**2. A review of HPLC methods used to determine feed additives**

Feed additives are commonly used in animal nutrition, e.g. in order to supplement the animals' requirement for nutrients (amino acids), useful micro components (vitamins), to prevent invasive diseases (e.g. coccidiosis – coccidiostats), to reduce oxidation processes of feed's components (antioxidants), to enhance the dietary value and quality in food products of animal origin (amino acids, carotenoids– egg yolk coloration).

166 Chromatography – The Most Versatile Method of Chemical Analysis

studies (reproducibility) is taking advantage of Horwitz equation [3].

spectrophotometric measurement rather than on HPLC methods [13,8].

testing feed additives in order to solve the problem of interpreting the results.

for the bias of the method on the basis of CRM studies or PT results.

**2. A review of HPLC methods used to determine feed additives** 

Feed additives are commonly used in animal nutrition, e.g. in order to supplement the animals' requirement for nutrients (amino acids), useful micro components (vitamins), to

chromatographic separation. In examining feed additives with the use of HPLC methods the most frequently used types are spectrophotometric detection (UV-VIS), detection with the help of diode array and fluorescent detection. The choice of optimal parameters for chromatographic separation is done during validation of the method. The analyses in this respect should be accompanied by an assessment of the method's robustness [2]. A practical way to verify the precision of a method in a laboratory (repeatability) and in interlaboratory

Using HPLC methods for examining feed additives was the subject matter of numerous studies on the basis of which official methods of analysing certain feed additives were developed. The studies presented the basic validation parameters for the methods of examining the content of fat-soluble vitamins [4-7], water-soluble vitamins [8,9], coccidiostats [4,10-12], and other feed additives, amino acids, methionine hydroxy analog and antioxidants [13,4,8]. However, in case of carotenoids such as canthaxanthin or apocarotenoic acid ester official methods of examining these additives are still based on

New requirements have been introduced regarding the validation parameters for the methods of analysing feed additives, e.g. those listed in the regulation No 882/2004 [14], taking into consideration, among others, the uncertainty of measurement. It is necessary to determine the uncertainty of measurement with a particular method in order to interpret adequately the result of examining feed additives in feedingstuffs and to assess acceptable tolerance in compliance with the requirements of the regulation No 939/2010 [15]. The new requirements in this area should be taken into account while validating the methods of

The aim of the present work was to offer a review of HPLC methods used for analysing active substances in certain feed additives, with regard to current requirements defined in the regulations. In some justified cases the results of the authors' own studies were presented, as well as the procedures for determining vitamins B1 and B2, canthaxanthin and methionine hydroxy analog (MHA). Some validation parameters were presented, such as the limit of the method's quantification (LOQ), linearity of the calibration curve, repeatability, within-laboratory reproducibility (intermediate precision), recovery and the uncertainty of measurement. Also, the results of verifying the developed methods and laboratories participating in proficiency testing (PT) were demonstrated. The ways of quality assurance of the tests in reference to HPLC methods were discussed. The work presents the method of assessing combined standard uncertainty of measurement with the use of experimental approaches based on within-laboratory reproducibility and calculations In case of some vitamins and antioxidants their maximum content in feed mixtures was determined, e.g. for vitamins A and D3, antioxidants (ethoxyquin, BHA, BHT), carotenoids (canthaxanthin, apocarotenoic acid ester and others). Maximum contents are subject to official control in reference to their conformity with the requirements related to the safety of feedingstuffs. Additionally, a feed manufacturer is obliged to declare the content of feed additives on the label of a premix or feed mixture. Thus, it is necessary to have access to analytical methods for testing the content of feed additives in a wide range of concentrations in the preparations containing additives, premixes and feed mixtures.

