R.E. Campbell

Invited review: Annatto usage and bleaching in dairy foods

Annatto is a yellow/orange colorant that is widely used in the food industry, particularly in the dairy industry. Annatto, consisting of the carotenoids bixin and norbixin, is most commonly added to produce orange cheese, such as Cheddar, to achieve a consistent color over seasonal changes. This colorant is not all retained in the cheese, and thus a percentage remains in the whey, which is highly undesirable. As a result, whey is often bleached. Hydrogen peroxide and benzoyl peroxide are the 2 bleaching agents currently approved for bleaching whey in the United States. Recent studies have highlighted the negative effect of bleaching on whey flavor while concurrently there is a dearth of current studies on bleaching conditions and efficacy. Recent international mandates have placed additional concern on the use of benzoyl peroxide as a bleaching agent. This review discusses the advantages, disadvantages, regulatory concerns, flavor implications, and optimal usage conditions of 2 widely used bleaching agents, hydrogen peroxide and benzoyl peroxide, as well as a few alternative methods including lipoxygenase, peroxidase, and lactoperoxidase systems.

The effect of bleaching agent on the flavor of liquid whey and whey protein concentrate

The increasing use and demand for whey protein as an ingredient requires a bland-tasting, neutral-colored final product. The bleaching of colored Cheddar whey is necessary to achieve this goal. Currently, hydrogen peroxide (HP) and benzoyl peroxide (BPO) are utilized for bleaching liquid whey before spray drying. There is no current information on the effect of the bleaching process on the flavor of spray-dried whey protein concentrate (WPC). The objective of this study was to characterize the effect of bleaching on the flavor of liquid and spray-dried Cheddar whey. Cheddar cheeses colored with water-soluble annatto were manufactured in duplicate. Four bleaching treatments (HP, 250 and 500 mg/kg and BPO, 10 and 20 mg/kg) were applied to liquid whey for 1.5 h at 60 degrees C followed by cooling to 5 degrees C. A control whey with no bleach was also evaluated. Flavor of the liquid wheys was evaluated by sensory and instrumental volatile analysis. One HP treatment and one BPO treatment were subsequently selected and incorporated into liquid whey along with an unbleached control that was processed into spray-dried WPC. These trials were conducted in triplicate. The WPC were evaluated by sensory and instrumental analyses as well as color and proximate analyses. The HP-bleached liquid whey and WPC contained higher concentrations of oxidation reaction products, including the compounds heptanal, hexanal, octanal, and nonanal, compared with unbleached or BPO-bleached liquid whey or WPC. The HP products were higher in overall oxidation products compared with BPO samples. The HP liquid whey and WPC were higher in fatty and cardboard flavors compared with the control or BPO samples. Hunter CIE Lab color values (L*, a*, b*) of WPC powders were distinct on all 3 color scale parameters, with HP-bleached WPC having the highest L* values. Hydrogen peroxide resulted in a whiter WPC and higher off-flavor intensities; however, there was no difference in norbixin recovery between HP and BPO. These results indicate that the bleaching of liquid whey may affect the flavor of WPC and that the type of bleaching agent used may affect WPC flavor.

Influence of bleaching on flavor of 34% whey protein concentrate and residual benzoic acid concentration in dried whey proteins

