Dairy Science & Technology · Dairy Chemistry · also Analytical Dairy Chemistry, Quality Control of Milk and Milk Products, Instrumental Methods of Analysis
Testing milk and dairy products: platform tests, reference methods, adulterant detection and modern instruments.
In the exams: ICAR AICE subject 34 Units 4 and 10; ICAR AIEEA-PG group 16 (tests for milk, butter, ghee, powder; adulterants, neutralisers and preservatives); FSSAI Food Analyst examination (dairy portion)
Dairy analysis is the set of tests used to decide whether milk and milk products are genuine, of the composition declared, safe and fit for processing. It ranges from the quick platform tests done in seconds at a village collection centre, through the classical chemical methods for fat, protein and solids, to the instruments of a modern laboratory: infrared analysers, chromatographs, atomic absorption spectrometers, electrophoresis systems and immunoassays. It also covers the discipline that makes results trustworthy: correct sampling, validated methods, reference materials and proficiency testing.
The subject matters because every litre of milk bought, every product released and every enforcement case depends on test results. Payment for milk rests on fat and solids-not-fat measurements; acceptance of tankers rests on acidity, alcohol stability, freezing point and residue tests; regulatory action against adulteration rests on tests that must stand up in court. The underlying chemistry is in the milk-chemistry subjects: Milk Composition and Physical Properties (lactometer, freezing point, acidity, dye-reduction tests), Milk Proteins and Enzymes (nitrogen fractions, phosphatase and peroxidase tests), Milk Lipids (fat constants and oxidation tests) and Lactose, Milk Salts and Vitamins. This manual refers to those subjects for the principles and concentrates on the methods, their procedures, limitations and interpretation. Reagent quantities and limits quoted are those of standard methods as examples; the official method and legal limit in each country must be followed, and India's FSSAI, the Bureau of Indian Standards, ISO and IDF, AOAC and Codex publish them.
The subject is examined in Units 4 and 10 of ICAR AICE-JRF/SRF subject 34 (Dairy Chemistry: adulteration and analytical methods; instrumental analysis), in group 16 of ICAR AIEEA-PG (tests for milk, butter, ghee and powder, adulterants, neutralisers and preservatives), and heavily in the FSSAI Food Analyst and Junior Analyst examinations and food safety officer papers. Expect questions on test principles and reagents, interpretation of positive results, calculation of fat, protein, total solids and analytical constants, principles of instruments, and validation terms such as limit of detection, recovery and z-score.
No analysis is better than the sample on which it is done. A sample must be representative of the lot, must not change between sampling and analysis, and must be identified so that the result can be traced. Milk separates on standing because fat rises, so it must be mixed thoroughly before sampling: by plunging a perforated plunger up and down in a can or tank, by pouring from one vessel to another several times, or by mechanical agitation of bulk tanks and tankers for a set time. Samples are taken with a dipper or a sampling tube that collects a column through the whole depth. Butter and cheese are sampled with a trier or borer that removes a core; powders with a sampling tube or spear reaching through the container; and products in retail packs by taking whole unopened packs. Composite samples, taken in proportion to quantity, are used for payment testing over a period. International guidance on sampling is given in ISO 707 (IDF 50), and sampling plans, which specify how many units to sample from a lot and how to decide acceptance, are given in ISO and Codex guidelines. Samples are put into clean, dry, leak-proof containers, labelled with the source, date, time and sampler, sealed when they may become evidence, and transported cold. Samples for microbiological testing must be taken aseptically with sterile equipment into sterile containers and kept at about 0 to 4 degrees Celsius without preservative, and are usually examined within 24 hours.
Samples for chemical composition may need a preservative to prevent souring before analysis. Preservatives used include bronopol and sodium azide in payment testing, and formalin or other permitted agents in some official procedures; potassium dichromate and mercuric chloride were used historically but are toxic and now avoided. The preservative must not interfere with the test: for example, preserved samples are unsuitable for microbiological or some residue tests. Before analysis, refrigerated milk samples are warmed to about 38 to 40 degrees Celsius, mixed gently to redistribute cream without churning, and cooled to the test temperature; a sample showing churned fat or clots cannot give a valid fat result. In regulatory enforcement, a food safety officer takes a sample in the presence of the seller, divides it into the number of parts prescribed by law, seals and labels each part, sends one to the food analyst and retains others for referee analysis if the result is disputed. The procedures, container types, preservative and quantity required for each product are set out in national rules, and failure to follow them can make an otherwise correct result legally invalid. Good sampling records include the temperature of the product, the method of mixing, any visible abnormality and the chain of custody. The same discipline applies inside plants, where line samples taken at the wrong point or time give misleading results about process control.