Table 1 presents examples of well-known HPLC methods for examining fat-soluble vitamins in feedingstuffs, including the official methods accepted by the European Commission. A commonly used method of preparing a sample for analysing the content of vitamin A is alkaline hydrolysis during which gelatin/sugar cross-linked beadlets which protect vitamin A in the form of retinol acetate are solved and then purified by liquid-liquid extraction. An interesting option for purifying vitamin A extracts from feed mixture with the use of the SPE technique was presented by Fedder & Ploger [7]. The step of alkaline hydrolysis is also used while determining vitamins E and D3. Chromatographic separation of vitamins, except for vitamin D3 where preliminary separation and fraction collection are necessary [5], does not present any serious problems. However, a problem may be posed by the quality and durability of a standard, as well as poor precision resulting from a too low of analytical weight [16]. It is necessary to verify vitamin A standards with the use of a spectrophotometric method [4].

The official methods of determining the content of water-soluble vitamins, such as B1, B2 and B6 are based on spectrophotometric or fluorometric methods [8,13]. The results of analyses using these methods may be biased with errors due to some interferences from other substances in the variable feed matrix. Recently HPLC methods to determine vitamins B1, B2, B6, nicotinic acid and nicotinamide in mineral preparations and mixtures [8], as well as vitamin B1 in feed mixtures and premixes [9] were published (Table 2). Due to the high limit of quantification for vitamin B1 amounting to 5 mg/kg according to Italian Official Method [9], it cannot be used for analysing vitamin B1 in typical feed mixtures to which it is normally added at the amount of 2-4 mg/kg. Moreover, the method quoted above makes it possible to examine vitamin B1 added to feedingstuffs but not the total content of this vitamin, regarding its presence in feed materials. It is thus necessary to have access to chromatographic methods enabling the examination of water-soluble vitamins present in feed materials and added in the form of feed additives. The procedures of HPLC methods of vitamins B1 and B2 developed during the authors' own studies are presented later in the chapter [17,18].


Using High Performance Liquid Chromatography (HPLC) for Analyzing Feed Additives 169

Performance Parameters

Range, mg/kg:

B1= 320-7940; B2= 868- 15990; B6=627-11530; NA=4520-77850; NSA= 3665-61230; RSDr(%) = 2.1- 5.1; RSDR(%) = 4.2-30.2

LOQ=5 mg/kg; Range,

mg/kg: 7 – 484; RSDr(%)=4.2-4.7 RSDR(%)=5-13 Rec.(%)=88-97

Analyte, matrix reference

Vitamin B1, B2, B6, NA, NSA in premixes and mineral feeds [8]

Vitamin B1 in feedingstuffs and premixes [9] Extraction, extract clean up and chromatography parameters

Extraction with methanol -Titriplex solution, clean-up on membrane filter 0.45 μm, reverse phase HPLC coupled to UV or diode array detector, column Nucleosil 250 x 3.0 mm, 5 μm packing; mobile phase: mixture of water solution of acetonitrile and acetic acid

Extraction with methanol ; clean-up on SPE,

fluorescence detector, excitation at 360 nm,

limit of methods' quantification in order to control safe use of coccidiostats.

Numerous methods have been developed to examine coccidiostats in feedingstuffs with the use of high performance liquid chromatography. Examples of such methods are presented in Table 3. Satisfactory precision of such methods has been obtained, in conformity with that calculated from the Horwitz equation [3] and with the requirements of the Commission's Decision [19], which enables analyzing coccidiostats at the levels declared by manufacturers. Due to the hazard of cross-contamination with the remains of coccidiostats found in nontarget mixed feeds on the production line and the risk of carry-over the remains of contamination onto the products of animal origin, it is necessary to continue lowering the

Table 4 presents HPLC methods for testing other feed additives, such as antioxidants, amino acids, methionine hydroxy analog. The official AOAC method for determining ethoxyquin was verified in testing pet food and meat meal [13]. Due to the determination of maximum content of antioxidants in feed mixtures for animals used for food production there is a necessity to check this method in testing typical feed mixtures and premixes. Amino acids are present in typical feed materials as components of proteins. In order to determine amino acids in feed materials it is necessary to subject proteins to hydrolysis and next to separate amino acids using ion-exchange chromatography and apply derivatization. With intensive animal production it is necessary to supplement the deficiency of amino acids, such as lysine, methionine, threonine and tryptophan. New feed additives have been registered recently, such as arginine, valine and cysteine. The official AOAC methods make it possible to determine mainly the composition and content of amino acids in feedingstuffs after hydrolysis [13], yet validation parameters of the determination methods have not been defined for all synthetic amino acids. Moreover, the precision parameters of the method used to determine amino acids with sodium metabisulphite or the hydrobromic acid method were in many cases unsatisfactory, which was confirmed by the values of the Horwitz ratio higher than 2, e.g. in a feed mixture for broiler chickens the HorRat (H) values

