Commission Regulation (EC) No 273/2008 (repealed)Show full title

REFERENCE METHOD FOR THE DETERMINATION OF MILK FAT PURITY BY GAS CHROMATOGRAPHIC ANALYSIS OF TRIGLYCERDIES — REVISION 2

1. SCOPE AND FIELD OF APPLICATION

This standard specifies a reference method for the determination of milk fat purity using gas chromatographic analysis of triglycerides. Both vegetable fats and animal fats such as beef tallow and lard can be detected.

Using defined triglyceride equations, the integrity of milk fat is determined. Basically, the method applies to bulk bovine milk, or products made thereof, irrespective of feeding, breed or lactation conditions. Only exceptionally high feeding of pure vegetable oils, such as rapeseed oil, can result in a false positive result. Milk products obtained from individual cows can also cause a false positive result.

In particular, the method is applicable to fat extracted from milk products purporting to contain pure milk fat with unchanged composition, such as butter, cream, milk, and milk powder. Technological treatment of milk fat such as removal of cholesterol or fractionation can cause a false positive result. This is also true for milk fat obtained from skim milk or buttermilk. The method is not always applicable to fat extracted from cheese, because the ripening process can affect the fat composition so strongly that a false positive result is obtained.

Note 1: Butyric (n-butanoic) acid (C₄) occurs exclusively in milk fat and enables quantitative estimations of low to moderate amounts of milk fat in vegetable and animal fats to be made. Due to the large variation in C₄ in the approximate percentage mass fraction range 3,1 % and 3,8 %, however, it is difficult to provide qualitative and quantitative information for foreign fat to pure milk fat mass fractions of up to 20 % [1].

Note 2: Practically, quantitative results cannot be derived from the sterol content of vegetable fats, because they depend on the production and processing conditions. Even more, the qualitative determination of foreign fat using sterols is ambiguous.

2. DEFINITION

Milk fat purity: absence of vegetable and animal fats determined by the procedure specified in this standard.

Note: The purity is determined using S-values, which are calculated from the triglyceride composition. The triglyceride mass fractions are expressed as a percentage.

3. PRINCIPLE OF THE METHOD

The fat extracted from milk or milk products is analysed by gas chromatography using a packed or a short capillary column to determine the triglycerides (TGs), separated by total carbon numbers. By inserting the mass fraction, expressed as a percentage, of fat molecules of different sizes (C₂₄ to C₅₄, using even C numbers only) into suitable TG equations, S-values are calculated. If the S-values exceed the limits established with pure milk fat, the presence of foreign fat is detected.

Note 1: The suitability and equivalence of both packed and capillary columns have been demonstrated previously [2-4].

Note 2: The S-value is a sum of TG mass fractions being multiplied by defined factors respectively.

4. REAGENTS

All reagents shall be of recognized analytical grade.

4.1. Carrier gas: nitrogen or, alternatively, helium or hydrogen, all with a purity of at least 99,995 %.

4.2.

Fat standards, for standardizing a milk fat standard according to Clause 7.3.3.

4.2.1. Triglyceride standards, saturated, suited products are available commercially.

4.2.2. Cholesterol standard.

4.3. Methanol (CH₃OH), free of water.

4.4. n-Hexane (CH₃(CH₂)₄CH₃).

4.5. n-Heptane (CH₃(CH₂)₅CH₃).

4.6. Other gases: hydrogen, purity at least 99,995 %, free from organic impurities (C_nH_m < 1 µl/l); synthetic air, free from organic impurities (C_nH_m < 1 µl/l).

4.7. Anhydrous sodium sulfate (Na₂SO₄).

5. APPARATUS

Usual laboratory equipment and, in particular, the following:

5.1. High-temperature gas chromatograph

The high-temperature gas chromatograph shall be suited for temperatures of at least 400 °C and be equipped with a flame ionization detector (FID). Septa used in the injector shall withstand high temperatures and exhibit a very low degree of ‘bleeding’. For capillary GC, use an on-column injector. Always use graphite seals to connect the column as well as injector and/or detector inserts (where applicable).

