Process for producing modified whey powder

Technical Field of the Invention

The present invention relates to a process for producing a modified whey powder from whey permeate.

Background Art

There are, in general, two kinds of whey that accrue as byproducts during processing of milk in the manufacture of dairy products such as cheese and curd. These kinds of whey are commonly referred to as sweet whey and sour whey.

Sweet whey is also termed cheese whey and is produced during cheese making, when rennet (an enzyme derived from a calf's stomach) is used for curdling. The pH value of sweet whey can range between 5.2 and 6.7.

Sour whey comprises the whey types commonly known as acid whey and curd or cottage-cheese whey. Acid whey, also known as casein whey, originates from the manufacture of casein by means of lactic acid, lactic acid generating bacteria or hydrochloric acid. As suggested by their names, curd or cottage-cheese whey are produced during the manufacture of quark and cottage-cheese. Lactic acid generated by natural fermentation imparts a high acidity to the whey such that the pH values of these types of whey typically range from 3.8 to 4.6.

In some cases, sour sweet whey is also considered encompassed by this term although this may appear incorrect. If insufficient care is given to sweet whey (cheese whey) it becomes sour by continued natural fermentation and is then referred to as sour sweet whey. This fermentation process is usually not desired.

A typical composition of these types of whey is shown in the following table:

Sweet whey

sour whey

% solids

6.4-6.8

6.8

% lactose

4.8

4.3-4.4

% protein

0.75

0.8

% fat

0.05

< 0.01

% ash

0.6

0.80

pH value

6.1

4.6

values are % by weight Reference: Zadow, J.G.: Whey and lactose processing, Elsevier Applied Science, 1992

If the whey is treated by means of ultrafiltration to recover valuable proteins, the product stream with lower protein content is the so-called whey permeate. Whey permeate is therefore also a by-product of cheese manufacture which accrues during protein recovery from whey by means of ultrafiltration.

Its use in downstream dairy processes and food production is limited due to the high mineral content and the poor sweetness. Therefore, so far whey permeate is brought to the market as a low value product disposed at virtually no creation of value by using it in fertilizer applications and as an animal feed.

Applications of whey permeate creating more value are desirable. This could be accomplished by providing a method to transform whey permeate into a raw material for 'dry applications' such as confectionery, biscuits, powdered soft drinks and other categories.

In the past, sweet whey powder (SWP) has been employed as a low-price substitute for sucrose in confectioneries and, in particular, in chocolate. However, any further increase of the SWP content and, thus, a further reduction of the costs of the ingredients is limited by

• the mineral content which leads to undesirable off-flavours,
• the high content of a-lactose which is the crystalline form of lactose usually resulting from spray-drying and which causes a powdery mouthfeel and lack of sweetness, both of which is undesirable in chocolate products, and, independently,
• the whey proteins present in SWP are not available for other economically more beneficial applications such as valuable nutrition products although they appear to have no functionality in chocolate products.

In particular in view of the latter issue, it is desirable to replace the SWP in chocolate products by another raw material that is available at lower costs, for instance, a powdery raw material derived from whey, in particular, whey permeate.

It is essential that such a raw material for dry applications has immaculate flavour properties and processability, i.e. no salty or metallic off-flavour and a low tendency of caking. The latter is important not only for later applications but also during manufacture, for instance, by spray-drying. Advantageous caking properties can be achieved if the formation of amorphous forms of lactose during drying is avoided. Furthermore, whey permeate originating from cottage cheese and casein production is known to be difficult to dry due to its high content of lactic acid. It tends to agglomerate and form lumps upon spray-drying. Independently, a method has to be found to lower the mineral content of whey permeate.

The object of removing minerals in whey and whey permeate has for instance been addressed in the following prior art documents:

• Zadow (in "Whey and Lactose Processing", Elsevier 1992, pages 83-85) discloses the processes of ion-exchange, lactose crystallization and spray-drying and shows some linkage between them. The reference mentions the influence of crystallization on the hygroscopicity of the product and, hence, on the processability during spray-drying, but the reference does not teach that this step is not sufficient to achieve good processability, but a previous demineralization step is necessary. It could be shown that the caking properties of the powder decrease significantly if a higher degree of demineralization is accomplished.