reverse phase HPLC, coupled to a

emission at 430 nm

B1 – thiamine; B2 – riboflavin, B6 – pirydoxin; NA- nicotin acid; NSA – nicotinamid; **Table 2.** HPLC methods for the analysis of water-soluble vitamins in feeds

**Table 1.** HPLC methods for the analysis of fat-soluble vitamins in feeds


B1 – thiamine; B2 – riboflavin, B6 – pirydoxin; NA- nicotin acid; NSA – nicotinamid;

168 Chromatography – The Most Versatile Method of Chemical Analysis

Extraction, extract clean up and chromatography parameters

nm or UV detector (325 nm)

330 nm or UV detector (292 nm)

isopropanol (94.5+5+0.5)

nm

Sample hydrolyze with ethanolic KOH, extraction into light petroleum, evaporation and dissolution in methanol, reversed phase HPLC, C18 column (250 x 4 mm) 5 μm or 10 μm packing , mobile phase: methanol and water 98+2 (v/v), fluorescence (or UV) detector: excitation 325 nm; emission 475

Sample hydrolyze with ethanolic KOH solution, extraction into light petroleum, evaporation and dissolution in methanol, reversed phase HPLC, C18 column (250 x 4 mm) 5 μm or 10 μm packing , mobile phase: methanol and water 98+2 (v/v), fluorescence detector: excitation 295 nm; emission

Feed saponification and extraction with diethyl ether; evaporation and solvation in methanol; reverse phase preparative chromatography; eluat collection with vitamin D3 , evaporation and next solvation in n-hexan or isooctane; normal phase chromatography, column 250 mm x 4 mm, Si-60, 5 μm packing; UV detection at 264 nm; mobile phase: preparative column: methanol -water (92+8)), analytical column – n-hexan – dioxan -

Sample extraction with chloroform, transfer vitamin K substances to free menadion; clean-up with Celite and sodium sulphate anhydrous; normal phase chromatography, Si-60 column 250 mm x 4 mm, 10 μm packing; UV detection at 251

Sample hydrolyze with ethanolic KOH solution; clean-up on SPE column; elution in ethyl acetate, evaporation and dissolution in methanol; reverse phase chromatography, ODS2 column 250 mm x 4.6 mm, 5 μm packing; UV detection at 325 nm

(vitamin A) and 292 nm (vitamin E).

**Table 1.** HPLC methods for the analysis of fat-soluble vitamins in feeds

Performance parameters

LOQ=2000 IU/kg; SDr (%): 3.0-8.1; SDR (%): 6.2-20.0;

LOD=2 mg/kg; LOQ=10 mg/kg; SDr (%): 2.2-4.1; SDR (%): 4.8-12.7

LOQ=1000 IU/kg; RSDr to 5000 IU/kg: 1000 IU/kg; 5000- 20000 IU/kg:20% 20000-100000 IU/kg: 15%;

>100000 IU/kg: 10%

LOQ=0.5 mg/kg RSDr at 1 mg/kg: 10%; SDr at 8 mg/kg: 4%; SDr at the level >2500 mg/kg: 3%;

Range: Vitamin A=1250-20000 U/kg; Vitamin E=3-

300 mg/kg RSDip= 21% (vit. A), 11% vit. E; Rec: vit. A - 80%, vit. E - 110%

Analyte, matrix reference

Vitamin A in feedingstuffs and

premixes, Commission Regulation, 2009

Vitamin E in feedingstuffs and

Vitamin D3 in feedingstuffs and premixes [5]