Platform tests are quick tests done at the reception dock (platform) of a collection centre or plant to accept or reject milk before it is mixed with good milk. The organoleptic test comes first: the receiver smells the milk as the can lid is opened, looks at its colour and appearance and for foreign matter, and notes its temperature; abnormal odours such as sour, rancid, feed, chemical or putrid lead to rejection or further testing. The clot-on-boiling (COB) test heats a few millilitres of milk to boiling in a test tube; milk that clots or shows flakes is rejected because it would coagulate in processing. It detects milk with developed acidity roughly above about 0.2 per cent lactic acid, as well as colostrum and some abnormal milks. The alcohol test mixes equal volumes of milk and ethanol, commonly 68 per cent; flakes indicate reduced protein stability from developed acidity, high ionic calcium, colostrum or mastitis, as explained in Milk Composition and Physical Properties, and the test is used especially for milk destined for high-heat processing. The alcohol-alizarin test adds alizarin indicator to the alcohol so that the colour also shows the pH: a lilac to red-brown colour without flakes for normal milk, yellow-brown with flakes for acid milk and violet for alkaline (possibly neutralised or mastitic) milk. Titratable acidity can also be checked quickly at the platform.
The sediment test shows visible dirt: a fixed volume of milk, often 500 millilitres, is drawn through a cotton or filter disc, and the dirt on the disc is graded against standard discs; it reflects milking and handling hygiene rather than bacterial load. The lactometer test, with temperature correction and calculation of solids-not-fat, is described in Milk Composition and Physical Properties; at the platform it screens for watering and, together with a fat test, for skimming. Fat is measured at the platform by the Gerber method or, increasingly, by electronic and ultrasonic analysers linked to automatic payment systems. Dye-reduction tests (methylene blue and resazurin), also described in Milk Composition and Physical Properties, estimate microbial activity, and the resazurin test can be read after ten minutes as a rapid platform test. Rapid strip or kit tests for antibiotic residues, aflatoxin M1 and common adulterants are now widely used at chilling centres. The value of platform tests lies in speed and in removing clearly unfit milk before it contaminates a bulk tank; none of them is definitive, and suspect milk is sampled for laboratory confirmation. Results and rejections must be recorded and communicated to suppliers, because platform testing is also a means of improving quality at farm level through feedback and payment incentives.
The Gerber method is the routine volumetric fat test in most of the world. In a Gerber butyrometer, 10 millilitres of sulphuric acid of specific gravity about 1.807 to 1.812 is mixed with 10.75 millilitres of milk (delivered by an 11 millilitre pipette) and 1 millilitre of amyl alcohol. The acid dissolves the proteins and destroys the fat globule membranes, generating heat, so that the fat is released; amyl alcohol helps the fat separate cleanly and prevents charring of fat and sugar. The butyrometer is stoppered, shaken, centrifuged, typically for 4 to 5 minutes at about 1,100 revolutions per minute, and tempered in a water bath at about 65 degrees Celsius; the fat column is then read directly as per cent fat on the graduated stem. Special butyrometers are used for cream (with weighed samples), skim milk and cheese. Sources of error include acid of the wrong strength (too weak gives incomplete digestion and a cloudy column; too strong chars the fat), wrong temperature at reading, poor mixing of the sample and impure amyl alcohol. The Babcock method, used mainly in the United States, uses 18 grams of milk (a 17.6 millilitre pipette) and about 17.5 millilitres of sulphuric acid of specific gravity about 1.82 to 1.83, without amyl alcohol; hot water is added to bring the fat into the graduated neck, and the fat column is read at about 55 to 60 degrees Celsius, often with a coloured mineral oil to flatten the meniscus.
The reference method for fat in milk is the Röse-Gottlieb method (ISO 1211 / IDF 1), a gravimetric solvent extraction. Ammonia is added to the milk to dissolve the casein and neutralise acidity, ethanol to break the emulsion and precipitate proteins, and the fat is extracted with diethyl ether followed by light petroleum. The solvent layer is separated, the extraction repeated, the solvents evaporated and the fat dried and weighed. The Mojonnier method is a widely used apparatus-based form of this extraction. For products in which fat is bound to protein or enclosed in a matrix, such as cheese and some processed products, acid hydrolysis methods are used: the Schmid-Bondzynski-Ratzlaff method digests the sample with hydrochloric acid before ether extraction, and the Weibull-Berntrop method uses acid hydrolysis followed by extraction of the dried residue. Gravimetric methods are slower but are not affected by the assumptions built into volumetric methods, so they are used to calibrate Gerber butyrometers, infrared analysers and other routine instruments, and to settle disputes. Fat results are expressed by mass per cent; for payment and labelling, the method used must be stated, because different methods can give slightly different results on the same milk. Routine laboratories check their Gerber results regularly against Röse-Gottlieb on the same samples and correct for bias.
