Dairy Science & Technology · Dairy Chemistry · also Chemistry of Milk, Physical Chemistry of Milk, Dairy Chemistry I
What milk is made of, how composition varies, and the physical properties used to test and process it.
In the exams: ICAR AICE subject 34 Unit 1; ICAR AIEEA-PG group 16 (composition and physico-chemical properties of milk and colostrum); ASRB NET Dairy Chemistry
Milk is the secretion of the mammary gland that feeds the newborn. To a dairy technologist it is also a raw material whose value, keeping quality and behaviour in the plant are set by what it contains and how those components are arranged. This subject describes the gross composition of milk, how it differs between species and between animals, what colostrum and abnormal milks look like, and the physical properties (density, freezing point, acidity, buffering, viscosity, surface tension, conductivity, refractive index, thermal behaviour, redox potential and colour) that are measured at reception, used to detect adulteration, and relied on when milk is heated, cooled, pumped and concentrated.
The subject matters in practice because almost every routine decision in a dairy uses it. A reception chemist accepts or rejects a tanker on fat, solids-not-fat, acidity and freezing point. A milk payment system prices milk on fat and SNF or on fat and protein. A process engineer sizes a heat exchanger with the specific heat and viscosity of milk, and an analyst explains why buffalo milk gives a whiter, firmer paneer than cow milk. Numbers quoted in this manual are typical textbook values; actual composition varies with species, breed, feed, season, stage of lactation and udder health, and legal minima differ between countries.
The subject is examined as Unit 1 of ICAR AICE-JRF/SRF subject 34 (Dairy Chemistry), within group 16 of ICAR AIEEA-PG (composition and physico-chemical properties of milk and colostrum), in ASRB NET Dairy Chemistry and in food analyst and food safety officer papers, which test reception tests and adulteration. Outside India the same material appears in food science degree examinations, dairy technician and apprenticeship qualifications, and in GATE food technology as part of food chemistry. Expect recall of composition figures and constants, calculation of SNF from lactometer readings, added water from freezing point and titratable acidity, and reasoning about why a property changes in mastitic, watered or souring milk.
Milk is not a simple solution. It is at once an oil-in-water emulsion, a colloidal suspension and a true solution. Fat is dispersed as globules, mostly 0.1 to 15 micrometres across with a volume-weighted mean of about 3 to 4 micrometres in cow milk, each surrounded by the milk fat globule membrane (MFGM), a layer of phospholipids and proteins derived from the secretory cell. Casein is present as casein micelles, roughly 50 to 500 nanometres across with an average near 150 to 200 nanometres, held together by colloidal calcium phosphate; there are of the order of 10 to the power 14 or 15 micelles per millilitre. Whey (serum) proteins are dispersed as individual molecules a few nanometres in size. Lactose, most of the soluble salts, non-protein nitrogen compounds such as urea, and water-soluble vitamins are in true solution in the milk serum. Because the phases differ so much in size, they can be separated: fat by gravity or centrifugation (cream separation), casein by acid or rennet, whey proteins by membranes, and lactose and salts by crystallisation and ion exchange. The partition of each component between phases also explains many properties: the white colour comes from light scattering by micelles and globules, the viscosity mainly from micelles and fat, and the freezing point from the dissolved molecules and ions.
Typical cow milk contains about 87 to 87.5 per cent water and 12.5 to 13 per cent total solids (TS). Of the solids, fat is the most variable component, commonly 3.5 to 4.5 per cent in herd milk but ranging from below 3 to above 6 per cent between individual cows and breeds. Total protein is about 3.2 to 3.5 per cent, of which casein forms roughly 80 per cent (about 2.6 per cent of milk) and whey proteins roughly 20 per cent. Lactose, the main carbohydrate, is about 4.6 to 4.9 per cent and is the least variable major constituent because it largely sets the osmotic pressure of milk. Ash (minerals) is about 0.7 per cent; calcium is about 1.2 grams per litre and phosphorus about 0.9 to 1.0 gram per litre, a large part of each held in the micelles. Solids-not-fat (SNF) is total solids minus fat, typically 8.5 to 9 per cent in cow milk; it is the sum of protein, lactose, ash and minor constituents. Minor constituents include citrate (about 1.6 to 1.8 grams per litre), non-protein nitrogen (about 5 per cent of total nitrogen, roughly half of it urea), dissolved gases, vitamins, enzymes and somatic cells. Crude protein is conventionally calculated as Kjeldahl nitrogen multiplied by 6.38 for milk. Legal standards for fat and SNF are set nationally and differ by country, by species and sometimes by region within a country, so a value that is acceptable in one jurisdiction may fail in another.