WO 02/50089 discloses a method for purification of lactose in a whey product comprising two demineralization steps and an additional crystallization. The second demineralization step includes the addition of alcohol in order to precipitate minerals. Hence, this reference is related to the production of edible high-purity lactose (99.8 %), whereas the present invention aims to provide a whey powder that has a lactose content higher than 80 %, preferably higher than 85 %, but lower than 95 %. Treating the whey with ion-exchange resin(s) is mentioned as a method suitable for the first demineralization step in order to remove divalent ions, although disadvantages of an ion-exchange step are discussed. According to the present invention both divalent and monovalent ions are removed by an ion exchange process.

US 2003/0000894 discloses a process for treating liquids, such as a citric acid fermentation broth, including a nanofiltration step and an ion-exchange step.

US 06,475,390 and EP-A-01541032, both emanating from PCT/AU98/00588, disclose a process for purifying biological molecules, such as lactose, from dairy streams, such as sweet cheese whey permeate or acid whey permeate, which combines two demineralization steps: a cation exchanger is used to remove the divalent cations in the first step and nanofiltration is used to remove monovalent ions in the second step. The permeate of the nanofiltration step is used to regenerate the ion exchanger resins. In contrast, according to the present invention also an anion exchanger is used in order to accomplish a high degree of demineralization in terms of anions.

EP-A-0083325 discloses a process for the manufacture of a sweetener, in which lactose is dissolved in water and subsequently hydrolyzed to glucose and galactose by means of a strongly acidic cation exchanger.

US 4,971,701, US 6,033,700, EP-A-0315135, and EP-A-0835610 disclose processes for removing at least a portion of salts contained in whey by means of electro-deionization using ion-exchange membranes. US 6,033,700 and EP-A-0835610 mention that demineralised milks and derivatives may be useful for replacing skimmed milk in the manufacture of confectionery-chocolate. Acid whey and sweet whey obtained from an ultrafiltration step are mentioned as starting materials for the disclosed deionization process. However, neither of these prior art references contains any teaching regarding the organoleptic properties of the obtained demineralised whey products and their suitability in confectionery such as chocolate and other products.

The present invention is directed to provide a solution to the problems outlined hereinabove, viz. to provide a process for producing whey powder suitable to be used in biscuits, powdered soft drinks and in confectionery, in particular, in chocolate products.

Summary of the Invention

The present invention relates to a process by which blends of whey and whey permeates can be converted to powders having a low mineral content, a low hygroscopicity, and advantageous organoleptic properties by a combination of ion-exchange demineralization and crystallization.

Detailed Description of the Invention

I. Process

a) Providing Whey Permeate

In the context of the present invention, the term "whey permeate" relates to whey that has been separated from the curd. Suitable techniques for separating whey from curd include ultrafiltration (UF) and nanofiltration (NF). Depending on the technique by which the whey permeate has been obtained, it will in the following be referred to as "UF whey permeate" and "NF whey permeate", respectively. Ultrafiltration is a variety of membrane filtration in which hydrostatic pressure forces a liquid against a semipermeable membrane. Suspended solids and solutes of high molecular weight are retained, while water and low molecular weight solutes pass through the membrane. This separation process is suitable for purifying and concentrating macromolecular solutions, i.e. solution containing compounds having a molecular weight in the range of about 10³ to 10⁶ Da, especially protein solutions.

In the context of the present invention, the term "whey permeate" encompasses any whey permeate, i.e. UF whey permate, NF whey permeate and blends thereof. Since whey protein can be marketed as valuable nutrition products, it is advantageous from an economic point of view to separate whey protein from whey prior to further processing of the whey.

In a preferred embodiment of the present invention, a blend of whey permeates from different cheese manufactures is used.

The present invention relates to whey permeates or blends of them with high variability in composition of incoming whey. Generally, the whey permeates used in the present invention have low protein content and high mineral content.

The mineral content is commonly also referred to and determined as ash content. The term "ash" thus comprises all compounds that are not considered organic or water. These are the compounds that remain (as "ashes") after a sample is burned, and consist mostly of metal oxides. It includes salts of such as salts comprising ions of alkali metals, alkaline earth metals and transition metals. Specific examples of ions included in ash include Na, K, Mg, Ca, A1, Mn, Fe, Si, and ions derived from phosphorus, but further ions may be present as well.