The Kjeldahl method is the reference method for protein in milk. The sample is digested with concentrated sulphuric acid, with potassium sulphate to raise the boiling point and a copper catalyst, until all organic nitrogen is converted to ammonium sulphate and the digest is clear. The digest is made alkaline with sodium hydroxide, the ammonia is distilled into boric acid solution containing an indicator, and the trapped ammonia is titrated with standard acid. Nitrogen per cent equals (sample titre minus blank titre) times the normality of the acid times 1.4007, divided by the sample mass in grams; it is multiplied by 6.38 to give crude protein. For a 5 gram sample needing 15.5 millilitres of 0.1 normal acid with a blank of 0.2 millilitre, nitrogen is 15.3 x 0.1 x 1.4007 / 5 = 0.429 per cent, and crude protein 2.73 per cent. The nitrogen fractions (non-casein and non-protein nitrogen) are determined by the same procedure on filtrates, as described in Milk Proteins and Enzymes. The Dumas combustion method, which burns the sample and measures nitrogen gas, is faster and avoids hazardous chemicals. Rapid methods include dye binding, in which an acid dye such as amido black or Orange G binds the basic amino acid residues and terminal groups of proteins at low pH, the protein-dye complex is removed and the unbound dye is measured photometrically; and formol titration, in which neutral formaldehyde blocks amino groups, releasing hydrogen ions that are titrated with alkali, the titre being converted to protein with an empirical factor. Rapid methods must be calibrated against Kjeldahl.
Total solids are determined by drying a weighed sample to constant mass. The reference method for milk (ISO 6731 / IDF 21) dries about 5 grams of milk, with or without sand, in a flat-bottomed dish, first on a boiling water bath and then in an oven at about 102 degrees Celsius until constant weight. Total solids per cent equals the mass of the dried residue divided by the mass of milk, times 100: a 5.000 gram sample leaving 0.640 gram of residue has 12.8 per cent total solids. Errors arise from incomplete drying, from browning of lactose at too high a temperature, from absorption of moisture by the residue before weighing (it is cooled in a desiccator) and from loss of volatile acids in sour milk. Solids-not-fat is calculated as total solids minus fat. At reception, SNF is often estimated instead from the corrected lactometer reading and fat content using the formulas given in Milk Composition and Physical Properties, or measured by infrared analysers. Moisture in butter is determined by heating a weighed sample until water is driven off; in cheese and powders by oven drying under specified conditions, and in powders also by Karl Fischer titration or vacuum oven methods. Because lactose in powders may be partly crystalline as the monohydrate, the drying conditions determine whether water of crystallisation is included, so methods for powders specify conditions precisely. Ash is determined by incineration at about 550 degrees Celsius, as discussed in Lactose, Milk Salts and Vitamins.
Infrared milk analysers measure fat, protein, lactose and total solids within seconds and are the workhorses of payment laboratories, herd-recording laboratories and plant quality control. They are based on absorption of mid-infrared radiation by specific chemical bonds. Fat is measured at about 5.7 micrometres, where the carbonyl (C=O) of ester bonds absorbs (fat A), and at about 3.5 micrometres, where carbon-hydrogen bonds of fatty acid chains absorb (fat B); protein at about 6.5 micrometres, the amide II band of peptide bonds; and lactose at about 9.6 micrometres, where hydroxyl groups absorb. Water absorbs strongly throughout, so measurements are made against a water reference. Because fat globules scatter infrared light, the milk is homogenised inside the instrument to a small globule size before it enters the measuring cell, and the sample is warmed to about 40 degrees Celsius. Older instruments used optical filters at fixed wavelengths; modern Fourier-transform infrared (FTIR) instruments record the whole spectrum, which allows more components to be predicted and abnormal samples to be flagged. Near-infrared (NIR) spectroscopy, using overtone absorptions, is used for solid and powdered products such as cheese, butter and milk powder, often without sample preparation. Separate instruments based on the speed of ultrasound through milk, which depends on composition, are widely used at village collection centres because they are cheap and robust.
Infrared analysers do not measure components absolutely; they predict them from absorbance through calibration equations, so their accuracy depends entirely on calibration. Calibration uses a set of milk samples whose composition has been determined by reference methods (Röse-Gottlieb for fat, Kjeldahl for protein, gravimetric drying for total solids, and a reference lactose method), covering the range of composition expected. The instrument's readings are compared with reference values and the slope and intercept of each equation are adjusted so that the instrument agrees with the reference across the range; intercorrection factors correct for the influence of each component on the others' measurements, for example the contribution of fat to the protein signal. Calibration is checked routinely with pilot samples of known composition, and the instrument's repeatability, linearity, homogeniser efficiency and carry-over between samples are checked at set intervals. Many countries operate reference laboratory systems that distribute calibration sample sets to payment laboratories so that all instruments agree. Infrared results can be wrong for abnormal milks and for adulterated milks: added substances such as urea, sugar, salts or vegetable fat can distort predictions or go undetected, and composition far outside the calibration range is unreliable. FTIR spectra can, however, be used to screen for some adulterants and abnormal milks by comparing a sample's spectrum with the spectra of normal milk. Laboratories therefore combine infrared screening with reference methods and specific adulterant tests.
