Milford, MA) and identified based on the retention time of TAG standards, which were then presented as percentage areas.

Determination of fatty acid composition

Fatty acid methyl esters (FAME) were prepared by dissolving a 50 mg portion of oil in 0.8 mL of hexane and adding a 0.2 mL portion of a 1 M solution of sodium methoxide (PORIM, 1995) followed by analysis on a gas chromatograph (Agillent Technologies, Singapore) fitted with an flame ionization detector (FID). The gas chromatography (GC) condition and fatty acid calculation were according to Marikkrar et al.^[27]

Results and discussion

Spectral analysis

Fats and oils are typically consisted of fatty acids esters of glycerol and some minor components. The ¹H-NMR spectra of mixtures of fatty acid compounds such as TAG found in vegetable oils as well as animal fats and methyl esters are characterized by several salient regions of the fatty acid chains containing the signals of specific type of protons. The signals of these types of protons in the ¹H-NMR spectra of vegetable oils were used to quantitate individual unsaturated fatty acids (USFA).

Figure 1 shows the typical ¹H-NMR spectra of pure lard (A) and pure butter (B). Both spectrums of these samples consist of nine signals of variable intensity depending on the proportions of different acyl groups. These spectra signals are mainly due to the protons of the main components such as TAG, and their area is proportional to the number of hydrogen atoms of each kind presents in the sample. The range of chemical shifts at 0 to 4 ppm cover most the free fatty acids and fatty acids that attached to the glycerol while chemical shifts range 4 to 6 ppm is assigned to the signals of proton at glycerol backbone in butter and lard. The spectra of fats sample usually have the similar profile, with difference only apparent in the intensities of the peak. However, careful observation on the extended spectra shows significant differences between sample.^[20] General assignment of the signals to these spectra to the different kind of hydrogen atoms is well-known^[20,28,29] and is given in Table 1.

Table 1. Assignments of signals of ¹H-NMR spectra from samples. Signal number agrees with those in Fig. 3.

Signal	Chemical shift (ppm)	Functional group
1	0.83–0.93	–CH3 (saturated, oleic, and linoleic acyl chains)
2	0.93–1.03	–CH3 (acyl chains)
3	1.22–1.42	–(CH2)n– (acyl chains)
4	1.52–1.70	–OCO–CH2–CH2– (acyl chains)
5	1.94–2.14	–CH2–CH = CH– (acyl chains)
6	2.23–2.36	–OCO–CH2–(acyl chains)
7	2.60–2.84	= HC–CH2–CH = (acyl chains)
8	4.10–4.32	–CH2OCOR (glyceryl group)
9	5.20–5.26	>CHOCOR (glyceryl group)
10	5.26–5.40	–CH = CH– (acyl chains)

Figure 1. ¹H-NMR spectra of A: pure lard; and B: pure butter.

Signal 1 is produced by the overlapping of the triplet signals of methyl group hydrogen atoms of saturated, oleic (n-9) and linoleic (n-6) acyl groups and signal 2 is a triplet due to the methyl hydrogen atoms of n-3 acyl groups. The difference in chemical shifts between both methyl hydrogen atom signals is due to their proximity to the double bond of the chain. Signal 3 is due to the hydrogen atoms of those methylene groups either in position β or further from olefinic groups, or in position γ or further from carbonilic groups inside the triglyceride molecule. Signals 4 and 6 are due to methylene hydrogen atoms in β- and α-position, in relation to the carbonyl group, respectively. Signal 5 is due to α-methylene hydrogen atom in relation to a single-double bond, also named allylic hydrogen atoms, and signal 7 is due to the overlapping of the signals of α-methylene hydrogen atom in relation to two double bonds, also named bis-allylic hydrogen atoms. Signal 8 is due to the hydrogen atoms on 1 and 3 carbon atoms of the glyceryl group, and signal 9 is due to the hydrogen atom on the carbon atom 2 of the glyceryl group. This latter signal overlaps slightly with signal 10 due to olefinic hydrogen atoms of the different acyl groups.