Composition drives the price and the use of milk. Many milk payment schemes pay on two axes, fat and SNF (common in South Asia) or fat and protein (common in Europe, Oceania and North America), because these components determine the yield of butter, cheese, powder and condensed products. Standardisation adjusts fat and sometimes protein to a target by blending with skim milk or cream, and Pearson's square or mass-balance equations are used to calculate the quantities. Composition also predicts processing behaviour: high casein raises cheese yield, high whey protein raises fouling in heat exchangers, high lactose and low protein favour browning in concentrated products, and a high ratio of fat to SNF affects the stability of evaporated milk. At reception, gross composition is measured quickly by infrared analysers calibrated against reference methods: the Gerber or Röse-Gottlieb methods for fat, Kjeldahl for nitrogen and gravimetric drying for total solids. In smaller collection centres in India and elsewhere, fat is measured by Gerber or by ultrasonic and electronic analysers, and SNF is estimated from the lactometer reading and fat using formulas described in chapter 5. Because each estimate rests on assumptions about normal milk, an abnormal result is a reason to check the milk, not merely to adjust the price.
Milk composition differs between species in ways that matter for products. Buffalo milk, the main milk of South Asia and important in Egypt and Italy, typically contains 6 to 8 per cent fat, about 4 to 4.5 per cent protein, about 4.8 to 5 per cent lactose and 9.5 to 10 per cent SNF. It has more casein and more colloidal calcium and phosphate than cow milk, larger fat globules on average, and higher titratable acidity. It is whiter because the buffalo converts beta-carotene to vitamin A and secretes little carotene into milk fat. These features explain why buffalo milk gives firmer curd, higher yields of paneer, khoa and mozzarella, and whiter ghee. Goat milk has a gross composition close to cow milk in fat and protein but somewhat less lactose; its fat globules are smaller on average, it lacks the cryoglobulin-type agglutinin that clusters cow fat globules during cold creaming, so it creams slowly, and its fat is richer in short and medium-chain fatty acids (caproic, caprylic and capric), which give the characteristic goaty flavour when lipolysed. It is also white for the same reason as buffalo milk. Sheep milk is the richest of the common dairy species, with about 6 to 8 per cent fat, about 5.5 to 6 per cent protein and SNF near 11 to 12 per cent, so its cheese yield per litre is much higher than that of cow milk.
Camel milk is produced in arid regions of Africa, the Middle East and South Asia. Its fat content varies widely, from about 2 to over 5 per cent, with protein about 3 to 3.5 per cent and lactose about 4.5 to 5 per cent. It contains no beta-lactoglobulin, a low proportion of kappa-casein and larger casein micelles, so it coagulates poorly with calf rennet and makes cheese only with special coagulants and calcium addition. It is notably richer in vitamin C than cow milk, which is valuable in diets with few fresh vegetables. Human milk differs from the milk of ruminants in being low in protein (about 0.9 to 1.2 per cent in mature milk) and ash (about 0.2 per cent) but high in lactose (about 7 per cent). Its proteins are mainly whey proteins, with a casein to whey ratio that changes during lactation and is often quoted near 40 to 60 in mature milk, compared with about 80 to 20 in cow milk. Human milk has no beta-lactoglobulin, much more lactoferrin and lysozyme than cow milk, secretory IgA as the main immunoglobulin, and a large and varied fraction of oligosaccharides. These differences are why cow milk must be modified, by diluting protein and minerals, adding lactose or other carbohydrates and adjusting whey protein content, before it can be used as the base for infant formula.
Species differences have direct consequences for technology, law and nutrition. Because buffalo and cow milk differ so much in fat and SNF, many countries, including India, set separate legal standards for cow milk, buffalo milk and mixed milk, and price them differently. The ratio of casein to fat governs cheese yield and the fat in dry matter of the cheese, so cheesemakers using buffalo or sheep milk standardise differently from those using cow milk. The colour of fat helps detect admixture: cow ghee and butter are yellow from beta-carotene, buffalo products white. Species identification in mixed or fraudulently labelled milk and cheese is done by electrophoresis or chromatography of the caseins and whey proteins, by immunological methods and by DNA-based tests; the aim is often to detect cheaper cow milk in buffalo, sheep or goat products. Nutritionally, the low protein and mineral content of human milk suits the slow growth and immature kidneys of the human infant, while the high protein and ash of ruminant milks suits fast-growing calves, lambs and kids. When comparing figures between sources, remember that composition varies with breed, feed and stage of lactation within each species, so published averages for one species overlap at their edges with those of another.