Adulteration of milk takes three forms: dilution (adding water or removing fat), adding substances to hide dilution or spoilage, and substituting cheaper materials. Added water is detected most reliably by the freezing point, and lactometer readings and fat and SNF results reveal watering and skimming, as described in Milk Composition and Physical Properties. Substances added to restore the lactometer reading after watering include sugar, starch, salt, urea, ammonium sulphate, glucose and skim milk powder. Starch is detected by adding iodine solution to boiled and cooled milk: a blue colour indicates starch. Cane sugar is detected by the resorcinol (Seliwanoff) test, in which milk heated with resorcinol and hydrochloric acid turns red because of the fructose in sucrose. Glucose is detected with modified Barfoed's reagent followed by phosphomolybdic acid, which gives a blue colour. Urea is detected with p-dimethylaminobenzaldehyde, which forms a yellow colour with urea; because milk naturally contains some urea, quantitative methods and limits are needed to distinguish natural from added urea. Neutralisers, such as sodium carbonate, bicarbonate and hydroxide, added to hide developed acidity, are detected by the rosalic acid test, in which a rose-red colour indicates added alkali, and by the alkalinity of ash, which is raised above that of normal milk.
Preservatives added illegally to extend keeping quality include formalin, hydrogen peroxide, boric acid and borax, salicylic and benzoic acids. Formalin is detected by the Hehner test, in which concentrated sulphuric acid containing a trace of ferric chloride is run under the milk and a violet ring at the junction indicates formaldehyde, or by the chromotropic acid test. Hydrogen peroxide is detected by the p-phenylenediamine test (blue colour) or the vanadium pentoxide test (pink to red). Boric acid and borax turn turmeric paper red, the colour turning greenish blue with ammonia. Detergents, which may enter milk from poor rinsing or be added deliberately, especially in synthetic milk, are detected with methylene blue and chloroform: anionic detergents form a blue complex that moves into the chloroform layer. Synthetic milk, a mixture of water, urea, detergent, vegetable oil, sugar and sometimes skim milk powder, has been reported in parts of South Asia; it is detected by combining tests for urea, detergent, vegetable fat and abnormal freezing point and composition. Vegetable fat in milk is detected in the extracted fat by the tests described in chapter 7. Adulteration with milk of a cheaper species, such as cow milk in buffalo milk or cheese, is detected immunologically, for example by an antiserum test against cow milk proteins, by electrophoresis of the caseins, or by DNA-based tests. Melamine, added to raise apparent protein in nitrogen-based tests, requires instrumental methods such as liquid chromatography-mass spectrometry. Many qualitative tests are compiled in national manuals such as the FSSAI Manual of Methods of Analysis.
Ghee and butter fat are adulterated with cheaper fats: vegetable oils and hydrogenated vegetable fat (vanaspati), animal body fats such as tallow and lard, and in some cases mineral oil. Because milk fat is chemically distinctive, its identity is tested by analytical constants, whose typical values are given in Chemistry of Dairy Products. The butyro-refractometer (BR) reading, taken at 40 degrees Celsius, measures refractive index on an arbitrary scale; ghee reads about 40 to 45, while most vegetable oils read higher. The Reichert-Meissl (RM) value is the number of millilitres of 0.1 normal alkali needed to neutralise the steam-volatile, water-soluble fatty acids from 5 grams of fat; it reflects butyric acid and is high for milk fat and near zero for most other fats. The Polenske value is the corresponding titre for the steam-volatile, water-insoluble acids collected on the filter and dissolved in alcohol; it is low for ghee but high for coconut and palm kernel oils. The Kirschner value measures butyric acid more specifically. In the procedure, the fat is saponified with glycerol and sodium hydroxide, acidified, and distilled under standard conditions; the distillate is filtered, the filtrate titrated for the RM value and the filter washings for the Polenske value. The saponification value, milligrams of potassium hydroxide to saponify 1 gram of fat, equals (blank titre minus sample titre) times normality of acid times 56.1 divided by grams of fat; it is high for milk fat because of short-chain acids. The iodine value measures unsaturation and is low for milk fat.