5.2. Chromatography column

5.2.1. Packed column

Use a glass column of internal diameter 2 mm and length 500 mm, packed with a stationary phase of 3 % OV-1 on 125 µm to 150 µm (100 to 120 mesh) Gas ChromQ(1). The preparation, silanization, packing and conditioning of the packed column is described in Annex A.

Alternatively a capillary column may be used (5.2.2).

5.2.2. Capillary column

Use a short capillary column, e.g. of length 5 m with a non-polar stationary phase that can withstand temperatures up to 400 °C or more(2)). Condition the column by performing 20 analyses of a milk fat solution (7.2) within 2 to 3 days by using the settings given in 7.3.4.2. After that the response factors (7.3.3) shall be close to 1 and less than 1,20.

Note: Columns with different dimensions and a different non-polar, highly temperature-resistant phase can be used, as long as their performance is consistent with this standard. See also 7.3.4.2.

5.3. Extrelut column, of capacity 1 ml to 3 ml, filled with silica gel, needed for the extraction of milk fat according to 7.1.3 only.

5.4. Graphite seals, capable of withstanding temperatures of at least 400 °C; to be used for the connection of the GC column as well as for the injector and/or detector inserts.

5.5. Water bath, capable of maintaining a temperature of 50 °C ± 2 °C.

5.6. Oven, capable of operating at 50 °C ± 2 °C and 100 °C ± 2 °C.

5.7. Microlitre pipette.

5.8. Graduated pipette, of capacity 5 ml.

5.9. Round-bottomed flask, of capacity 50 ml.

5.10. Erlenmeyer flask, of nominal capacity 250 ml.

5.11. Funnel.

5.12. Fine-pored filter paper.

5.13. Rotary evaporator.

5.14. Ampoules, of nominal capacity 1 ml, fitted with a polytetrafluoroethylene-lined aluminium crimp cap or screw cap.

5.15. Injection syringe, with syringe plunger not reaching into the tip of the needle (packed column GC).

Note: With these syringes better repeatability of the results is obtained.

5.16. Analytical balance, capable of weighing to the nearest 1 mg, with a readability of 0,1 mg.

6. SAMPLING

A representative sample should have been sent to the laboratory. It should not have been damaged or changed during transport or storage.

Sampling is not part of the method specified in this International Standard. A recommended sampling method is given in ISO 707 | IDF 50 [5].

7. PROCEDURE

7.1. Preparation of test samples

Use for test sample preparation one of the three following methods of milk fat extraction.

7.1.1. Isolation from butter or butteroil

Melt 50 g to 100 g of test sample at 50 °C using a water bath (5.5) or an oven (5.6). Place 0,5 to 1,0 g of sodium sulfate (4.7) in a folded filter paper (5.12). Preheat a 250 ml Erlenmeyer flask (5.10) and a funnel (5.11) with inserted filter paper in the oven (5.6) set at 50 °C. Filter, while maintaining the preheated flask, funnel and inserted filter device in the oven, the fat layer of the molten sample. Take care that no serum is transferred.

Only in cases where a limited amount of test sample is available, a smaller test sample may be used and the procedure should be adapted accordingly. However, the handling of a smaller test portion involves a higher risk of obtaining a non-representative sample.

Note 1: Butter can be obtained from cream by churning and thorough washing of the resulting butter grain.

Note 2: The milk fat obtained using the procedure in 7.1.1 will be almost free of phospholipids.

7.1.2. Extraction according to the Röse–Gottlieb gravimetric method

Extract the fat fraction from the test sample by using the gravimetric method described in one of the Standards ISO 1211 | IDF 001D, ISO 2450 | IDF 016C or ISO 7328 | IDF 116A.

Note: If phospholipids are present in the milk fat obtained, a cholesterol peak will be obtained which is increased by approximately 0,1 %. The TG composition standardized to 100 % including cholesterol is thereby influenced only to a negligible extent.