The ¹H-NMR spectra of butter with 3% of lard to 80% of lard are shown in Fig. 2. The entire range of spectra looks similar with the naked eye, unless one observes very closely, especially in highlighted peak region 2.60–2.84 ppm. This is due to the similar composition between both of the samples. Taking into account the differences between the butter and lard spectra, it is obvious that peak intensities at 2.60–2.84 ppm are different. Therefore, these frequencies were selected to be optimized for analysis of lard in butter because ¹H-NMR spectra variation was observed between lard and butter. Qualitatively, it is easy to differentiate between 3% lard and 80% lard in this region. The spectra of lard having a peak at 2.63 ppm which is known the overlapping of the signals of α-methylene hydrogen atom in relation to two double bonds, also named bis-allylic hydrogen atoms, whereas the spectra of butter has no peak at the 2.63 ppm.

Figure 2. ¹H-NMR spectra of butter and lard mixtures showing changes in the height ratio at 2.60–2.84 ppm due to the increase of lard percentages at 0 to 10 ppm.

Fatty acid profile

As pointed out previously, the area of these ¹H-NMR spectra signal is proportional to the number of protons of both samples butter and lard. The fatty acid profile of butter and lard is the major factor influencing the chemical and physical properties and subsequently their various applications. Evidently, NMR techniques alone will not be able to provide detailed information about the molar percentage of any specific fatty acid present in a mixture of TAG. This compositional information can only be provided by the chromatographic methods. The fatty acid composition of butter and lard are listed in Table 2. The major fatty acid composition of butter were myristic, palmitic, and oleic acids with value 14.79 ± 0.05, 31.29 ± 0.05, and 22.77 ± 0.13, respectively, while lard consist of palmitic, oleic, and linoleic with values 23.37 ± 0.22, 40.15 ± 0.31, and 17.29 ± 0.09, respectively. Butter has more saturated fatty acids (SFA) 61.68 compared to lard with a USFA value of 60.2.

Table 2. Fatty acid composition of butter and lard by gas chromatography mass spectrometry analysis.

FA^†	Butter	Lard
C14:0	14.79 ± 0.05	1.11 ± 0.02
C16:0	31.29 ± 0.05	23.37 ± 0.22
C16:1	0.00 ± 0.00	1.50 ± 0.02
C18:0	11.62 ± 0.04	13.26 ± 0.18
C18:1	22.77 ± 0.13	40.15 ± 0.31
C18:2	1.64 ± 0.00	17.29 ± 0.09
C18:3	0.74 ± 0.00	0.89 ± 0.02
SFA	61.68	38.34
USFA	25.15	60.2

TAG composition

TAG are the main constituent of vegetable oil and animal fat, and make up most of the fats digested by humans. They are important in that they allow the uptake and transport of fat-soluble vitamins. They also play a role in metabolism (unused, saturated, or monounsaturated fatty acids are stored by the body as triglycerides). However, triglyceride intake should be monitored because high levels of triglycerides have been linked to an increased risk of heart disease and stroke. Table 3 compiled the TAG composition of butter and lard. Butter is actually a complex mixture of different saturated triglycerides. The profile differences can be attributed to the fact that these products are from different sources; butter is processed from cow milk or cream; whereas, lard is rendered from pig fatty tissues. The TAG peak identification in this study was based on retention time of TAG standards. Butter consists of shorter length of fatty acid compared to lard.

Table 3. Triacylglycerol (TAG) composition of lard, butter, and other animal fats.