Vitamin K3 in feedingstuffs, premixes and feed additives [6]

Vitamin A and E, feedingstuffs [7]

premixes, Commission Regulation, 2009

[4]

[4]

**Table 2.** HPLC methods for the analysis of water-soluble vitamins in feeds

Numerous methods have been developed to examine coccidiostats in feedingstuffs with the use of high performance liquid chromatography. Examples of such methods are presented in Table 3. Satisfactory precision of such methods has been obtained, in conformity with that calculated from the Horwitz equation [3] and with the requirements of the Commission's Decision [19], which enables analyzing coccidiostats at the levels declared by manufacturers. Due to the hazard of cross-contamination with the remains of coccidiostats found in nontarget mixed feeds on the production line and the risk of carry-over the remains of contamination onto the products of animal origin, it is necessary to continue lowering the limit of methods' quantification in order to control safe use of coccidiostats.

Table 4 presents HPLC methods for testing other feed additives, such as antioxidants, amino acids, methionine hydroxy analog. The official AOAC method for determining ethoxyquin was verified in testing pet food and meat meal [13]. Due to the determination of maximum content of antioxidants in feed mixtures for animals used for food production there is a necessity to check this method in testing typical feed mixtures and premixes. Amino acids are present in typical feed materials as components of proteins. In order to determine amino acids in feed materials it is necessary to subject proteins to hydrolysis and next to separate amino acids using ion-exchange chromatography and apply derivatization. With intensive animal production it is necessary to supplement the deficiency of amino acids, such as lysine, methionine, threonine and tryptophan. New feed additives have been registered recently, such as arginine, valine and cysteine. The official AOAC methods make it possible to determine mainly the composition and content of amino acids in feedingstuffs after hydrolysis [13], yet validation parameters of the determination methods have not been defined for all synthetic amino acids. Moreover, the precision parameters of the method used to determine amino acids with sodium metabisulphite or the hydrobromic acid method were in many cases unsatisfactory, which was confirmed by the values of the Horwitz ratio higher than 2, e.g. in a feed mixture for broiler chickens the HorRat (H) values

amounted to 1.7-3.6, with mean 2.5, while satisfactory H values are within the range of 0.5 > H > 2. This requires further studies with the use of high performance liquid chromatography in order to determine the total content of amino acids after hydrolysis and added amino acids.

Using High Performance Liquid Chromatography (HPLC) for Analyzing Feed Additives 171

Performance Parameters

Method range: 0.5-300 mg/kg; SDr (%): 4.5-32; SDR (%): 4.5-55; Rec. 60-83%

Method range: 10-200 mg/kg; SDr (%): 2.1-11.5; SDR (%): 2.7-21.5; Rec. 83-103%

Feedingstuffs: SDr (%): 1.6-1.9; SDR (%): 2.2-6.3; Feed materials: SDr (%): 0.8-1.3; SDR (%): 4.1-5.1;

Broiler feed: SDr (%): 1.1-4.7; SDR (%): 6.0-19.8; HorRat: 1.7-3.6

LOD=0.2 g/kg LOQ=0.5 g/kg

~ 2.5

Analyte, matrix reference

Ethoxyquin in pet-food and meat

Phenolic antioxidants\* in fats

Tryptophan in feedingstuffs and premixes[4]

Amino acids in feeds\*\* [13]

feedingstaffs and premixes [8]

MHA in

hydroxy analog

5.00

column derivatisation

BHT – 3,5-di-*tert-*butyl—4-hydroxytoluene; OG, DG – octyl and dodecyl gallate

**Table 4.** HPLC methods for the analysis of other feed additives in feeds

A and D3 , as well as feed colorants, such as canthaxanthin.

phosphoric acid solution ( 23+77)

\*PG-propyl gallate; THBT – 2,4,5-trihydroksybutyrophenone; TBHQ – *tetr-*butylhydroquinone; NDGA –

nordihydroguaiaretic acid; Ionox 100 – 2,6-di-*tert-*butyl-4-hydroxymethylphenol BHA- 3-*tert-* butyl-4-hydroxyanisole;