Constants overlap when adulteration is partial or when blends are designed to imitate milk fat, so sterol tests and modern methods are used to confirm. Milk fat contains cholesterol but no significant phytosterols, while vegetable fats contain phytosterols such as beta-sitosterol and no significant cholesterol, as noted in Milk Lipids. In the phytosterol acetate test, sterols are isolated from the unsaponifiable matter, converted to acetates and recrystallised; the melting point of the acetates is measured, and a value above about 117 degrees Celsius indicates phytosterols, because cholesteryl acetate melts lower, near 114 to 115 degrees Celsius. Gas chromatography of sterols detects small additions of vegetable fat more sensitively. The Baudouin test detects sesame oil: furfural and hydrochloric acid give a crimson colour; in India, sesame oil was at one time required in vanaspati as a marker, so a positive Baudouin test in ghee indicated vanaspati. Animal body fats contain cholesterol and little butyric acid, so they lower RM and saponification values without giving phytosterols; their detection relies on constants, opacity or solidification tests, and on gas chromatography of fatty acids and triglycerides, including the position of palmitic acid in lard. Gas chromatography of fatty acid methyl esters gives the full fatty acid profile, in which butyric acid content and ratios between fatty acids characterise milk fat, and triglyceride analysis by carbon number, used in the European Union reference method for foreign fats, detects small proportions of most foreign fats. Because natural constants vary with species, season and region, legal limits in some countries are set regionally.
Residues of veterinary drugs in milk, mainly antibiotics used to treat mastitis, cause allergic reactions in sensitive consumers, contribute to antimicrobial resistance and inhibit starter cultures in cheese and yoghurt making. Maximum residue limits (MRLs) are set by Codex and national authorities; for benzylpenicillin, for example, the Codex and European Union MRL in milk is 4 micrograms per kilogram. Testing proceeds in tiers. Microbial inhibitor tests are broad screens: milk is added to agar seeded with spores of a sensitive thermophilic organism, Geobacillus stearothermophilus (formerly Bacillus stearothermophilus var. calidolactis), with a pH indicator such as bromocresol purple, and incubated; if no inhibitor is present the organism grows and produces acid, changing the colour from purple to yellow, while inhibitors keep the medium purple. These tests detect many beta-lactams at or near their MRLs but are less sensitive to some other classes. Receptor-based and immunological rapid tests, often in lateral-flow strip format, detect specific classes such as beta-lactams and tetracyclines within minutes and are widely used on tankers before unloading. ELISA kits quantify specific residues. Positive screening results are confirmed and quantified by liquid chromatography-tandem mass spectrometry (LC-MS/MS), which identifies individual compounds at very low levels. Natural inhibitors in milk, such as lysozyme and lactoferrin in colostrum and abnormal milk, can cause false positives in microbial tests; heating the sample inactivates them.
Pesticide residues reach milk through feed, water and treatment of animals and premises. The persistent organochlorine insecticides, such as DDT and hexachlorocyclohexane isomers, are fat-soluble and accumulate in milk fat, so they are analysed in the extracted fat and limits are often expressed on a fat basis; although banned or restricted in most countries, they persist in the environment. Organophosphates, carbamates, pyrethroids and other modern pesticides are less persistent. Analysis involves extraction of the residues, clean-up to remove fat and other interferences (for example by gel permeation chromatography or dispersive solid-phase extraction in the QuEChERS approach) and determination by gas chromatography with electron-capture detection for organochlorines, nitrogen-phosphorus or flame photometric detection for organophosphates, and increasingly by gas or liquid chromatography with tandem mass spectrometry for multi-residue analysis. Heavy metals of concern include lead, cadmium, mercury and arsenic. They are determined after digestion of the sample, usually by microwave-assisted acid digestion, by atomic absorption spectrometry with graphite furnace atomisation for lead and cadmium, cold-vapour atomic absorption for mercury and hydride generation for arsenic, or by inductively coupled plasma mass spectrometry for many elements together. Codex sets a maximum level for lead in milk of 0.02 milligrams per kilogram, and national limits are similar or set separately. Contamination control in the laboratory, using acid-washed plastics and reagent blanks, is essential at these low levels.
Aflatoxin M1 is the hydroxylated metabolite of aflatoxin B1, the mycotoxin produced by Aspergillus flavus and Aspergillus parasiticus on feeds such as maize, groundnut cake and cottonseed cake stored in warm, humid conditions. When a dairy animal eats feed contaminated with aflatoxin B1, it converts part of it in the liver to aflatoxin M1, which is excreted in milk; the carry-over is commonly a few per cent of the ingested B1, and M1 appears in milk within hours and largely disappears within a few days of removing the contaminated feed. Aflatoxin M1 is considered a possible human carcinogen and is of particular concern for infants. It is stable to pasteurisation, UHT treatment and drying, and because it binds to casein it is concentrated in cheese relative to the milk from which the cheese was made. Legal limits differ: the European Union sets 0.050 micrograms per kilogram for milk (and a lower limit for infant foods), whereas Codex, the United States and India set 0.5 micrograms per kilogram. Control depends on preventing mould growth and aflatoxin formation in feed by drying and proper storage, testing feeds, and using binders such as hydrated sodium calcium aluminosilicate in feed, as described in Milk Composition and Physical Properties and veterinary nutrition subjects.