The whey permeate used in the present invention is furthermore characterized by the following features:

• Ash/lactose ratio: 0.13 or higher, for instance 0.16 or higher
• Ash/dry matter ratio: 7% or higher,

The absolute lactose content depends on the concentration and therefore on the total solids content of the incoming product.

Preferably, the whey permeate is pasteurized prior to further processing.

The meaning of the term "pasteurizing" is well known to the skilled person. In general, it relates to any method of heating food for the purpose of killing harmful organisms such as bacteria, viruses, protozoa, molds, and yeasts. Unlike sterilization, pasteurization is not intended to kill all micro-organisms in the food. Instead, pasteurization aims to achieve a reduction of the number of viable organisms by several orders of magnitude so that they are unlikely to cause disease (assuming the pasteurized product is refrigerated and consumed before its expiration date).

There are two widely used methods to pasteurize milk: high temperature/short time (HTST), and ultra-high temperature (UHT). HTST is by far the most common method. It involves holding the milk at a temperature of 72 °C for at least 15 seconds. UHT involves holding the milk at a temperature of 138 °C for at least two seconds. Pasteurization methods are usually standardized and controlled by national food safety authorities. There are different standards for different dairy products, depending on the fat content and the intended usage.

Generally, in the present invention, any pasteurization method suitable for achieving a reduction of the number of viable microorganisms by the factor 10^-5 in milk can be employed. This is considered adequate for destroying almost all yeasts, mold, and spoilage bacteria and also to ensure adequate destruction of common pathogenic heat-resistant organisms including particularly Mycobacterium tuberculosis, which causes tuberculosis and Coxiella burnetii that might be present in whey permeate. Processes must be designed so that the whey permeate is heated evenly, and no part of the whey permeate is subject to a shorter time or a lower temperature.

In the present invention, it is preferred to apply a HTST pasteurization treatment to the whey permeate. A typical procedure may include heating the whey permeate from its storage temperature to 74 °C for 13.5 seconds and cooling it immediately afterwards to a temperature of about 4°C.

b) Demineralization

The whey permeate is demineralised by an ion exchange process, i.e. by passing the whey permeate over a cation exchange resin and an anion exchange resin.

The term "ion exchange" relates to a reversible chemical process wherein ions are exchanged between a solution and an ion exchanger, that is usually an insoluble solid or gel. Typical ion exchangers are ion exchange resins, zeolite, montmorillonite, clay. Since the properties of ion exchange resins can be adjusted according to the requirements of a specific process, ion exchange resins are preferred with respect to the present invention. The resins should allow an even removal of different ions, e.g. as achieved with Lewatit resins S100 and MP 62, or also act ion-specifically if, due to pre-treatment, some ions were already removed.

Numerous types of ion exchange resins suitable for the present invention are available on the market under various tradenames.

The temperature of the whey permeate during the demineralization step can be in the range of from about 4 to about 10 °C. Typically, conductivity, Brix value and pH value of the whey permeate is monitored during demineralization. The degree of demineralization DD is, depending on the constitution of the incoming whey permeate or whey permeate blend, at least 90 %, preferably 95% and is calculated according to the following equation: $$\mathrm{DD}=\left(\mathrm{IB}-\mathrm{IA}\right)/\mathrm{IB}$$

wherein

IB is the concentration of ions in mol/l in the whey permeate before the ion exchange step and

IA is the concentration of ions in mol/l after the ion exchange step.

In particular, the following ion-specific degree of demineralization should be achieved:

Chloride:

95 % or more

Phosphorus-derived ions:

85 - 95 %

Calcium:

60 -70 %

Potassium:

90 % or more

Sodium:

90 % or more

Magnesium:

50 % or more

c) Crystallization of Lactose

The crystallization step comprises concentrating of the demineralised whey permeate prior to a rest period during which crystals of lactose are formed.