7.1.3. Extraction from milk using silica gel columns

Add, using a microlitre pipette (5.7), 0,7 ml of test sample tempered to 20 °C to a 1 ml to 3 ml Extrelut column (5.3). Allow to distribute uniformly on the silica gel for approximately 5 min.

To denature the protein–lipid complexes, add, using the graduated pipette (5.8), 1,5 ml of methanol (4.3) to the Extrelut column. Subsequently, extract the fat fraction from the test sample with 20 ml of n-hexane (4.4). Add the n-hexane slowly in small amounts. Collect the solvent draining off in a 50 ml round-bottomed flask (5.9) previously dried to a constant, known mass weighed to the nearest 1 mg, recording the mass to 0,1 mg.

Allow the column to drain until empty after the extraction. Distil off the solvents from the eluate on a rotary evaporator (5.13) with its water bath set at between 40 °C and 50 °C. After the solvents have been distilled off, dry and subsequently weigh the round-bottomed flask and its contents to the nearest 1 mg, recording the mass to 0,1 mg. Determine the fat mass yield by subtracting the mass of the dried empty round-bottomed flask from the mass obtained.

Note: Fat extractions by the Gerber, Weibull–Berntrop or Schmid–Bondzynski–Ratzlaff methods or isolation of milk fat using detergents (BDI method) are not suitable for TG analysis, because substantial quantities of partial glycerides or phosholipids can pass into the fat phase. Consequently, the application of this International Standard is limited regarding certain products, particularly cheese.

7.2. Preparation of sample solution

For gas chromatography with a packed column, prepare a 5 % (volume fraction) solution of the fat (obtained according to 7.1) in n-hexane (4.4) or n-heptane (4.5). Depending on the column dimensions, use a concentration of 1 % (0,53 mm, ID wide-bore) or lower for on-column injection with a capillary column.

Based on the column used and the mass of fat obtained in 7.1.3, determine the amount of solvent (4.4 or 4.5) to be added to the test sample material in the flask on the basis of weighing to the nearest 1 mg, and recording the mass to 0,1 mg. Completely dissolve the remainder.

Transfer approximately 1 ml of the sample solution into an ampoule (5.14).

7.3. Chromatographic triglyceride determination

7.3.1. Baseline drift

To minimize baseline rising, the column shall be conditioned as specified in 5.2.2 (capillary column) or in Annex A.4 (packed column).

Note: Because of the high column temperature, the analysis of TGs is particularly susceptible to a rise of the baseline in the high carbon-number range.

7.3.2. Injection technique

7.3.2.1. Packed column

To avoid discrimination effects, apply the hot-needle technique for improving the quantification of the high-boiling TG components. Fill the needle with air by drawing up the fat solution in the syringe. Insert the needle into the injector. Heat the needle up prior to injection for about 3 s. Then, rapidly inject the syringe content.

7.3.2.2. Capillary column

When using cold on-column injection (7.3.4.2), insert the needle of the syringe and inject immediately. The dwell time of the needle in the injection port should be such that broad tailing of the solvent peak is avoided.

Note: The optimum dwell time typically is about 3 s.

7.3.3. Calibration

7.3.3.1. General

For the calibration of test samples, perform two to three analyses of standardized milk fat at the beginning of every day. Use the last analysis of the standardized milk fat to determine the response factors, RF_si (mass fraction/area fraction) of the TGs and of cholesterol and apply these to the subsequent test samples (see 9.1):

(1)

where:

w _si

is the mass fraction, expressed as a percentage, of each TG or cholesterol in the standardized milk fat;

A _si

is the numerical value of the peak area of each TG or cholesterol in the standardized milk fat.

Use either 7.3.3.2 or 7.3.3.3 to obtain a standardized milk fat with a known TG composition.

7.3.3.2. Commercial milk fat standard

The best way to determine the response factor of each constituent of the test sample is to use a standardized milk fat with a certified TG composition.

Note: A suitable standard is CRM 519 (anhydrous milk fat) obtainable from the Institute for Reference Materials and Measurements (IRMM), Geel, Belgium(3)).