Methods for aflatoxin M1 must reach very low concentrations, below the EU limit of 0.050 micrograms per kilogram for many purposes. The reference method (ISO 14501 / IDF 171) uses immunoaffinity column clean-up followed by high-performance liquid chromatography with fluorescence detection. The milk is centrifuged to remove fat, passed through a column containing antibodies specific for aflatoxin M1, which binds it, the column is washed, and the toxin is eluted with a solvent and separated by reversed-phase HPLC; aflatoxin M1 fluoresces naturally, so fluorescence detection is very sensitive and selective. LC-MS/MS is increasingly used, especially for confirmation and for simultaneous determination of several mycotoxins. For screening, enzyme-linked immunosorbent assays (ELISA) in competitive format give quantitative results for batches of samples within about an hour, and lateral-flow strip tests give qualitative or semi-quantitative results in minutes at collection centres and plant reception. Thin-layer chromatography with fluorescence under ultraviolet light was the classical method and is still described in some manuals but is less sensitive and less precise. Screening tests are designed to give few false negatives at the legal limit; positive or borderline results are confirmed by HPLC or LC-MS/MS. Because aflatoxin M1 is distributed unevenly between fat and skim and binds to protein, sample preparation must follow the validated method exactly, and certified reference materials of milk powder containing known amounts of aflatoxin M1 are used to check accuracy.
Spectroscopic methods measure the interaction of electromagnetic radiation with matter. In ultraviolet-visible (UV-Vis) absorption spectroscopy, the absorbance of a solution at a chosen wavelength is related to concentration by the Beer-Lambert law: absorbance equals molar absorptivity times path length times concentration (A = εbc). If a compound with molar absorptivity 15,000 litres per mole per centimetre gives an absorbance of 0.45 in a 1 centimetre cell, its concentration is 0.45 divided by 15,000, or 30 micromoles per litre. In practice, concentration is usually read from a calibration curve of standards, which also shows the linear range, since the law holds only at moderate absorbances. Many dairy tests are colorimetric applications of UV-Vis: the yellow p-nitrophenol of the phosphatase test, colorimetric phosphorus, lactose and urea methods, the thiobarbituric acid test for oxidation, and colour reactions for adulterants. A colorimeter uses filters to select a band of wavelengths, while a spectrophotometer uses a monochromator. Infrared spectroscopy measures vibrations of chemical bonds; its application to milk analysers is described in chapter 5, and FTIR and NIR are also used to identify fats, detect adulteration and characterise powders. Fluorescence spectroscopy, in which molecules absorb light and emit at a longer wavelength, is highly sensitive and is used for riboflavin, aflatoxin M1 after HPLC separation, the fluorimetric phosphatase test and the detection of Maillard products.
Atomic spectroscopy determines elements. In atomic absorption spectrometry (AAS), the sample, usually after acid digestion or dry ashing, is atomised in a flame or graphite furnace, and the absorption of light from a hollow cathode lamp made of the element of interest is measured; each element absorbs at its own characteristic wavelengths, so the method is selective. Flame AAS, using an air-acetylene or nitrous oxide-acetylene flame, suits elements at milligram per litre levels such as calcium, magnesium, zinc, iron and copper; graphite furnace AAS, which atomises a small sample electrically, reaches microgram per litre levels for lead, cadmium and other trace elements. Chemical interferences must be controlled: phosphate in milk forms refractory compounds with calcium in the flame and lowers its signal, so lanthanum or strontium salts are added as releasing agents. Flame photometry (flame atomic emission) measures the light emitted by excited atoms of easily excited elements, chiefly sodium (at about 589 nanometres) and potassium (at about 766 nanometres), and sometimes lithium and calcium; it is simple and cheap and is used for sodium and potassium in milk, which change in mastitis. Inductively coupled plasma optical emission spectrometry (ICP-OES) and ICP mass spectrometry (ICP-MS) excite or ionise atoms in an argon plasma and determine many elements simultaneously, ICP-MS at very low concentrations. All atomic methods require matrix-matched standards or standard additions, reagent blanks and certified reference materials.
Chromatography separates the components of a mixture by distributing them between a stationary phase and a mobile phase; components that interact more strongly with the stationary phase move more slowly. Separation depends on differences in adsorption, partition, ion exchange, size or affinity. In thin-layer chromatography (TLC), the stationary phase is a thin layer of silica or another adsorbent on a plate; samples are spotted near one edge and the plate is developed in a closed tank with a solvent that rises by capillary action. Components are located by their colour, by fluorescence under ultraviolet light or by spraying a reagent, and are characterised by the retardation factor, Rf, the distance travelled by the component divided by the distance travelled by the solvent front: a spot 4.2 centimetres from the origin with the solvent front at 7.0 centimetres has an Rf of 0.60. TLC is simple and cheap and is used for sterols, for aflatoxins in older methods, for separating lipid classes and for screening dyes and preservatives, but it is mainly qualitative or semi-quantitative. High-performance TLC with densitometric scanning improves quantification. Column chromatography on a larger scale is used for clean-up of extracts, for example solid-phase extraction cartridges and immunoaffinity columns, before instrumental analysis.