Concentrating the demineralised whey permeate is preferably accomplished by partial evaporation which can be carried out ' by any method of evaporation commonly known in the art. Typically, the demineralised whey permeate is heated to an elevated temperature in order to decrease the time necessary for the partial evaporation step. Evaporation is controlled by monitoring the density of the demineralized whey permeate. After the desired density of the demineralised whey permeate has been accomplished, the demineralised whey permeate is transferred to a crystallization vessel. If evaporation has been carried out at elevated temperature, this temperature should be maintained during the transfer in order to avoid premature crystallization outside the crystallization vessel. Premature crystallization, for instance, in the piping between the concentration equipment and the crystallization equipment is undesired as it may lead to the formation of deposits that complicate cleaning and maintaining of the equipment and might even provoke clogging. In the final stage of the evaporator a high temperature should be kept to avoid spontaneous crystallization of the lactose (higher than ca. 45°C) .

The content of total solids after evaporation, e.g. in a four-stage evaporator, should be as high as possible, at least 55 %, preferably at least 60%. To determine the total content of solids, the product is dried at 102 °C. Drying time depends on the product (3 to 6.5 hours). The total solids content is calculated as the ratio between weight after and before drying.

The rest period is essential in order to form crystals of lactose. Only if the major part of lactose, determined as alpha-lactose-monohydrate, is crystallized before spray-drying of the concentrated whey permeate, a satisfying result can be obtained during drying.

The crystallization, in e.g. a Terlet crystallization tank, can be carried out between 2°C and 20°C. At least 80% of the lactose should be in crystal form. The degree of crystallization is determined according to the procedure of K. Roetman and J. J. Mol as disclosed in Voedingsmiddelentechnologie, 7 (1974), W44-W45 (in Dutch language). The crystallization time can be between 2 hours and 24 hours depending on the incoming material. The preferred total solids content prior and after crystallization is 55 % or higher, more preferably 60% or higher. Thus, for instance, a mean crystal size d₅₀ of 60 µm and d₉₀ of 200 µm can be obtained.

d) Spray-drying the whey permeate

The partially crystallized whey permeate from step c) is spray-dried under following conditions:

• Inlet air temperature: 160 -190 °C, preferably 170 - 185 °C more preferably 180 - 185 °C
• Outlet temperature: 70 - 90 °C, preferably 75 - 85 °C, more preferably 80 - 85 °C.

For this process step, a spray with or without integrated fluid beds (external or internal) can be used. Such apparatus are commercially available.

II. Modified whey powder

The MWP according to the present invention can be characterized by the following parameters:

• Lactose content: 80 - 95 %, preferably 85 - 95%, most preferably 85 - 90 %
• Protein content: 0 - 3 %, preferably 0.5 - 2 %, most preferably 0.5 - 1.5 %
• Ash content: 0.5 - 2%, preferably 0.5 - 1.5 %, most preferably 0.9 - 1,5 %

In particularly preferred embodiments, the ash content is further specified in that the following components are present in the following ranges:

• Sodium: 6000 mg/kg or less, more preferably 5000 mg/kg or less
• Chloride: 3620 mg/kg or less, more preferably 1000 mg/kg or less
• Potassium: 10000 mg/kg or less, more preferably 4000 mg/kg or less

The protein content is determined by the Kjeldahl method.

The lactose content is determined by hydrolyzing it to D-glucose and D-galactose using the enzyme β-galactosidase and water. D-galactose is oxidized by nicotinamide-adenine-dinucleotide (NAD) to D-galactonic acid in the presence of the enzyme β-galactose dehydrogenase (Gal-DH). In this step, NADH is formed in an amount stoichiometric to the amount of D-galactose. The additional absorbance of NADH at a wavelength of 340 nm compared to the oxidized form NAD allows the determination of the lactose amount by means of the increase of the absorbance a t 340 nm.

The ash content is determined by ashing the sample in a muffle oven at 550 °C and calculating the ratio of the weight after ashing and the weight before ashing.

The content of potassium and sodium is analyzed by Inductively coupled Plasma Mass Spectroscopy (ICP-MS).

The content of chloride is determined by titration.

III. Product comprising the modified whey powder

The MWP obtainable by the process according to the present invention is suitable as an ingredient of compositions for the manufacture of confectionery or biscuits, Products with low water content. e.g. crackers and products where a salty or metallic taste affects organoleptic perception. It can be used as a sugar and sweet whey powder (SWP) replacer but also as a replacer for skim milk powder (SMP). In particular, it can replace sugar and/or SWP without negative effects on organoleptic properties of such products, such as, for instance, flavour and mouthfeel.