7.3.3.3. Laboratory milk fat standard

Prepare about 1 g of a mixture of the fat standards (see 4.2, containing at least the saturated TGs, C₂₄, C₃₀, C₃₆, C₄₂, C₄₈ and C₅₄, as well as cholesterol; plus, preferably, C₅₀ and C₅₂) by weighing to the nearest 1 mg, recording the mass to 0,1 mg, to obtain a relative TG composition similar to milk fat.

Analyse repeatedly a solution of the fat standards mixture in n-hexane (4.4) or n-heptane (4.5) according to 7.3.4. In the same sequence, analyse repeatedly averagely composed milk fat.

Determine the TG response factors from the fat standards mixture. Intermediate response factors of TGs not present in the mixture can be calculated by mathematical interpolation. Apply the response factors obtained to the milk fat, in order to obtain a standardized composition. The standardized milk fat thus obtained has a stock life of several years, if stored under nitrogen at a maximum temperature of -18 °C.

7.3.4. Chromatographic conditions

Note: Use of either packed or capillary columns generally results in a resolution similar to Figure 1. Splitting of the even-numbered TGs is not normally observed and shall be avoided.

7.3.4.1. Packed column

(a)

Temperature programme: Set the initial oven temperature at 210 °C. Maintain it at that temperature for 1 min. Then increase the temperature at a rate of 6 °C/min to 350 °C. Maintain it at that (final) temperature for 5 min.

(b)

Detector and injector temperatures: Set both at 370 °C.

(c)

Carrier gas: Use nitrogen at a constant flow rate of about 40 ml/min. Adjust the exact carrier gas flow in such a manner that C₅₄ is eluted at 341 °C.

(d)

Duration of analysis: 29,3 min.

(e)

Injection volume: Inject 0,5 µl of a 5 % (volume fraction) sample solution.

If no TG analyses are carried out, maintain the initial oven temperature as given in a), the detector and injector temperatures as in b) and the carrier gas flow rate as in c) at constant level, also overnight and during weekends and holidays. This ensures best performance of the column.

7.3.4.2. Capillary column

(a)

Temperature programme: Set the initial oven temperature at 80 °C. Maintain it at that temperature for 0,5 min. Then increase the temperature at a rate of 50 °C/min to 190 °C and subsequently at a rate of 6 °C/min to 350 °C. Maintain it at that (final) temperature for 5 min.

(b)

Detector temperature: Set at 370 °C.

(c)

Carrier gas: Use nitrogen at a constant flow rate of about 3 ml/min.

(d)

Duration of analysis: 34,4 min.

(e)

Injection volume: Inject 0,5 µl of a 1 % (volume fraction) sample solution.

Maintain these settings during standby to ensure best performance (see 7.3.4.1).

The analytical settings given in 7.3.4.2 are suitable for use with a wide-bore column (0,53 mm ID) as specified in 5.2.2. Different conditions may be applied if another column dimension or phase is used.

8. INTEGRATION, EVALUATION AND CONTROL OF THE ANALYTICAL PERFORMANCE

Evaluate the chromatogram peaks with an integration system capable of baseline drawing and reintegration. Figure 1 shows a correctly integrated chromatogram, whereas Figure 2 demonstrates a sporadic error in the baseline ending after C₅₄ that influences the percentages of all TGs. Nevertheless, exclude peaks eluting after C₅₄ from the evaluation.

Combine TGs with an odd acyl-C number (2n + 1) with the preceding even-numbered TG (2n). Do not take into account the low C₅₆ content. Multiply the area percentages of the remaining TGs including cholesterol by the corresponding response factors of the standardized milk fat (latest calibration) and normalize altogether to 100 % according to 9.1.

Figure 1

Example of a triglyceride chromatogram of milk fat with baseline set correctly

Figure 2

Example of a triglyceride chromatogram of milk fat with baseline set incorrectly

To control measuring conditions, compare with the coefficients of variation, CVs, expressed as percentages, of the various TGs given in Table 1 which are based on 19 consecutive analyses of the same milk fat sample.