TAG^†	Lard	Chicken fat	Beef fat	Mutton fat	Butter
CClCl	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	4.57 ± 0.52
CpCpLa	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	6.06 ± 0.42
CpCLa	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	5.02 ± 0.64
CCLa	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	12.99 ± 0.02
CLaLa	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	12.34 ± 0.14
LaLaLa	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	15.8 ± 0.18
LaLaM	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	10.46 ± 0.11
LaLaO	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	5.02 ± 0.60
LaLaP	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	3.65 ± 0.06
LaMO	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	5.38 ± 0.07
LaMP	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	5.07 ± 0.06
LaOO	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	4.40 ± 0.00
LaPO	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	6.20 ± 0.02
MOO	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	2.88 ± 0.03
POO	23.37 ± 0.07	28.68 ± 0.02	18.26 ± 0.03	18.58 ± 0.01	1.64 ± 0.02
PPO	8.05 ± 0.12	18.55 ± 0.31	10.95 ± 0.06	11 ± 0.13	1.8 ± 0.02
POS	18.14 ± 0.37	4.38 ± 0.05	22.67 ± 0.14	20.4 ± 0.23	0.4 ± 0.01
PPS	0.91 ± 0.03	1.00 ± 0.00	7.36 ± 0.32	6.4 ± 0.32	0.19 ± 0.00
POL	21.6 ± 0.58	12.52 ± 0.23	0.95 ± 0.12	0.95 ± 0.12	0.00 ± 0.00

Note: TAG: triacylglycerol; C: Capric; La: lauric; M: myristic; P: palmitic; S: stearic; O: oleic; L; linoleic.

The TAG distributional profiles of butter and lard are compared. The predominant TAGs of butter were CCLa (12.99 ± 0.02), CLaLa (12.34 ± 0.14), and LaLaLa (15.8 ± 0.18); while lard was POO (23.37 ± 0.07), palmitoyloleoyllinoleoyl (POL; 21.6 ± 0.58), and POS (18.14 ± 0.37). This is an agreement with the previous findings of Rashood et al.^[30] and Yanty et al.^[31] Lard itself has its own special characteristic in that it has high predominance SFA, especially palmitic acid at the sn-2 position, unlike most other fats and oils.^[32] The three TAGs namely LaLaLa, CLaLa, and CCLa were the only TAG species occurring in amounts higher than 10%. Most of the remaining peaks of the chromatogram represent either minor or trace levels of TAG species.

¹³C-NMR and heteronuclear multiple bond correlation (HMBC)

The high resolving power of ¹³C-NMR spectroscopy allows many of the positional isomers of long-chain fatty acids or esters to be differentiated. For example, in the spectrum of lard (Fig. 3), the shifts of the carbonyl region appear at δ173.29 signals correspond to different carbonyl carbons from fatty acid. In general, the chemical shift in this signal depends on the TAG, the stereospecific confirmation (fatty acid in α [sn-1,3] or β [sn-2] position in acylglycerol), and the distance (Δ) of the nearest double bond to the carbonyl carbon. Since the positional distribution of fatty acids in TAG seems to be unique for each species.^[33] The carbonyl region can be of value for lard identification. Additionally, it may be possible to discover if the lard has been subjected to adulteration from other species from this region of the spectra. In general, stereospecific analysis of TAG has shown that the positional distribution of fatty acids varies among plant and animal fat.^[34,35] In vegetable oils, linoleic acid is mainly present in sn-2 position, while SFA are concentrated in the sn-1 and sn-3 positions.^[36] In animals, the stereospecifity varies among species^[33,37] and among parts of the same animal. Studies of lard has shown that palmitic acid are preferentially esterified at the sn-2,^[38,39] while saturated and monounsaturated fatty acid are preferably esterified at the sn-1,3 positions.^[34,35] From the olefinic region 132.5–126.6 ppm of the ¹³C-NMR spectrum, information about USFA can be obtained.^[37] Since fatty acid composition varies among species and geographical origin, addition of different types of oil can be seen in this region. Olefinic carbon atoms in n-3 fatty acids display characteristic peaks because of the influence on the chemical shift from any methyl end group in the vicinity. The two resonances at 129.49 and 129.35 ppm, respectively, are unique for fatty acid with a double bond at n-3 position, and these peaks allow quantification of the relative concentration of n-3 fatty acid in lipid mixtures. As for the carbonyl region at 61.89 where resonances from glycerol carbons appear gives valuable information about the glycerol carbon resonances are influenced by the stereospecific conformation and the nature of the esterified fatty acid, namely the distance to the nearest double bond. For TAG the resonances from glycerol carbons esterified at sn-1, 3 and sn-2 positions are separated by approximately 7 ppm. Consistent with findings from the carbonyl regions, observations in this region confirm that several glycerol ester species in addition to TAGs are present in the lard.