A preferred product comprising the MWP prepared by the process of the present invention is chocolate. In consumer tests, chocolate formulations in which at least part of the sugar and SMP was replaced with MWP prepared by the process according to the present invention were shown to have organoleptic properties similar or even better than standard formulations.

A particularly preferred chocolate formulation comprises from about 6 to 20 % by weight of MWP according to the present invention. Thus, the content of sugar and SWP can be reduced by about 20 % by weight. A typical chocolate formulation comprises the following amounts of ingredients:

Ingredients

Cocoa Ingredient.

Sugar

Milk Ingredient

MWP

[%]

28 %

41 %

16 %

6 %

The term "Cocoa Ingredient" is used for a combination of cocoa liquor and cocoa butter. The term "Milk Ingredient" is used for a combination of skimmed milk powder and anhydrous milk fat.

Examples

Processing of MWP from whey permeate

Ten MWP batches were produced whey permeate according to the following procedures.

Comparative Examples

M1

M2

M3

M4

M5

M6

M7

M8

M9

M10

Pasteurization

+

+

+

+

+

+

+

+

+

+

Heat / pH treatment

↓

↓

+

+

+

↓

↓

+

↓

+

Ion Exchange

+

↓

↓

↓

↓

+

↓

↓

↓

↓

Nanofiltration

↓

+

+

↓

↓

↓

+

+

↓

↓

Evaporation

+

+

+

+

+

+

+

+

+

+

Crystallization / spray drying

+

+

+

+

+

↓

↓

↓

↓

↓

Roller drying

+

+

+

+

+

+ indicated treatment carried out

↓ no treatment

Description of the individual processing steps

a) Pasteurization

25,000 L of whey permeate were provided in two trucks at a temperature of 8.2 °C and 8.1 °C, respectively. After unloading and prior to processing, the whey permeate was pasteurized for 13.5 seconds at 74 °C in a pasteurizer (Alfa Laval) having a capacity of 10,000 L/h. After pasteurizing, the whey permeate was cooled to about 4 °C.

b) Heat / pH treatment

Directly after pasteurization and cooling, the pH of 15,000 L of pasteurized whey permeate was adjusted to pH 8.0 with aqueous NaOH (concentration of NaOH: 33 % by weight). After storage overnight, the pH had decreased to 7.7 and the pH was re-adjusted to 8.0. In total, 65.7 kg of NaOH (concentration of NaOH: 33 % by weight) was added.

The temperature of whey was increased to 61-63 °C by means of an APV heat exchanger having a capacity of 5500 L/h. The heat exchanger was fouled by calcium phosphate precipitating from the whey permeate. After pasteurization of 7,500 L of whey permeate, the heat exchanger was cleaned with nitric acid. No additional cleaning was necessary for heating of 15,000 L of whey permeate.

After about 30 minutes, the heated whey permeate was centrifuged in a bactofuge (Westfalia Separator AG) having a capacity of 4,500 L/h. Every 600 seconds the sediment was removed with about 11 L of water. After centrifugation, the visible remaining calcium phosphate in the supernatant was determined with a laboratory centrifuge (6 min, 3,300 g).

After centrifugation, the pH value of the supernatant of batches M3 and M8 was decreased to 5 using aqueous HCl (concentration of HCl: 33 % by weight).

c) Ion exchange

Pasteurized whey permeate was demineralised on a cation exchanger (Lewatit resin S100 (K1)) and an anion exchanger (Lewatit resin MP62 (A1)). Conductivity, Brix value and pH of the whey permeate were checked regularly (every 6 minutes at the end of the run) in order to determine whether the capacity of the ion exchange columns was still sufficient to remove the ions. The temperature of the whey permeate remained between 4 and 10 °C during the process and the sample was cooled with ice water to 4 °C after ion exchange. The columns were rinsed with water and the diluted whey was collected until a Brix value of 5 °Bx was reached. The capacity of the ion exchange columns decreased after the first run. The regeneration procedure was extended to increase the capacity of the columns. During the first runs the anion exchange column was the limiting column being the slowest step. The output of the ion exchange runs are given in the following table:

Batch

Run 1 [L]

Run 2 [L]

Run 3 [L]

Run 4 [L]

Total [L]

M1

400¹⁾

580

700²⁾

700²⁾

2380

M6^§)