If the CVs are considerably higher than the values given in Table 1, the chromatographic conditions are not appropriate.

Note: The values given in Table 1 are not mandatory, but are indicative for quality control purposes.

In case, however, that higher CV-values are accepted, the repeatability and reproducibility limits given in Clause 10 shall nonetheless be complied with.

Table 1

Coefficients of variation of triglyceride contents (19 consecutive analyses)

Triglyceride	CV %
C24	10,00
C26	2,69
C28	3,03
C30	1,76
C32	1,03
C34	0,79
C36	0,25
C38	0,42
C40	0,20
C42	0,26
C44	0,34
C46	0,37
C48	0,53
C50	0,38
C52	0,54
C54	0,60

9. CALCULATION AND EXPRESSION OF RESULTS

9.1. Triglyceride composition

9.1.1. Calculation

Calculate the mass fraction of each TG (for i = C₂₄, C₂₆, C₂₈, C₃₀, C₃₂, C₃₄, C₃₆, C₃₈, C₄₀, C₄₂, C₄₄, C₄₆, C₄₈, C₅₀, C₅₂, and C₅₄) plus cholesterol, w_i , expressed as a percentage, of the total TG content of the test sample by using the following equation:

(2)

where

A_i

is the numerical value of the peak area of each TG in the test sample;

RF _si

is the response factor of each TG determined by calibration (7.3.3).

9.1.2. Expression of test results

Express the results to two decimal places.

9.2. S-values

9.2.1. Calculation

9.2.1.1. Calculate the S-values, expressed as a percentage, by inserting the calculated w_i (9.1.1) of the appropriate TG percentages into Equations (3) to (7). Use all equations irrespective of the kind of foreign fat suspected.

9.2.1.2. Soy bean, sunflower, olive, rapeseed, linseed, wheat germ, maize germ, cotton seed and fish oil.

S = 2,098 3 · w _C30 + 0,728 8 · w _C34 + 0,692 7 · w _C36 + 0,635 3 · w _C38 + 3,745 2 · w _C40 – 1,292 9 · w _C42 + 1,354 4 · w _C44 + 1,701 3 · w _C46 + 2,528 3 · w _C50 (3)

9.2.1.3. Coconut and palm kernel fat.

S = 3,745 3 · w _C32 + 1,113 4 · w _C36 + 1,364 8 · w _C38 + 2,154 4 · w _C42 + 0,427 3 · w _C44 + 0,580 9 · w _C46 + 1,292 6 · w _C48 + 1,030 6 · w _C50 + 0,995 3 · w _C52 + 1,239 6 · w _C54 (4)

9.2.1.4. Palm oil and beef tallow.

S = 3,664 4 · w _C28 + 5,229 7 · w _C30 – 12,507 3 · w _C32 + 4,428 5 · w _C34 – 0,201 0 · w _C36 + 1,279 1 · w _C38 + 6,743 3 · w _C40 – 4,271 4 · w _C42 +6,373 9 · w _C46 (5)

9.2.1.5. Lard.

S = 6,512 5 · w _C26 + 1,205 2 · w _C32 + 1,733 6 · w _C34 + 1,755 7 · w _C36 + 2,232 5 · w _C42 + 2,800 6 · w _C46 + 2,543 2 · w _C52 + 0,989 2 · w _C54 (6)

9.2.1.6. Total.

S = – 2,757 5 · w _C26 + 6,407 7 · w _C28 + 5,543 7 · w _C30 – 15,324 7 · w _C32 + 6,260 0 · w _C34 + 8,010 8 · w _C40 – 5,033 6 · w _C42 + 0,635 6 · w _C44 + 6,017 1 · w _C46 (7)

9.2.2. Expression of test results

Express the results to two decimal places.

9.3. Detection of foreign fat

Compare the five S-values obtained in 9.2.1 with the corresponding S-limits given in Table 2.

Consider the test sample as a pure milk fat, when all five S-values fall inside the limits mentioned in Table 2. However, if any S-value falls outside the corresponding limits, the sample is considered to contain a foreign fat.