Previous studies have shown that bleaching negatively affects the flavor of 70% whey protein concentrate (WPC70), but bleaching effects on lower-protein products have not been established. Benzoyl peroxide (BP), a whey bleaching agent, degrades to benzoic acid (BA) and may elevate BA concentrations in dried whey products. No legal limit exists in the United States for BP use in whey, but international concerns exist. The objectives of this study were to determine the effect of hydrogen peroxide (HP) or BP bleaching on the flavor of 34% WPC (WPC34) and to evaluate residual BA in commercial and experimental WPC bleached with and without BP. Cheddar whey was manufactured in duplicate. Pasteurized fat-separated whey was subjected to hot bleaching with either HP at 500 mg/kg, BP at 50 or 100 mg/kg, or no bleach. Whey was ultrafiltered and spray dried into WPC34. Color [L*(lightness), a* (red-green), and b* (yellow-blue)] measurements and norbixin extractions were conducted to compare bleaching efficacy. Descriptive sensory and instrumental volatile analyses were used to evaluate bleaching effects on flavor. Benzoic acid was extracted from experimental and commercial WPC34 and 80% WPC (WPC80) and quantified by HPLC. The b* value and norbixin concentration of BP-bleached WPC34 were lower than HP-bleached and control WPC34. Hydrogen peroxide-bleached WPC34 displayed higher cardboard flavor and had higher volatile lipid oxidation products than BP-bleached or control WPC34. Benzoyl peroxide-bleached WPC34 had higher BA concentrations than unbleached and HP-bleached WPC34 and BA concentrations were also higher in BP-bleached WPC80 compared with unbleached and HP-bleached WPC80, with smaller differences than those observed in WPC34. Benzoic acid extraction from permeate showed that WPC80 permeate contained more BA than did WPC34 permeate. Benzoyl peroxide is more effective in color removal of whey and results in fewer flavor side effects compared with HP and residual BA is decreased by ultrafiltration and diafiltration.

The effect of starter culture and annatto on the flavor and functionality of whey protein concentrate.

The flavor of whey protein can carry over into ingredient applications and negatively influence consumer acceptance. Understanding sources of flavors in whey protein is crucial to minimize flavor. The objective of this study was to evaluate the effect of annatto color and starter culture on the flavor and functionality of whey protein concentrate (WPC). Cheddar cheese whey with and without annatto (15 mL of annatto/454 kg of milk, annatto with 3% wt/vol norbixin content) was manufactured using a mesophilic lactic starter culture or by addition of lactic acid and rennet (rennet set). Pasteurized fat-separated whey was then ultrafiltered and spray dried into WPC. The experiment was replicated 4 times. Flavor of liquid wheys and WPC were evaluated by sensory and instrumental volatile analyses. In addition to flavor evaluations on WPC, color analysis (Hunter Lab and norbixin extraction) and functionality tests (solubility and heat stability) also were performed. Both main effects (annatto, starter) and interactions were investigated. No differences in sensory properties or functionality were observed among WPC. Lipid oxidation compounds were higher in WPC manufactured from whey with starter culture compared with WPC from rennet-set whey. The WPC with annatto had higher concentrations of p-xylene, diacetyl, pentanal, and decanal compared with WPC without annatto. Interactions were observed between starter and annatto for hexanal, suggesting that annatto may have an antioxidant effect when present in whey made with starter culture. Results suggest that annatto has a no effect on whey protein flavor, but that the starter culture has a large influence on the oxidative stability of whey.

Effect of bleaching whey on sensory and functional properties of 80% whey protein concentrate

Whey is a highly functional food that has found widespread use in a variety of food and beverage applications. A large amount of the whey proteins produced in the United States is derived from annatto-colored Cheddar cheese. Color from annatto is undesirable in whey and must be bleached. The objective of this study was to compare 2 commercially approved bleaching agents, benzoyl peroxide (BP) and hydrogen peroxide (HP), and their effects on the flavor and functionality of 80% whey protein concentrate (WPC80). Colored and uncolored liquid wheys were bleached with BP or HP, and then ultrafiltered, diafiltered, and spray-dried; WPC80 from unbleached colored and uncolored Cheddar whey were manufactured as controls. All treatments were manufactured in triplicate. The WPC80 were then assessed by sensory, instrumental, functionality, color, and proximate analysis techniques. The HP-bleached WPC80 were higher in lipid oxidation compounds (specifically hexanal, heptanal, octanal, nonanal, decanal, dimethyl disulfide, and 1-octen-3-one) and had higher fatty and cardboard flavors compared with the other unbleached and BP-bleached WPC80. The WPC80 bleached with BP had lower norbixin concentrations compared with WPC80 bleached with HP. The WPC powders differed in Hunter color values (L, a, b), with bleached powders being more white, less red, and less yellow than unbleached powders. Bleaching with BP under the conditions used in this study resulted in larger reductions in yellowness of the powders made from whey with annatto color than did bleaching with HP. Functionality testing demonstrated that whey bleached with HP treatments had more soluble protein after 10 min of heating at 90°C at pH 4.6 and pH 7 than the no-bleach and BP treatments, regardless of additional color. Overall, HP bleaching caused more lipid oxidation products and subsequent off-flavors compared with BP bleaching. However, heat stability of WPC80 was enhanced by HP bleaching compared with control or BP-bleached WPC80.