750

630

580

160¹⁾

2120

§) comparative example

1) desalted whey permeate was divided over batches M1 and M6

2) decrease in product flow to 30 % of maximum flow after about 45 minutes

In order to increase the overall capacity, the product flow in the second part of the runs was decreased. This resulted in an increase in binding capacity of the columns of more than 20%. The demineralised whey permeate had the following properties:

After cation exchange:

pH start

1.5 - 2.0

pH end

2 - 4

Conductivity start

10 - 11 mS

Conductivity end

6 - 10 mS

Solids content start

about 8.5 °Bx

Solids content end

about 8.5 °Bx

After anion exchange:

pH start

about 10

pH end

about 4.5

Conductivity start

about 1.1 mS

Conductivity end

about 1.3 mS

Solids content start

about 7.5 °Bx

Solids content end

about 8.0 °Bx

d) Nanofiltration

The whey was partly desalted and concentrated on a two-stage nanofiltration unit using the following membranes (in parallel and/or in series):

• 1.1 Desal 5 DK 38-40 C 5.6 m² spiral wound Osmonics
• 1.2 NF-3838/48 FF 5.6 m² spiral wound Filmtec, Dow chemical
• 2.1 Desal 5 DK 5.6 m² spiral wound Osmonics
• 2.2 NF-3838/48 FF 5.6 m² spiral wound Filmtec, Dow Chemical

All batches were demineralized with a product flow of about 350 L/h and a retenate flow of about 95 L/h as shown in the following table:

Batch

M3^§) + M8^§)

M2^§) + M7^§)

Product flow 1.1 [L/h]

350-330

35-300

Permeate flow 1.1 [L/h]

90-85

90-77

Permeate flow 1.1 [L/h]

105-95

100-78

Permeate flow 1.1 [L/h]

29-32

30-29

Permeate flow 1.1 [L/h]

18-32

15-30

Retentate flow [L/h]

93-82

99-83

Pout end of run [bar]

40.3

39.7

Volume before nanofiltration [L]

4500

4270

Volume after nanofiltration [L]

1205

1195

Concentration factor

3.7

3.6

Solids content end [°Bx]

27

27

§) comparative example

The temperature of the whey permeate was increased to about 10 °C before nanofiltration and cooled with ice water to about 4 °C after nanofiltration.

e) Evaporation

Batches M1, M4 to M6, M9, and M10 were heated to ca. 74 °C before evaporation on a 4-stage falling film evaporator (NIRO250). The product flow was about 1,700 L/h. Only two stages were used because the amount of whey permeate was too small to use complete capacity.

Batches M4 and M9, M5 and M10 were pooled and evaporated in two steps. After concentrating the batches were split. Batches M9 and M10 were kept at 60 °C and roller dried on the same day and batches M4 and M5 were transferred to the crystallization tank.

Batches M1 and M6 were preconcentrated to 23-31 °Bx and further concentrated on a different 4-stage falling film evaporator (Holvrieka).

Batches M1 to M3 and M6 to M8 were concentrated to about 52-55 °Bx. These batches were heated to about 74 °C before being evaporated on a 4-stage falling film evaporator (Holvrieka). Product flow was about 260-320 L/h. Using this evaporator, the batches could be successfully evaporated. Batches M2 and M7 which were only demineralised by means of nanofiltration gave some fouling (white precipitate, presumably caused by calcium phosphate deposition).

Processing parameters employed for operating the NIR0250 evaporator are shown in the following table:

Solids content

Batch

Starting Amount [L]

at start [°Bx]

at end [°Bx]

M1

2400

7

28*)

M6^§)

2115

8

23

M4^§) + M9^§)

4500

7 20

20 53

M5^§) + M10^§)

4800

8 22

22 55

§) comparative example

*) achieved with circulation in evaporator

Processing parameters employed for operating the Holvrieka evaporator are shown in the following table:

Solids content

Batch

Starting Amount [L]

at start [°Bx]

at end [°Bx]

M1

640

28

52

M2^§)#)

560

25

54

M3^§)

600

26

52

M6^§)

600

23

52

M7^§)#)

620

25

53

M8^§)

575

25

53

^§) comparative example

^#) a white precipitate formed, acid cleaning was necessary after 2 hours

f) Lactose crystallization / spray drying

The concentrated whey permeate was held at a temperature of about 60 °C during transportation to the crystallization tank(s) in order to avoid premature lactose crystallization outside of the tanks. The whey permeate was quickly cooled to 4 °C and stirred overnight at 4 °C.