Gas chromatography (GC) separates volatile compounds carried by an inert gas, such as helium, hydrogen or nitrogen, through a long, narrow capillary column coated with a liquid stationary phase, in an oven whose temperature is programmed to rise. Compounds are identified by their retention times compared with standards and are quantified by peak areas, usually with an internal standard. Detectors include the flame ionisation detector (FID), general for organic compounds and standard for fatty acids; the electron-capture detector (ECD), very sensitive to halogenated compounds such as organochlorine pesticides; nitrogen-phosphorus and flame photometric detectors for organophosphates; and the mass spectrometer, which identifies compounds by their mass spectra. Dairy applications include the fatty acid profile of milk fat after conversion to fatty acid methyl esters (FAME), sterols, triglycerides by carbon number, flavour volatiles by headspace sampling, and pesticide residues. High-performance liquid chromatography (HPLC) pumps a liquid mobile phase at high pressure through a column packed with fine particles; reversed-phase columns with octadecylsilyl (C18) packing and water-organic solvent mixtures are the most common. Detectors include ultraviolet-visible absorbance, fluorescence, refractive index (for sugars such as lactose, which lack chromophores), evaporative light-scattering and mass spectrometry. HPLC is used for lactose and oligosaccharides, organic acids, whey proteins and caseins, vitamins, aflatoxin M1, antibiotic residues, furosine and preservatives. Ion-exchange chromatography separates amino acids and charged molecules, and size-exclusion chromatography separates proteins and peptides by size.
Electrophoresis separates charged molecules, chiefly proteins, by their movement in an electric field through a gel. In sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE), proteins are heated with SDS, a detergent that binds to them in proportion to their length and gives them a uniform negative charge per unit mass, and usually with a reducing agent such as beta-mercaptoethanol that breaks disulphide bonds; the proteins then move through the gel according to molecular mass alone, smaller ones faster. Their masses are estimated against marker proteins, and bands are stained with Coomassie blue or silver. In milk, SDS-PAGE separates the caseins (about 19 to 25 kilodaltons), beta-lactoglobulin (about 18 kilodaltons), alpha-lactalbumin (about 14 kilodaltons), serum albumin, immunoglobulin chains and lactoferrin. Alkaline urea-PAGE, in which urea dissociates the caseins and they separate by charge, resolves the individual caseins and their breakdown products and is widely used to follow proteolysis in cheese and UHT milk. Isoelectric focusing separates proteins by isoelectric point in a pH gradient and resolves genetic variants and species-specific proteins; the European Union reference method for detecting cow milk in cheeses made from sheep, goat or buffalo milk uses isoelectric focusing of gamma-caseins after plasmin digestion. Capillary electrophoresis performs similar separations in narrow capillaries with automated detection and good quantification.
Immunoassays use the specific binding of antibodies to their target (antigen) to detect proteins, toxins and residues. In an enzyme-linked immunosorbent assay (ELISA), one partner is immobilised on a plastic plate, the sample is added, and binding is revealed by an enzyme, commonly horseradish peroxidase or alkaline phosphatase, linked to an antibody or antigen, which converts a substrate such as tetramethylbenzidine into a coloured product measured photometrically. Large molecules with several epitopes, such as milk proteins measured as allergens or for species identification, are measured by sandwich ELISA, in which the analyte is captured between two antibodies and the colour is directly proportional to concentration. Small molecules such as aflatoxin M1 and antibiotics, which have only one epitope, are measured by competitive ELISA, in which analyte in the sample competes with a labelled or immobilised analyte for a limited number of antibody sites, so the colour is inversely proportional to concentration. Lateral-flow (immunochromatographic) tests apply the same principles on a strip: the sample flows along a membrane carrying labelled antibodies, typically on gold nanoparticles, past a test line and a control line. In the competitive format used for residues and aflatoxin M1, a visible test line means a negative result and a missing or faint line a positive result; the control line must always appear for the test to be valid. Immunoassays are fast and sensitive but can cross-react with related compounds, so positive results are confirmed by chromatographic methods. DNA-based methods such as the polymerase chain reaction complement immunoassays for species identification.