\*\*Sodium metabisulphite method and hydrobromic acid method not applicable to determination of tyrosine and tryptophan; acid hydrolysis method not applicable for methionine, cysteine and tryptophan; MHA - methionine

**3. A description of feed matrix and active substances in feed additives** 

The difficulty in determining certain feed additives is related with their low stability. In order to obtain a more durable form, resistant to the manufacturing conditions of feed mixtures, the additives are secured by protective coating. This concerns primarily vitamins

meal [13]

[13]

Extraction, extract clean up and chromatography parameters

acetonitrile-methanol (1 +1, v/v)

Extraction with acetonitrile without clean-up, reversed phase HPLC, C18 column (250 x 4.6 mm) 5 μm packing , mobile phase: acetonitrile and 0.01 M ammonium acetate (70 + 30, v/v); fluorescence detector: excitation 360 nm; emission 432

Extraction with acetonitrile, extract is concentrated and diluted with 2-propanol; reversed phase gradient HPLC, C18 column with guard column; mobile phase: (1) 5% acetic acid in water, (2)

For total tryptophan alkaline hydrolise with saturated barium hydroxide solution; for free tryptophan extraction under mild acid conditions; reversed phase HPLC with fluorescence detector, excitation 280 nm, emission 356 nm; C18 column (125 x 4 mm) 3 μm packing; mobile phase: acetic acid and 1,1,1-trichloro-2-methyl-2-propanol solution, pH

Performic acid oxidation of the sample to oxidize cystine and methionine; amino acids liberation from protein by hydrolysis with 6 M HCl; dilution with sodium citrate buffer; amino acid separation on ionexchange chromatograph with ninhydrin post-

Extraction with water solution of acetonitrile; reversed phase HPLC with UV detection at 210 nm; RoSil-NH2 column (250 mm x 4.6 mm, 5 μm packing) with guard column; mobile phase: acetonitrile with


SDr - standard deviation of repeatability; SDR - standard deviation of reproducibility; rec. – recovery; LOD – limit of determination; LOQ – limit of quantification; DMF – N,N-dimethylformamide;

**Table 3.** HPLC methods for the analysis of coccidiostats in feeds


\*PG-propyl gallate; THBT – 2,4,5-trihydroksybutyrophenone; TBHQ – *tetr-*butylhydroquinone; NDGA –

nordihydroguaiaretic acid; Ionox 100 – 2,6-di-*tert-*butyl-4-hydroxymethylphenol BHA- 3-*tert-* butyl-4-hydroxyanisole; BHT – 3,5-di-*tert-*butyl—4-hydroxytoluene; OG, DG – octyl and dodecyl gallate

\*\*Sodium metabisulphite method and hydrobromic acid method not applicable to determination of tyrosine and tryptophan; acid hydrolysis method not applicable for methionine, cysteine and tryptophan; MHA - methionine hydroxy analog

**Table 4.** HPLC methods for the analysis of other feed additives in feeds

170 Chromatography – The Most Versatile Method of Chemical Analysis

Extraction, extract clean up

and chromatography parameters

ammonium acetate buffer solution

v/v), UV detection at 305/392 nm

acids.

[10,4]

Lasalocid, monensin, salinomycin and narasin, poultry

feed [12]

Lasalocid, poultry feeds, premixes [11,4]

Robenidine, feedingstuffs, premixes [4]

Diclazuril, feedingstuffs, premixes [4]

Analyte, matrix reference

Halofuginone, medicated feeds

amounted to 1.7-3.6, with mean 2.5, while satisfactory H values are within the range of 0.5 > H > 2. This requires further studies with the use of high performance liquid chromatography in order to determine the total content of amino acids after hydrolysis and added amino

> Ethyl acetate extraction, purification by ion-exchange chromatography, reversed phase HPLC with UV detection at 243 nm, C18 column (300 x 10 mm) 10 μm packing, mobile phase: mixture of acetonitrile and