Within one species, composition varies between breeds, between animals of the same breed and even within one milking. Among cattle, Jersey and Guernsey milk is richer in fat and protein (Jersey fat commonly near 5 per cent) than Holstein-Friesian milk (fat commonly 3.5 to 4 per cent), while Holsteins give much more volume. Indian Bos indicus breeds such as Sahiwal, Gir and Red Sindhi generally give milk of moderate to high fat. Fat varies most between animals, protein less and lactose least. Genetic variants of the milk proteins, inherited from the parents, also alter composition and processing properties and are covered in the subject on milk proteins. Within a milking, the first milk drawn (fore-milk) is low in fat, often 1 to 2 per cent, and the last milk (strippings) is rich, often 6 to 10 per cent, because fat globules are held back in the alveoli until the end of milk let-down. Incomplete milking therefore lowers the fat of the milk delivered and raises the fat of the next milking. Long and unequal milking intervals reduce the fat content of the larger milking and raise that of the shorter one. Fat content also tends to fall slightly and SNF to fall as the animal ages, partly because udder infections become more common in older animals.
Feeding affects fat more than protein and protein more than lactose. Rations low in physically effective fibre or high in rapidly fermentable starch, or diets with high levels of unsaturated oils, reduce milk fat, a condition called milk fat depression; certain biohydrogenation intermediates produced in the rumen, notably trans-10, cis-12 conjugated linoleic acid, inhibit fat synthesis in the mammary gland. Underfeeding energy lowers protein content. Pasture feeding increases unsaturated fatty acids and carotene, making butter softer and yellower. Season affects composition through feed, temperature and stage of lactation: in many temperate and tropical regions, fat and SNF are lowest in hot summer months and highest in winter, and heat stress lowers yield and solids. Stage of lactation has a large and predictable effect. Colostrum in the first days is very rich in protein and solids. After that, fat and protein fall to their lowest values around peak yield, roughly the second month, and then rise steadily as yield falls, so late-lactation milk is richer in fat and protein. Very late in lactation, milk may become saltier, with more sodium and chloride, less lactose, more plasmin activity and poorer rennet coagulation, behaving rather like mild mastitic milk.
Mastitis, inflammation of the udder usually caused by bacterial infection, is the most important cause of abnormal composition in commercial herds. Damage to the secretory cells and the tight junctions between them lowers the synthesis of lactose, casein and fat and lets blood components leak into milk. The typical changes are: lactose falls; casein falls as a share of total protein while serum albumin, immunoglobulins and other blood proteins rise, so total protein may change little; sodium and chloride rise while potassium falls; pH rises towards or above 6.8; electrical conductivity rises; somatic cell count (SCC), mainly leucocytes, rises from a normal level of under about 100,000 to 200,000 cells per millilitre to hundreds of thousands or millions; and enzymes such as plasmin, catalase and N-acetyl-beta-D-glucosaminidase increase. Because the fall in lactose is roughly balanced osmotically by the rise in sodium and chloride, the freezing point of mastitic milk changes little. These changes make mastitic milk poor for processing: lower cheese yield, weaker curd, more proteolysis and lipolysis during storage, shorter shelf life of pasteurised and UHT milk, and off-flavours, including a salty taste. Somatic cell limits for bulk milk are set nationally and differ between countries. Screening tests such as the California Mastitis Test (CMT), electrical conductivity measurement in milking systems and laboratory SCC counting detect affected quarters and herds.
Colostrum is the first secretion of the mammary gland after parturition. It differs sharply from normal milk and changes rapidly over the first few days, approaching normal composition in about five days in the cow. First-milking cow colostrum typically contains about 22 to 25 per cent total solids and about 14 per cent protein or more, much of it immunoglobulin, mainly IgG1; immunoglobulin content in good colostrum is often 50 grams per litre or more, though it varies widely between cows. Lactose is low, about 2.5 to 3 per cent, ash is higher, about 1 per cent, and fat is variable. Colostrum is rich in vitamin A and carotene (so it is deep yellow), in other fat-soluble vitamins, and in lactoferrin, growth factors and leucocytes. Its density is higher than that of milk, and its titratable acidity is high because of its protein and phosphate content, not because of bacterial acid. Because whey proteins and immunoglobulins denature on heating, colostrum coagulates when boiled, which is why it is unsuitable for normal processing and why most legal definitions require milk to be free from colostrum. For the calf, colostrum is essential: ruminant placentas do not pass antibodies, so the calf depends on absorbing colostral immunoglobulins through the gut in the first day of life. Colostrum quality on farm is estimated with a colostrometer, which reads specific gravity, or a Brix refractometer. Surplus colostrum is used in traditional foods in South Asia and as a raw material for colostrum supplements.