Figure 3. ¹³C nuclear magnetic resonance (NMR) spectrum of lard.

¹³C-NMR spectra of fatty acids provided a large number of signals spread over a wide range of chemical shifts, which made the spectrum appear complicated, but much more informative. Techniques of correlating signals between ¹H-NMR and ¹³C-NMR spectra provided HMBC, is some of the latest technique in NMR spectroscopy from which structural details can be derived in great detail. To reconfirm these assignments, the HMBC spectrum was recorded (Fig. 4). This technique shows correlations of nuclei (carbon to carbon or carbon to proton) which are two or three bonds apart. Figure 5 shows the structural correlation between proton and carbon. The signal at δ_c 133.2 (C-1) was correlated to the proton signals at δ_H 5.37 (H-1). The signals at δ_c 123.8 (C-2) correlated to the proton signals at δ_H 5.43 (H-2) and the signal at δ_c 24.3 (C-3) correlated to the proton signals at δ_H 2.63 (H-3). These two- or three-bond connections reconfirmed the assignments of the carbon atom of the POL.

Figure 4. Heteronuclear multiple bonds correlation (HMBC) spectrum of lard.

Figure 5. Structural correlation between proton and carbon.

Conclusion

The TAG composition of lard as a chemical marker for Halal authentication was successfully identified using high-resolution NMR spectroscopy. The suitability of ¹H-NMR together with HPLC provided a high-performance approach for the determination of butter adulterated with lard in their entirety of all proton bearing components. Peaks in the region of 2.60–2.84 ppm shows special characteristics, present only in lard. Therefore, these frequencies were selected to be optimized for analysis of lard in butter, because ¹H-NMR spectra variation was observed between lard and butter. Only lard have its own unique characteristics which only polyunsaturated fatty acids would give signals 7 at δ 2.63 which corresponded to the chemical shift of the double-allylic methylene protons. In the same way, the intensity of signal at 2.63 ppm, due to methylenic protons in a position α to two double bonds, that is to say, due to the linoleic group. Techniques of correlating signals between ¹H and ¹³C-NMR spectra provided HMBC, which permitted confirmation of signals. This technique shows correlations of nuclei (carbon-to-carbon or carbon-to-proton) which are two or three bonds apart. These two- or three-bond connections reconfirmed the assignments of the carbon atom of the POL. This method is rapid, non-destructive, and easy to use. Great precision of a method is a requirement of quality control in the dairy product industry .Thus, the combination of ¹H-NMR and HPLC could be a powerful alternative for identification of lard adulteration in butter.

Funding

Nurrulhidayah Ahmad Fadzillah acknowledges MyPhD for the scholarship to pursue the Ph.D. program in Halal Products Research Institute, Universiti Putra Malaysia (UPM). This research was supported by Universiti Putra Malaysia (UPM), also providing the funding support awarded to Professor Dr. Russly Ab Rahman through the RUGS 9316900 grant. She also acknowledges the Research Management Centre, Universiti Teknologi Malaysia for providing her with a postdoctoral research fellowship (project number QJ130000.21A2.02E45).

Additional information

Funding

Nurrulhidayah Ahmad Fadzillah acknowledges MyPhD for the scholarship to pursue the Ph.D. program in Halal Products Research Institute, Universiti Putra Malaysia (UPM). This research was supported by Universiti Putra Malaysia (UPM), also providing the funding support awarded to Professor Dr. Russly Ab Rahman through the RUGS 9316900 grant. She also acknowledges the Research Management Centre, Universiti Teknologi Malaysia for providing her with a postdoctoral research fellowship (project number QJ130000.21A2.02E45).