> Methanol extraction without clean-up, derivatization with 2,4-dinitrophenylhydrazide (DNP) in acid medium at 55 °C, ODS column (150 x 4.6 mm, 5 μm); eluent: methanol – 1.5% aqueous acetic acid (90:10,

> Extraction into acidified (HCl) methanol, agitation in ultrasonic bath at 40 ºC, filtration through a 0.45 μm filter, reversed phase HPLC, C18 column (125 x 4 mm) 5 μm packing , mobile phase: mixture of phosphorus buffer solution and methanol 5+95 (v/v), fluorescence

> Extraction into acidified (HCl) methanol, clean-up on an aluminum oxide column; reversed phase HPLC, UV detection at 317 nm; C18 column (300 x 4 mm) 10 μm packing; mobile phase: mixture of acetonitrile and sodium and potassium phosphate solution

> Extraction with acidified methanol with internal standard; purification on C18 solid phase extraction cartridge (SPE), evaporation and dissolution in DMF;

reversed phase gradient HPLC, Hypersil ODS column, 100 mm x 4.6 mm, 3 μm packing; mobile phase: (1) aqueous solution of ammonium acetate and tetrabutyl-ammonium hydrogen sulphate, (2)

SDr - standard deviation of repeatability; SDR - standard deviation of reproducibility; rec. – recovery; LOD – limit of

acetonitrile, (3) methanol

determination; LOQ – limit of quantification; DMF – N,N-dimethylformamide; **Table 3.** HPLC methods for the analysis of coccidiostats in feeds

detector: excitation 310 nm; emission 419 nm

Performance Parameters

LOQ = 1 mg/kg ; RSDr (%): 2.0-4.7; Rec.(%) 75.3-98.0; at the level of 3

LOQ = 40 mg/kg conc. range 50-150 mg/kg; RSDr (%): 4- 10; Rec. 85-100%

LOD=5 mg/kg; LOQ=10 mg/kg; RSDr (%): 2.1-5.4; RSDR (%): 5.0-10.7; Rec : feed ≥ 80%; premixes ≥ 90%

LOQ=5 mg/kg SDr (%): 3.3-5.4; SDR (%): 9.7-10.1; Rec. for blanc sample ≥ 85%

LOD=0.1 mg/kg; LOQ=0.5mg/kg; SDr (%): 1.9-17.3; SDR (%): 7.4-18.6; Rec. for blanc sample ≥ 80%

mg/kg

## **3. A description of feed matrix and active substances in feed additives**

The difficulty in determining certain feed additives is related with their low stability. In order to obtain a more durable form, resistant to the manufacturing conditions of feed mixtures, the additives are secured by protective coating. This concerns primarily vitamins A and D3 , as well as feed colorants, such as canthaxanthin.

Vitamin A is produced in the form of gelatin-and-sugar beadlets or fat beadlets. Each beadlet contains ca. 0.5-0.6 μg of vitamin A, as calculated for retinal (ca. 2 IU). The distribution of beadlets in the feed is not equal and the feed enriched in vitamin A tends to segregate vitamin beadlets during the process of manufacturing and transporting the feedingstuff, especially in case of loose products. On the other hand, pelleting feed mixtures or subjecting them to other barothermal processes, such as extrusion or expanding reduces vitamin segregation, yet it lowers their durability at the same time. Ultimately, the unequal distribution of vitamins in feed may affect the precision and accuracy of results of analyses. Grinding the samples may improve the distribution of vitamin A, yet it will also increase the risk of its oxidation. Vitamin A is chemically unstable and its content and biological activity are reduced along with the presence of oxygen from the air, light, humidity, inorganic acids, choline hydrochloride, microelements and peroxides created in the processes of fat oxidation. It is recommended that samples should be ground immediately prior to an analysis into 1 mm particles. Further grinding of the sample before determining the content of vitamins may lead to their decomposition. A useful guideline regarding the preparation of samples for analyses, including the analyses of unstable feed additives such as vitamins, is provided by the currently issued ISO/FDIS International Standard 6498 [20].