Abnormal milk is milk whose composition, appearance, flavour or safety departs from normal. The main types are colostral milk, discussed above; late-lactation milk, which is saltier, with lower lactose and casein and poorer coagulation; mastitic milk, described in chapter 3; and salty milk from individual animals with high chloride. Milk containing blood is pink or shows a red sediment after injury or udder bleeding at calving. Ropy or slimy milk results from growth of exopolysaccharide-producing bacteria. Milk may carry off-flavours from feed (onion, garlic, silage), from absorption of odours, from oxidation (cardboard or metallic flavour caused by copper and light), from lipolysis (rancid, butyric flavour), from sunlight (light-induced flavour) or from bacterial growth (sour, malty, fruity). Milk from animals treated with antimicrobials or other veterinary drugs, before the withdrawal period has passed, is abnormal in a safety sense: residues can cause allergic reactions, select resistant bacteria and inhibit starter cultures in cheese and yoghurt making, so it must be kept out of the supply. Contaminants such as aflatoxin M1 (from mouldy feed), pesticides and heavy metals also make milk unfit. Abnormal milk is detected at farm and reception by smell and appearance, the clot-on-boiling and alcohol tests, conductivity, somatic cell counts and rapid residue tests.
Density is mass per unit volume; specific gravity is the ratio of the density of milk to that of water at a stated temperature. The density of cow milk at 20 degrees Celsius is usually about 1.028 to 1.032 grams per millilitre and that of buffalo milk somewhat higher, because buffalo milk carries more SNF. Density reflects the densities and proportions of the components: milk fat is lighter than water (about 0.92 grams per millilitre at 20 degrees Celsius), while the non-fat solids are much heavier (about 1.6). So adding fat or skimming milk change density in opposite directions: fat lowers density, SNF raises it. Skim milk is denser than whole milk (about 1.035), cream is lighter, and adding water brings density towards 1.000. Density changes with temperature, mainly because water expands, and also because milk fat crystallises and melts slowly. Freshly drawn milk cooled quickly can show a lower density for some hours until fat crystallisation and gas loss are complete (the Recknagel phenomenon), so lactometer readings should not be taken on milk just after milking or rapid cooling. Density is needed in the plant to convert between volume and mass, since milk is often bought by litre and processed by kilogram.
A lactometer is a hydrometer designed for milk. The Quevenne lactometer is graduated from 15 to 45, its reading being (specific gravity minus 1) multiplied by 1000, and it is calibrated at 60 degrees Fahrenheit (15.6 degrees Celsius). A reading taken at another temperature is corrected by adding 0.1 for each degree Fahrenheit above 60 and subtracting 0.1 for each degree below, within a working range of about 50 to 70 degrees Fahrenheit. A Quevenne reading of 29 at 64 degrees Fahrenheit therefore gives a corrected lactometer reading (CLR) of 29.4, a specific gravity of 1.0294. In India, BIS lactometers calibrated at 27 degrees Celsius are widely used, with their own correction tables. Readings should be taken on well-mixed milk, after the lactometer has floated free and the milk surface is still, at the top of the meniscus. Normal cow milk usually gives a CLR of about 28 to 32 and buffalo milk somewhat higher. A low reading suggests added water or added fat; a high reading suggests skimming or addition of solids such as sugar, starch, salt or urea. A lactometer reading alone cannot separate watering from creaming, because adding water and adding cream both lower the reading, nor can it detect a combination of skimming and watering that leaves the reading normal. It must be read together with the fat content.
Because density depends on fat and SNF, SNF and total solids can be estimated from the fat content and the CLR. Richmond's formula, used with readings corrected to 60 degrees Fahrenheit, is SNF per cent = CLR divided by 4 plus 0.21 times fat per cent plus 0.14, and total solids per cent = CLR divided by 4 plus 1.21 times fat per cent plus 0.14. For milk with 4.0 per cent fat and a CLR of 29.4, SNF = 7.35 + 0.84 + 0.14 = 8.33 per cent and total solids = 12.33 per cent. For BIS lactometers calibrated at 27 degrees Celsius, Indian practice uses SNF per cent = CLR divided by 4 plus 0.25 times fat per cent plus 0.44; milk with 4.5 per cent fat and a CLR of 28 then gives 7.0 + 1.125 + 0.44 = 8.57 per cent SNF. The two formulas are tied to their own lactometer and temperature and must not be mixed: using the 27 degree formula on a 60 degree Fahrenheit reading gives an answer that is too high. Other formulas, such as Fleischmann's, express total solids in terms of fat and specific gravity. All such formulas assume normal milk; added sugar, salt or urea raise density and give a falsely high SNF, which is why reception laboratories confirm suspicious samples by gravimetric total solids and specific adulterant tests.
The freezing point is the temperature at which ice first forms in milk. It is lower than that of water because dissolved molecules and ions lower the freezing point, a colligative property that depends on the number of dissolved particles, not their kind. In milk, lactose and the soluble salts, especially chloride, account for most of the depression; fat globules and casein micelles, being colloidal, contribute almost nothing. The freezing point of normal cow milk lies in a narrow range, commonly quoted as about minus 0.512 to minus 0.550 degrees Celsius, with an average near minus 0.52 to minus 0.53 degrees Celsius. It is so constant because milk is secreted in osmotic equilibrium with blood, and the osmotic pressure of blood is tightly regulated. When lactose falls, as in mastitis or late lactation, sodium and chloride rise to keep the osmotic pressure, so the freezing point hardly changes. This constancy makes the freezing point the most reliable single test for added water. Older literature and some standards use the Hortvet scale; Hortvet degrees are close to, but not the same as, degrees Celsius, so a limit stated in one scale must not be compared with a measurement in the other without conversion.