The batches were crystallized in a crystallization tank having a capacity of 1,000 L (Terlet). After 4 hours of crystallization in the tank, 220 L of batch M4 was transported to two tanks having a capacity of 110 L each (Terlet).

After overnight crystallization, the samples taken from each batch were analyzed using a light microscope. Almost all the crystals were smaller than 100 µm and tomahawk-shaped. In batch M2 larger crystals were present, that presumably are partly composed of calcium phosphate. After heating the sample to 90 °C the lactose crystals dissolved (checked by means of the refractive index), but the supernatant remained turbid. The crystals from batch M1 were the smallest. This lactose solution contained the smallest amount of minerals and the crystals formed faster than in the other concentrated whey solutions.

The batches were spray-dried using a NIR025 spray-dryer equipped with a rotating wheel atomizer. The speed of rotation was 19,000 rpm. Further relevant parameters of operating are shown in the following table:

Preheat temperature [°C]

Permeate flow [L/h]

Outlet temperature [°C]

Moisture content [%]**)

Amount of powder [kg]

Amount of swept powder [kg]

M1

32

45

89

4.5

125

-

M2^§)

32

52

89

4.3

25

77

M3^§)

32

48

88-89

4.2

100

21

M4^§)

29

52

88

4.4

ca. 20

ca. 80*)

M5^§)

32

52

88

4.3

-

ca. 100

^§) comparative example

*) including large lumps

**) determined by Karl-Fischer titration

Observations:

Batch M1: After 1 hour some fouling of the cone occurred (warm, rainy weather). This could be cleaned easily.

Batch M2: After 3 hours the drying process was stopped because too much fouling occurred. Powder could easily be swept from the wall of the cone.

Batch M3: Drying could be carried out for a maximum of 5 hours due to fouling of the cone. In the bottom of the cone some lump formation and brown decolorization was observed.

Batch M4: Very sticky and difficult to dry. Fouling of the cone occurred. Decrease in outlet temperature to 85 °C gave more lump formation. Drying was stopped and the cone was' cleaned after 2 hours.

Batch M5: Very sticky and difficult to dry. The drying process was stopped 3 times after 1 hour each and the powder was swept. In the cone a powder layer of about 1.5 cm was formed.

The outlet temperature appeared very critical. If the outlet temperature was higher than approximately 90-92 °C, the powder became more rubbery and sticky. If the outlet temperature was lower than 86 °C the powder was too wet and became sticky as well.

The particle size of the different powders is given in the following table:

Batch

D (v, 0.5) [µm]

D (v, 0.1) [µm]

D (v, 0.9) [µm]

M1

72

27

134

M2^§)

66

18

136

M3^§)

64

16

131

M4^§)

50

15

698

M5^§)

40

12

102

^§) comparative example

*) two peaks, some lumps

D (v, 0.1) means that 10 % (of the whole volume or mass) of the powder is in particles that are below a certain value (e.g. 27 micrometer in batch M1)

D (v, 0.5) means that 50 % (of the whole volume or mass) of the powder is in particles that are below a certain value (e.g. 72 micrometer in batch M1)

D (v, 0.9) means that 90 % (of the whole volume or mass) of the powder is in particles that are below a certain value (e.g. 134 micrometer in batch M1)

g) Roller drying

The concentrated batches were dried on a drum dryer type T5/5 from GMF Gouda having a total surface of 1.5 m². The flakes were milled using a powder mill from GMF Gouda.

The following processing parameters were employed:

M6^§)

M7^§)

M8^§)

M9^§)

M10^§)

Steam pressure [bar]

3

3.4-4

3-3.4

2.5

2.5

Drum speed [rpm]

1.71

3-4

1.71

var.

var.

Distance of drums [mm]

0.2

0.2

0.1-0.3

var.

var.