Using High Performance Liquid Chromatography (HPLC) for Analyzing Feed Additives 173

2 months 96 °C 2 weeks 2 x 3 x 4

Vitamin retention, %\* Month 0 1 2 3 4 5 6 7 8

> 32 94 99

> 94 98 99

13 93 100

97 100 100

6 93 98

95 100 100

4 92 99

94 98 100

3 85 96

90 90 94

Feed storage time (Coelho [21], Table 10)

Total vitamin reten. %

vitamin retention. The product of the factors in columns 2, 3 and 4 (expressed as a fraction)

Pelleting temp., (Coelho[21], Table 11)

1 2 3 4 5

A beadlet 90 88 98 78 D3 beadlet 91 91 99 82 E acetate 50% 92 91 99 83 Thiamine 77 82 99 63 Riboflavin 91 84 99 76 B12 96 95 100 90 Ca pantothenate 87 84 99 72 Biotin 89 84 99 74 Niacin 90 86 99 77

Similarly, the authors observed in their own studies that the conditions and premix storage time in the laboratory affected the content of vitamins A and E. Fractioned samples of premix, each weighing 100 g, were stored for 8 months at room temperature (22 °C), in a fridge (5 °C) and in a freezer (-18 °C) (Table 6). In the samples stored at room temperature the content of vitamins after 8 months of storage was reduced to as little as 3% of the initial value. The analyses of vitamin content should be performed immediately after receiving the samples by the laboratory, otherwise the samples should be stored in a fridge until the

will let us estimate the retention of a particular vitamin in a particular feed (column 5).

Vitamin Vitamin, premix

analyses can be done.

Item

Vitamin A, initial value 2794000 IU/kg: - room 22 °C - refrigerator 5 °C - freezer -18 °C

Vitamin E, initial value 15.56 g/kg: - room 22 °C - refrigerator 5 °C - freezer -18 °C

(Coelho [21], Table 8)

**Table 5.** Vitamin stability in premixes and feeds (%), (Coelho, 2002)

100 100 100

100 100 100

**Table 6.** The results of laboratory retention of vitamins A and E, %

94 102 100

> 98 99 99

\* expanded uncertainty (k=2) of the result of examining is 16% for vitamin A and 12% for vitamin E

82 99 100

> 96 98 97

64 94 100

94 100 100

Vitamins are protected against oxidation by antioxidants (e.g. ethoxyquin) which are added to the materials creating beadlets. Some forms which are physically and chemically stable are created in this way, e.g. the oleic form of vitamin D3 and more stable compounds (menadione bisulfite, thiamine mononitrate and riboflavin phosphate). Feed additives used in feed manufacturing are introduced on mineral carriers or on wheat bran. The most frequently used mineral carrier is fodder chalk. In case of extracting additives from premixes with the use of diluted acids the influence of the carrier on the conditions of extraction should be considered. In such a situation it is recommended that the robustness of a method to slight changes in the analytical procedure or a change of matrix should be verified [2]. In analyzing feedingstuffs the interfering agent is fat which often occurs in significant amounts on feed mixtures (up to 10%). In high-fat samples containing more than 0.25g of fat in an analytical weight while determining fat-soluble vitamins, additional soaps are formed in the saponification process, which hinder the separation of the examined analyte.

It was possible to resolve the problem of interfering substances after using HPLC methods. However, the diversity of matrices and inhomogeneity of feedingstuffs pose numerous analytical problems while determining feed additives, such as vitamins or carotenoid colorants. The biggest difficulty is related to proper clean-up of the extract and selection of adequate conditions for chromatographic separation.

In case of vitamins, while examining the relevance of the producer's declaration and interpreting the result of the examination, one should be aware of the effect brought about by numerous factors on the content of vitamins in feedingstuffs. There were detailed studies in this respect conducted by Coelho [21]. Table 5, based on Coelho's article, presents the method of estimating the summary influence of different factors, such as the type of a premix and the time passed since the moment of its production (column 2), the type of conditions of hydro- and barothermal processing (column 3), feed storage time (column 4) on


vitamin retention. The product of the factors in columns 2, 3 and 4 (expressed as a fraction) will let us estimate the retention of a particular vitamin in a particular feed (column 5).