The freezing point is measured with a thermistor cryoscope. A small sample is supercooled, freezing is induced by vibration, and the temperature plateau reached as latent heat is released is read as the freezing point. Instruments are calibrated with standard salt solutions, and the reference method is described by ISO and IDF. Adding water raises the freezing point towards zero in proportion to the dilution. If T is the freezing point of genuine milk of that origin and T prime that of the sample, both as positive depressions below zero, the simple estimate of added water is (T minus T prime) divided by T, multiplied by 100. With a base of minus 0.525 and a sample at minus 0.483 degrees Celsius, the added water is 0.042 divided by 0.525, or 8.0 per cent of the sample. A variant used in some official methods multiplies by (100 minus total solids of the sample) instead of 100, which gives a smaller figure (7.0 per cent if the sample has 12 per cent total solids). Questions and reports must state which formula and base value are used. Many authorities do not declare milk watered unless the freezing point is above a threshold such as minus 0.520 or minus 0.525 degrees Celsius, to allow for natural variation.
A freezing point closer to zero than the legal limit usually means added water, whether deliberate or from rinse water left in equipment, but other causes must be considered. Very high-yielding animals, heat stress, high water intake, and some feeding conditions can raise the freezing point slightly; the natural range differs between regions and seasons, so many countries set their base values from surveys of authentic farm milk. Factors that depress the freezing point further, making watering harder to detect, include souring (lactic acid formation from lactose increases the number of dissolved particles), addition of salt, sugar or urea, and late-lactation or mastitic milk with high salt. Heating changes the freezing point only slightly through loss of carbon dioxide and shifts in calcium phosphate, and vacuum treatment raises it a little by removing dissolved gases. The freezing point is therefore always read with titratable acidity, lactometer reading and, where needed, tests for specific adulterants. In milk payment schemes, routine cryoscopy of every supplier's milk discourages watering and identifies faulty cooling or cleaning practices that leave water in tanks and pipelines.
The pH of fresh cow milk at 25 degrees Celsius is about 6.6 to 6.8, slightly acidic, usually near 6.7. Buffalo milk lies in a similar range. Colostrum has a lower pH, around 6.0 to 6.3, because of its high protein content, while mastitic milk has a higher pH, often 6.9 or above, because of the influx of bicarbonate and other blood constituents. pH falls when bacteria ferment lactose to lactic acid, and it is temperature dependent: heating milk lowers its pH because calcium phosphate precipitates and releases hydrogen ions, and cooling raises it, so pH must be reported with the temperature at which it was measured. pH matters in processing because it governs the stability of casein micelles. As pH falls, colloidal calcium phosphate dissolves, the negative charge on micelles is reduced, and at about pH 4.6 casein reaches its isoelectric point and precipitates; this is the basis of acid casein manufacture and of acid-coagulated products such as yoghurt, cottage cheese and paneer, where acid at elevated temperature is used. The heat stability of milk, the rate of rennet coagulation and the activity of enzymes also depend strongly on pH.
Titratable acidity (TA) is the amount of alkali needed to bring a measured quantity of milk to the end point of phenolphthalein, about pH 8.3. It is expressed as per cent lactic acid by convention, even though fresh milk contains almost no lactic acid. One millilitre of 0.1 normal sodium hydroxide neutralises 9 milligrams of lactic acid, so per cent lactic acid equals the volume of alkali in millilitres times its normality times 9, divided by the mass of the sample in grams. If 10 grams of milk need 1.6 millilitres of 0.1 normal alkali, TA is 0.144 per cent; if 10 millilitres are pipetted and the mass is taken as 10.3 grams, the result is 0.140 per cent. Fresh cow milk usually shows 0.13 to 0.15 per cent and buffalo milk somewhat more, about 0.14 to 0.16 per cent or higher. Several countries use other units: degrees Soxhlet-Henkel (millilitres of 0.25 normal alkali per 100 millilitres, fresh milk about 6.5 to 7.5), degrees Thörner (millilitres of 0.1 normal alkali per 100 millilitres, fresh milk about 16 to 18) and degrees Dornic (tenths of a millilitre of N/9 alkali per 10 millilitres, each degree equal to 0.01 per cent lactic acid, fresh milk about 14 to 16). One degree Soxhlet-Henkel equals 2.5 degrees Thörner and 2.25 degrees Dornic.