Gap width [mm]

50-100

40

50-100

100-500

100-500

Feed temperature [°C]

65-80

66-76

65-89

92

92

Temperature of ca. knife [°C]

ca. 99

-

-

-

-

Product temperature [°C]

ca. 74

-

ca. 70

-

-

Structure/colour

Yellow golden flakes

Yellow golden flakes

Yellow golden flakes

Dark, liquid

Dark, liquid

Capacity [kg/h]

20

20

15-20

-

-

D.m. content*) [%]

0.4-1.0

1.5-2.2

1.8-3.3

2.7

n.d.

Amount of product [kg]

108

-

-

0

0

^§) comparative example

*) Dry matter content

Batches M9 and M10 contained the largest amount of salt and no flakes could be obtained, although the processing conditions (steam pressure, drum speed, gap width, feed temperature) were varied.

From batch M6 good dry flakes were obtained.

Batches M7 and M8 were slightly more sticky and brown after production. Presumably, this related to the higher amount of minerals.

The layer of product on the drums depended on the viscosity and dry matter content of the product in the gap and on the gap width. In order to achieve a good and constant powder quality, the gap between the drums had to be small to minimize moisture loss and brown discoloration. The gap width had to be constant in order to achieve a constant layer thickness on the drums. During the production, the holes became partly blocked by the formation of lactose crystals which were removed manually.

Modified whey powder

Thus, modified whey powder batches were obtained which had the following composition:

Batch

Protein

Lactose

Ash

Potassium

Sodium

Chloride

1%]

[%]

[%]

[mg/kg]

[mg/kg]

[mg/kg]

M1

2.0

90.7

1.0

2420

736

< 30

M2

2.31

71.20

4.62

11600

2450

380

M3

2.58

79.00

4.76

13400

6410

620

M4

4.69

67.60

1.21

38700

8790

2490

M5

2.57

76.00

8.14

25100

12100

1630

M6

1.98

85.10

1.64

1960

700

30

M7

2.61

82.40

5.37

14600

3070

40

M8

2.69

75.30

4.73

14700

7060

620

Chocolate formulations:

The following formulations were prepared and tested for consumer acceptance. All formulations were moulded into 100g size format.

Amount of ingredient

Ratio of components [%]⁺⁾

No.

MWP batch

Cocoa Ingr.

Sucrose

Milk Ingr.

MWP

α-lactose / total lactose

minerals / lactose

monovalent ions / all ions

1^§)

-

27.7

45.3

17,3

-

n.a.

n.a.

n.a.

2^§)#)

-

27.7

45.3

17,3

-

n.a.

n.a.

n.a.

3

M1

27.7

40.5

16.1

6

82

0.90

31.5

4^§)

M2

27.7

40.5

16.1

6

88

3.79

53.5

5^§)

M6

27.7

40.5

16.1

6

15

1.05

30.2

6^§)

M7

27.7

40.5

16.1

6

17

4.02

53.4

7^§)

M8

27.7

40.5

16.1

6

13

3.86

77.04

8

M1

28.2

40

16,13

6

82

0.90

31.5

9^§)

M6

28.2

40

16,13

6

15

1.05

30.2

10^§)

M6

27.7

37.2

15,34

10

15

1.05

30.2

^§) comparative example

^#) product was remoulded

⁺⁾ based on the Modified Whey Product M1 to M6, respectively

Test procedure:

Consumers (N = 150) were invited to evaluate the chocolate formulations on basis of 5 criteria. A 9-point hedonic scale was employed for evaluation.

Liking

No.

overall

melting in mouth

sweetness

milk flavour

chocolate flavour

aftertaste

8

7.05

6.93

6.68

6.93

6.68

6.78

3

6.81

6.45

6.58

6.72

6.50

6.62

1^§)#)

6.81

6.68

6.57

6.53

6.49

6.50

2^§)#)

6.81

6.58

6.59

6.75

6.56

6.63

10^§)

6.78

6.56

6.49

6.72

6.61

6.69

9^§)

6.77

6.57

6.66

6.91

6.47

6.77

7^§)

6.75

6.34

6.44

6.60

6.38

6.30

5^§)

6.66

6.45

6.30

6.58

6.32

6.40

4^§)

6.58

6.21

6.42

6.49

6.34

6.31

6^§)

6.54

6.51

6.28

6.56

6.23

6.32

^§) comparative example

^#) product was remoulded

As a result, chocolate formulations no. 3 and 8, both containing MWP of type 1, were evaluated as best accepted by consumer, in terms of overall liking.