172 Chromatography – The Most Versatile Method of Chemical Analysis

which hinder the separation of the examined analyte.

adequate conditions for chromatographic separation.

Vitamin A is produced in the form of gelatin-and-sugar beadlets or fat beadlets. Each beadlet contains ca. 0.5-0.6 μg of vitamin A, as calculated for retinal (ca. 2 IU). The distribution of beadlets in the feed is not equal and the feed enriched in vitamin A tends to segregate vitamin beadlets during the process of manufacturing and transporting the feedingstuff, especially in case of loose products. On the other hand, pelleting feed mixtures or subjecting them to other barothermal processes, such as extrusion or expanding reduces vitamin segregation, yet it lowers their durability at the same time. Ultimately, the unequal distribution of vitamins in feed may affect the precision and accuracy of results of analyses. Grinding the samples may improve the distribution of vitamin A, yet it will also increase the risk of its oxidation. Vitamin A is chemically unstable and its content and biological activity are reduced along with the presence of oxygen from the air, light, humidity, inorganic acids, choline hydrochloride, microelements and peroxides created in the processes of fat oxidation. It is recommended that samples should be ground immediately prior to an analysis into 1 mm particles. Further grinding of the sample before determining the content of vitamins may lead to their decomposition. A useful guideline regarding the preparation of samples for analyses, including the analyses of unstable feed additives such as vitamins,

is provided by the currently issued ISO/FDIS International Standard 6498 [20].

Vitamins are protected against oxidation by antioxidants (e.g. ethoxyquin) which are added to the materials creating beadlets. Some forms which are physically and chemically stable are created in this way, e.g. the oleic form of vitamin D3 and more stable compounds (menadione bisulfite, thiamine mononitrate and riboflavin phosphate). Feed additives used in feed manufacturing are introduced on mineral carriers or on wheat bran. The most frequently used mineral carrier is fodder chalk. In case of extracting additives from premixes with the use of diluted acids the influence of the carrier on the conditions of extraction should be considered. In such a situation it is recommended that the robustness of a method to slight changes in the analytical procedure or a change of matrix should be verified [2]. In analyzing feedingstuffs the interfering agent is fat which often occurs in significant amounts on feed mixtures (up to 10%). In high-fat samples containing more than 0.25g of fat in an analytical weight while determining fat-soluble vitamins, additional soaps are formed in the saponification process,

It was possible to resolve the problem of interfering substances after using HPLC methods. However, the diversity of matrices and inhomogeneity of feedingstuffs pose numerous analytical problems while determining feed additives, such as vitamins or carotenoid colorants. The biggest difficulty is related to proper clean-up of the extract and selection of

In case of vitamins, while examining the relevance of the producer's declaration and interpreting the result of the examination, one should be aware of the effect brought about by numerous factors on the content of vitamins in feedingstuffs. There were detailed studies in this respect conducted by Coelho [21]. Table 5, based on Coelho's article, presents the method of estimating the summary influence of different factors, such as the type of a premix and the time passed since the moment of its production (column 2), the type of conditions of hydro- and barothermal processing (column 3), feed storage time (column 4) on Similarly, the authors observed in their own studies that the conditions and premix storage time in the laboratory affected the content of vitamins A and E. Fractioned samples of premix, each weighing 100 g, were stored for 8 months at room temperature (22 °C), in a fridge (5 °C) and in a freezer (-18 °C) (Table 6). In the samples stored at room temperature the content of vitamins after 8 months of storage was reduced to as little as 3% of the initial value. The analyses of vitamin content should be performed immediately after receiving the samples by the laboratory, otherwise the samples should be stored in a fridge until the analyses can be done.


\* expanded uncertainty (k=2) of the result of examining is 16% for vitamin A and 12% for vitamin E

**Table 6.** The results of laboratory retention of vitamins A and E, %

**Table 5.** Vitamin stability in premixes and feeds (%), (Coelho, 2002)