The titratable acidity of fresh milk is called natural or apparent acidity. It arises from the buffering constituents of milk that react with alkali between the pH of milk and the phenolphthalein end point: casein (about half of the natural acidity), phosphates, citrates, whey proteins and dissolved carbon dioxide. Because it reflects composition, natural acidity is higher in milk with more protein and SNF, such as buffalo milk, late-lactation milk and colostrum, and lower in milk with low SNF, such as watered or mastitic milk. Developed or real acidity is the acidity added by bacterial fermentation of lactose to lactic acid after milking. Total titratable acidity is the sum of the two. For this reason, a single TA value does not prove that milk is fresh or stale: a high-SNF milk may show a high TA without any souring, and watering lowers TA. Reception decisions therefore combine TA with the clot-on-boiling test, which detects milk that will coagulate on heating, and the alcohol test, in which milk mixed with an equal volume of ethanol (commonly 68 per cent, but strengths of 68 to 75 per cent are used) flocculates if its casein is destabilised by acid, by calcium imbalance or by colostrum. TA is also used to follow starter activity in cheese and yoghurt making and to monitor cream ripening.
A buffer resists a change in pH when acid or base is added. Milk is a well-buffered fluid, which is why fresh milk can take up a substantial amount of lactic acid with only a small fall in pH, and why cheese and yoghurt making need a large amount of acid to reach coagulation. The buffering components are those with acid-base groups whose pKa values lie in the range of interest. In milk these are the proteins, especially casein, through the side chains of histidine, glutamic and aspartic acids and the phosphate groups of phosphoserine residues; inorganic phosphate, whose second dissociation has a pKa near 6.8 to 7.2 under milk conditions; citrate, with pKa values near 3.1, 4.8 and 6.4; and carbonate. The relationship between pH and the ratio of conjugate base to acid for each buffer is given by the Henderson-Hasselbalch equation: pH equals pKa plus the logarithm of the ratio of base to acid. A buffer works best within about one pH unit of its pKa, so each component buffers a particular region of the pH scale.
Buffer capacity, or the buffer index, is the amount of strong acid or base needed to change the pH of a given volume by one unit, measured as the slope of a titration curve. When milk is titrated with acid from its natural pH, the buffer index rises to a maximum at about pH 5.0 to 5.5. The main reason is that, as pH falls, colloidal calcium phosphate in the casein micelles dissolves, releasing phosphate ions that take up hydrogen ions; casein side chains and citrate also contribute in this region. When acidified milk is back-titrated with alkali, the curve does not retrace its path. The buffering maximum on the way back is shifted to a higher pH, near 6.0 to 6.3, because the calcium phosphate that dissolved during acidification precipitates in a different form and at different pH values from the original colloidal phosphate. This hysteresis is a direct sign that colloidal calcium phosphate is involved in milk buffering. Removing colloidal calcium phosphate, for example by dialysis or by adding a chelating agent, alters the shape of the curve, and heating, which shifts phosphate into the colloidal phase, also changes buffering.
Buffer capacity explains several everyday observations in the dairy. Milk richer in protein and phosphate, such as buffalo milk, sheep milk and concentrated milk, needs more acid to reach a given pH, so starter cultures must produce more lactic acid to make yoghurt or cheese, and titratable acidity values for the same pH are higher. Ultrafiltered milk used for cheese is strongly buffered because protein and colloidal minerals are concentrated, so the pH falls slowly during fermentation and the final pH of the cheese can be higher than intended unless the process is adjusted. In cheese, the buffering of the curd near pH 5 helps hold the pH fairly steady during ripening. Buffering also matters for heat stability: added stabilising salts such as phosphates and citrates, used in evaporated milk and processed cheese, act partly by changing calcium binding and pH. In nutrition, the buffering capacity of milk was once the reason for using it to relieve gastric acidity, but its protein and calcium later stimulate acid secretion, so this practice is no longer advised. For analysts, knowing the buffering of milk prevents the mistake of reading titratable acidity as a measure of pH or of freshness alone.
Viscosity is the resistance of a fluid to flow, the ratio of shear stress to shear rate. Water at 20 degrees Celsius has a viscosity of about 1.0 millipascal second; whole milk is about 2 millipascal seconds and skim milk somewhat less. Casein micelles make the largest contribution to the viscosity of milk, followed by fat globules, with whey proteins and lactose contributing less. Normal milk behaves as a Newtonian fluid, its viscosity independent of shear rate, except at low temperature and low shear rates, where raw milk shows slight shear thinning because cold agglutination clusters fat globules. Cream, concentrated milk, yoghurt and ice cream mix are non-Newtonian: their apparent viscosity depends on shear rate and often on time. Viscosity falls steeply as temperature rises, roughly halving between about 20 and 50 degrees Celsius, which is why milk is heated before separation and why evaporators and pumps are sized with temperature in mind. Viscosity rises with total solids and fat, with homogenisation (which increases the number of fat particles and coats them with casein), with heat treatment that denatures whey proteins, and with souring and age gelation. It is measured with capillary (Ostwald type), falling-ball or rotational viscometers; rotational instruments are needed for non-Newtonian products.