Though individual Equations (3) to (6) are more sensitive for certain foreign fats than total Equation (7) (see Table B.1), a positive result obtained with only one of Equations (3) to (6) does not allow conclusions to be drawn on the kind of foreign fat.

Annex B describes a procedure for the calculation of the content of vegetable or animal fat in the adulterated milk fat. This procedure is not validated and is informative only.

Table 2

S-limits for pure milk fats

a Calculated on a 99 % confidence level, so that foreign fat addition is only indicated if the detection limits of the relevant equation are exceeded (see Table B.1).
Foreign fat	Equation	S-limitsa
Soy bean, sunflower, olive, rapeseed, linseed, wheat germ, maize germ, cotton seed, fish oil	(3)	98,05 to 101,95
Coconut and palm kernel fat	(4)	99,42 to 100,58
Palm oil and beef tallow	(5)	95,90 to 104,10
Lard	(6)	97,96 to 102,04
Total	(7)	95,68 to 104,32

10. PRECISION

10.1. Interlaboratory test

The repeatability and reproducibility values were determined on the basis of Equations (3) to (7) using pure milk fat and may not be applicable to matrices other than those given.

10.2. Repeatability

The absolute difference between two single test results, obtained using the same method on identical test material in the same laboratory by the same operator using the same equipment within a short interval of time, will not exceed the limits listed in Table 3 in more than 5 % of cases.

Table 3

Repeatability limits, r, for Equations (3) to (7)

Foreign fat	Equation	r %
Soy bean, sunflower, olive, rapeseed, linseed, wheat germ, maize germ, cotton seed, fish oil	(3)	0,67
Coconut and palm kernel fat	(4)	0,12
Palm oil and beef tallow	(5)	1,20
Lard	(6)	0,58
Total	(7)	1,49

10.3. Reproducibility

The absolute difference between two single test results, obtained using the same method on identical test material in different laboratories with different operators using different equipment, will not exceed the limits listed in Table 4 in more than 5 % of cases.

Table 4

Reproducibility limits, R, for Equations (3) to (7)

Foreign fat	Equation	R %
Soy bean, sunflower, olive, rapeseed, linseed, wheat germ, maize germ, cotton seed, fish oil	(3)	1,08
Coconut and palm kernel fat	(4)	0,40
Palm oil and beef tallow	(5)	1,81
Lard	(6)	0,60
Total	(7)	2,07

11. UNCERTAINTY OF MEASUREMENT

With the repeatability, r, and the reproducibility, R, the expanded uncertainty for an S-value can be calculated.

Inclusion of the expanded uncertainty (based on duplicate analyses) into the S-limits of Table 2 results in extended S-limits which are given in Table 5.

Table 5

Extended S-limits for pure milk fats including the expanded uncertainty

Foreign fat	Equation	Extended S-limits
Soy bean, sunflower, olive, rapeseed, linseed, wheat germ, maize germ, cotton seed, fish oil	(3)	97,36 to 102,64
Coconut and palm kernel fat	(4)	99,14 to 100,86
Palm oil and beef tallow	(5)	94,77 to 105,23
Lard	(6)	97,65 to 102,35
Total	(7)	94,42 to 105,58

12. TEST REPORT

The test report shall specify:

• all information necessary for the complete identification of the sample,

• the sampling method used, if known,

• the test method used, with reference to this International Standard,

• all operational details not specified in this International Standard, or regarded as optional, together with details of any incidents which may have influenced the test result(s),

• the test result(s) obtained, and, if the repeatability has been checked, the final quoted result obtained.

(1)

Example of a suitable product available commercially. This information is given for the convenience of users of this International Standard and does not constitute an endorsement of this product.

(2)

CP-Ultimetal SimDist (5 m × 0,53 mm × 0,17 µm) is an example of a suitable product available commercially. This information is given for the convenience of users of this International Standard and does not constitute an endorsement of this product.

(3)

Example of a suitable product available commercially. This information is given for the convenience of users of this International Standard and does not constitute an endorsement of this product.