A laboratory must show that each method it uses is fit for its purpose, by validation for new or modified methods and by verification that it can perform a standard method as published. The main performance characteristics are selectivity (the ability to measure the analyte without interference from other components); the working range and linearity, shown by a calibration curve; the limit of detection (LOD), the lowest concentration that can be reliably distinguished from zero, commonly estimated as 3.3 times the standard deviation of the blank or of the response at low level divided by the slope of the calibration curve; and the limit of quantification (LOQ), the lowest concentration that can be measured with acceptable precision and accuracy, commonly 10 times the standard deviation divided by the slope. With a standard deviation of 0.002 absorbance units and a slope of 0.05 units per milligram per litre, the LOD is about 0.13 and the LOQ 0.40 milligram per litre. Trueness is the closeness of the mean result to the true value, assessed by analysing certified reference materials or by recovery of added analyte: if a sample containing 0.50 units is spiked with 1.00 unit and 1.45 units are found, recovery is (1.45 minus 0.50) divided by 1.00, or 95 per cent. Precision is the closeness of repeated results, expressed as repeatability (same analyst, equipment and short time) and reproducibility (different laboratories); the repeatability limit, the maximum expected difference between two results, is about 2.8 times the repeatability standard deviation. Robustness and measurement uncertainty complete the picture.
Reference materials provide the link between a laboratory's results and true values. Certified reference materials (CRMs) have property values established by a recognised body with a stated uncertainty and traceability, for example milk powders certified for fat, protein, trace elements or aflatoxin M1; they are used to validate methods and check accuracy. In-house reference materials, such as stable pilot milks, are used daily for quality control, and results are plotted on control charts, such as Shewhart charts with warning limits at two standard deviations and action limits at three, so that drift or loss of control is noticed. Blanks, duplicate analyses and spiked samples are run with each batch. Proficiency testing (PT) is an external check in which a provider sends the same samples to many laboratories and compares their results with an assigned value. Performance is commonly expressed as a z-score: z equals the laboratory's result minus the assigned value, divided by the standard deviation for proficiency assessment. A z-score with an absolute value of 2 or less is satisfactory, between 2 and 3 questionable, and 3 or more unsatisfactory. A laboratory reporting 3.62 per cent fat against an assigned value of 3.50 per cent with a standard deviation of 0.05 per cent obtains z = 2.4, a questionable result that requires investigation. Accreditation to ISO/IEC 17025 requires validated methods, quality control, participation in proficiency testing, trained staff, calibrated equipment and documented procedures, and many regulators require food testing laboratories to be accredited. PT providers themselves are accredited to ISO/IEC 17043.
A student uses this subject in every dairy chemistry and quality control practical. You will sample milk correctly, run the platform tests, determine fat by Gerber and compare with Röse-Gottlieb, digest and distil Kjeldahl samples, dry samples for total solids, test spiked samples for common adulterants, determine ghee constants and run the phytosterol acetate test, and use a spectrophotometer, a chromatograph and an electrophoresis tank, recording results in a laboratory notebook to the standard an auditor would expect. Learn each test's principle, reagents, positive result and limitations, and practise the calculations, because food analyst and PG examinations ask them directly.
Five years into work as a reception chemist, quality manager or food analyst, you run a laboratory: calibrating infrared analysers against reference methods, keeping control charts, investigating a questionable proficiency-test z-score, choosing rapid tests for antibiotics and aflatoxin M1 at chilling centres, confirming positives by HPLC or LC-MS/MS, and preparing analysis reports for enforcement cases that must survive legal challenge. You also decide which adulterant tests to run when a supplier's composition or freezing point looks suspicious, and you train collection-centre staff to run platform tests consistently.
A senior laboratory head or quality director with twenty-five years of experience builds and maintains an accredited laboratory system, chooses instruments and methods, sets testing frequencies and sampling plans for a procurement network, negotiates specifications and dispute procedures with suppliers and customers, represents the organisation in standard-setting committees and regulatory audits, and responds to food safety incidents such as melamine, aflatoxin or residue crises with rapid, defensible testing. Such a person also budgets for instruments, staff and proficiency schemes and decides which tests to outsource.
A researcher with long experience develops and validates new analytical methods: spectroscopic fingerprinting for authenticity and adulteration, rapid biosensors and lateral-flow tests for residues and toxins, mass-spectrometric methods for proteins, peptides and contaminants, and chemometric models for infrared data. Such a person runs interlaboratory studies, produces reference materials, advises regulators on limits and methods, and evaluates the measurement uncertainty that underlies legal decisions, and trains analysts in the statistics of method comparison and validation. Methods are judged by whether they work in ordinary laboratories, not only in the research setting.
This manual explains the principles, main reagents and interpretation of each method; it does not reproduce full step-by-step procedures, reagent preparations or legal limits, which must be taken from the current official method (FSSAI, BIS, ISO/IDF, AOAC) and regulation. Limits quoted, such as MRLs and aflatoxin M1 levels, are examples and may change. Microbiological methods for milk and products are covered in the dairy microbiology subjects, and sensory evaluation in its own subject. Advanced techniques such as nuclear magnetic resonance, isotope-ratio mass spectrometry for origin, and biosensor development are mentioned only briefly. Ghee constants vary with species, season and region and are given as typical ranges.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.