Surface tension is the force per unit length at a liquid-air interface, caused by the unbalanced attraction of molecules at the surface. Water at 20 degrees Celsius has a surface tension of about 72 to 73 millinewtons per metre; milk is much lower, commonly quoted at about 45 to 52 millinewtons per metre, because proteins, phospholipids and free fatty acids adsorb at the surface. Skim milk has a slightly higher surface tension than whole milk, and cream lower. Surface tension falls as temperature rises and falls further when lipolysis releases free fatty acids, which are strongly surface active; a fall in surface tension is one sign of rancidity. Surface activity governs foaming: the foaming of milk for beverages, the overrun of ice cream and whipped cream, and the unwanted foam in pumps and tanks all depend on how proteins adsorb at air bubbles. Interfacial tension is the corresponding property at the boundary between two immiscible liquids, here milk fat and the aqueous phase. Without a membrane, the interfacial tension between fat and water is high; the native MFGM and, after homogenisation, the adsorbed casein and whey proteins lower it to a very small value, stabilising the emulsion against coalescence. Measurement uses the du Noüy ring, the Wilhelmy plate, drop-weight or pendant-drop methods.
Electrical conductivity measures how readily milk carries an electric current. It depends on the concentration and mobility of ions, mainly sodium, potassium and chloride, and is lowered by fat globules and other non-conducting particles that obstruct ion movement. Normal cow milk has a conductivity of about 4 to 5.5 millisiemens per centimetre at 25 degrees Celsius. Conductivity rises with temperature by roughly 2 per cent per degree, so it is reported at a reference temperature. In mastitis, sodium and chloride enter milk from blood, and conductivity rises; this is the basis of in-line mastitis detection in automatic and conventional milking systems, which compare the conductivity of the four quarters of each cow or track a cow's own baseline, since absolute values vary between animals. Conductivity rises as milk sours, because lactic acid ionises and calcium phosphate dissolves, and changes when milk is watered or concentrated. In the plant, conductivity probes distinguish milk from water and from cleaning solutions in clean-in-place (CIP) systems, controlling the changeover between product and rinse and the strength of caustic and acid solutions.
The refractive index is the ratio of the speed of light in a vacuum to its speed in the material. Water has a refractive index of 1.333 at 20 degrees Celsius, and milk about 1.344 to 1.348, the excess coming from dissolved and dispersed solids, especially lactose and proteins. Because it rises with solids, the refractive index of milk or of milk serum obtained by precipitating the proteins has been used to estimate solids and detect added water, although cryoscopy is more reliable. Brix refractometers estimate the immunoglobulin content of colostrum and the solids of concentrated milks and sweetened products. The refractive index of milk fat, read with a butyro-refractometer, is an identity test for ghee and butter fat in national standards. Thermal properties are essential for heat-exchanger design. The specific heat of whole milk is about 3.9 kilojoules per kilogram per kelvin, slightly less than that of water (4.18), and falls as fat and solids rise, so cream has a lower specific heat than skim milk. The thermal conductivity of milk is about 0.5 to 0.6 watts per metre per kelvin. Milk fat has a latent heat of melting, so the apparent specific heat of cream rises over the melting range of fat, between about 10 and 40 degrees Celsius.
The oxidation-reduction (redox) potential, Eh, measures the tendency of a system to accept or donate electrons, expressed in volts against a reference electrode. Fresh raw milk exposed to air has a redox potential of about plus 0.2 to plus 0.3 volts at its natural pH, set mainly by dissolved oxygen, ascorbic acid, riboflavin and traces of copper and iron. When bacteria grow, they consume oxygen and produce reducing compounds, and the potential falls. This is the principle of the dye-reduction tests for raw milk quality: methylene blue and resazurin are added to milk and incubated, and the time taken for the dye to lose its colour (methylene blue) or to change from blue through pink to colourless (resazurin) shortens as the bacterial load rises. These tests reflect the metabolic activity of the bacteria more than their number, and their grading limits are set by national standards. Heating lowers the redox potential, because whey protein denaturation exposes sulfhydryl groups and oxygen is expelled; this reducing environment protects heated milk products from oxidation and is why cooked milk resists oxidised flavour. Conversely, copper contamination, light and high oxygen raise the potential and promote lipid oxidation and oxidised flavours. Low redox potential is needed for the growth of anaerobic starter and spoilage organisms in cheese.
Milk appears white because it scatters light of all visible wavelengths strongly. The scattering particles are the casein micelles and the fat globules, whose sizes are comparable with or larger than the wavelength of visible light. In whole milk, fat globules and micelles together produce a uniform white. Skim milk looks slightly bluish because, with most fat removed, the smaller casein micelles scatter short (blue) wavelengths more strongly than long ones. Homogenised milk looks whiter and more opaque than unhomogenised milk because breaking fat globules into many smaller ones increases scattering. Whey, the liquid left after casein is removed, is a translucent greenish-yellow because of riboflavin (vitamin B2), which absorbs blue light and fluoresces. The yellow tint of cow milk, cream, butter and ghee comes from beta-carotene and other carotenoids dissolved in the fat, which vary with feed: pasture and green fodder give more yellow fat than dry feeding. Buffalo, goat, sheep and camel milks are whiter because these species convert most dietary carotene to colourless vitamin A. Strong scattering also sets the opacity of milk, which affects the depth to which ultraviolet light penetrates in non-thermal treatments.
Colour changes in milk products signal processing history and quality. Heating causes browning through the Maillard reaction between lactose and lysine residues of proteins, and through caramelisation at high temperatures. Pasteurised milk shows no visible browning, UHT milk a slight change, and in-container sterilised milk, evaporated milk and khoa progressively more, with beige to brown colours and cooked or caramel flavours. Browning increases with time, temperature, pH and solids content and continues slowly during storage. Light exposure, especially in clear containers, excites riboflavin, which then oxidises methionine to methional and causes a light-induced (sunlight) off-flavour, while riboflavin itself is destroyed; opaque or pigmented packaging protects against this. Pink or red colours indicate blood, and blue, yellow or red discolouration can result from contamination with pigment-producing bacteria. Colour is measured instrumentally on the CIE L*a*b* scale, in which L* represents lightness, a* the red-green axis and b* the yellow-blue axis; b* tracks carotene content and browning, and L* falls with browning. Optical measurements underpin many rapid analytical methods: infrared absorption for fat, protein and lactose, turbidity in the Aschaffenburg turbidity test for distinguishing sterilised milk, and light scattering for measuring fat globule and micelle size.
A student uses this subject in the first dairy chemistry practicals and in every later course. You will determine fat, SNF and total solids, read and correct a lactometer, calculate SNF by the formula that matches your lactometer, measure titratable acidity and pH, and run the clot-on-boiling, alcohol and dye-reduction tests. Learn the typical composition of cow, buffalo, goat, sheep, camel and human milk, the normal ranges of density, freezing point, pH and acidity, and the direction in which each changes with mastitis, watering, skimming and souring. Practise the calculations until they are automatic, because the same handful of formulas appears in every entrance and PG paper and in the plant training log.
Five years into work as a reception chemist, quality officer or food analyst, you decide every day whether a tanker or a can of milk is accepted, priced or rejected. You read the infrared analyser and the cryoscope together, recognise when a normal-looking lactometer reading hides skimming plus watering, investigate high freezing points traced to rinse water left in tankers, and explain to a supplier why late-lactation or mastitic milk is paid less. You also use composition to standardise milk, to predict cheese and paneer yields, and to set heat-treatment conditions that depend on viscosity and specific heat.
A plant or quality manager with twenty-five years of experience designs the milk procurement and payment system: which components to pay on, which tests to run at village, chilling-centre and plant level, how to set base freezing points from regional surveys, and how to handle disputes. Such a person uses seasonal composition data to plan product mix, decides on component-based payment that rewards protein for cheese plants, sets quality incentives for low somatic cell counts, and defends test methods before regulators and courts in adulteration cases.
A researcher with long experience works on the physical chemistry behind these properties: how colloidal calcium phosphate and micellar structure set buffering and stability, how breed, genetics and feeding change composition, rapid spectroscopic methods for authenticity and species identification, and the relationship between udder health, composition and processing quality. Such a person designs surveys to establish authentic composition and freezing-point baselines, evaluates new reception sensors, and advises standard-setting bodies on limits that are fair to farmers yet detect adulteration. Methods include multivariate analysis of large herd-test datasets, controlled feeding trials and reference-method validation studies.
This manual gives typical textbook values; it does not list the legal fat and SNF minima or freezing-point limits of individual countries or Indian states, which change and must be read from current regulations. Detailed analytical procedures (Gerber, Kjeldahl, infrared calibration, cryoscope operation) belong to the analysis subject and are described only in principle here. Lactose, milk salts and vitamins, the detailed chemistry of milk lipids and proteins, and colloidal behaviour are treated in their own subjects. Composition figures for camel, yak, mare and donkey milk vary widely in the literature and are given only in outline. Temperature-correction tables for BIS 27 degree lactometers and the exact Hortvet-to-Celsius conversion used by each authority are not reproduced.
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.