The use of lactoperoxidase for the bleaching of fluid whey

Lactoperoxidase (LP) is the second most abundant enzyme in bovine milk and has been used in conjunction with hydrogen peroxide (H₂O₂) and thiocyanate (SCN⁻) to work as an antimicrobial in raw milk where pasteurization is not feasible. Thiocyanate is naturally present and the lactoperoxidase system purportedly can be used to bleach dairy products, such as whey, with the addition of very little H₂O₂ to the system. This study had 3 objectives: 1) to quantify the amount of H₂O₂ necessary for bleaching of fluid whey using the LP system, 2) to monitor LP activity from raw milk through manufacture of liquid whey, and 3) to compare the flavor of whey protein concentrate 80% (WPC80) bleached by the LP system to that bleached by traditional H₂O₂ bleaching. Cheddar cheese whey with annatto (15 mL of annatto/454 kg of milk, annatto with 3% wt/vol norbixin content) was manufactured using a standard Cheddar cheesemaking procedure. Various levels of H₂O₂ (5-100 mg/kg) were added to fluid whey to determine the optimum concentration of H₂O₂ for LP activity, which was measured using an established colorimetric method. In subsequent experiments, fat-separated whey was bleached for 1h with 250 mg of H₂O₂/kg (traditional) or 20 mg of H₂O₂/kg (LP system). The WPC80 was manufactured from whey bleached with 250 mg of H₂O₂/kg or 20mg of H₂O₂/kg. All samples were subjected to color analysis (Hunter color values and norbixin extraction) and proximate analysis (fat, protein, and moisture). Sensory and instrumental volatile analyses were conducted on WPC80. Optimal LP bleaching in fluid whey occurred with the addition of 20mg of H₂O₂/kg. Bleaching of fluid whey at either 35 or 50°C for 1 h with LP resulted in > 99% norbixin destruction compared with 32 or 47% destruction from bleaching with 250 mg of H₂O₂/kg, at 35 or 50°C for 1 h, respectively. Higher aroma intensity and increased lipid oxidation compounds were documented in WPC80 from bleached whey compared with WPC80 from unbleached whey. Monitoring of LP activity throughout cheese and whey manufacture showed that LP activity sharply decreased after 30 min of bleaching (17.01 ± 1.4 to < 1 U/mL), suggesting that sufficient bleaching takes place in a very short amount of time. Lactoperoxidase averaged 13.01 ± 0.7 U/mL in unpasteurized, fat-separated liquid whey and 138.6 ± 11.9 U/mL in concentrated retentate (11% solids). Lactoperoxidase may be a viable alternative for chemical whey bleaching.

Effects of Starter Culture and Storage on the Flavor of Liquid Whey

UNLABELLED The primary off flavors in dried whey proteins have been attributed to lipid oxidation products. A deeper understanding of lipid oxidation in fluid whey is crucial to understand how to minimize off flavors in dried whey protein. The objectives of this study were to further elucidate the role of storage and starter cultures as sources of lipid oxidation in whey. Fluid Cheddar, Mozzarella, and rennet-set wheys were manufactured from skim and whole milk. Liquid wheys and milks were evaluated by descriptive sensory and volatile instrumental analysis within 2 h of manufacture and following storage for 3 d at 4 °C. Culture type greatly influenced the oxidative stability of liquid whey, with Cheddar and Mozzarella whey differing not only in sensory profile, but also in volatile compounds. The type of starter culture (Mozzarella compared with Cheddar) had more influence on flavor than the set type (acid compared with culture). Milks had lower relative abundances of volatile free fatty acids than their liquid whey counterparts. Volatile lipid oxidation products in wheys were higher than in their respective milks, but oxidation in both milks and wheys increased with storage time. Wheys from Cheddar starters displayed more oxidation products than wheys from Mozzarella starters. Starter media did not have an effect on the flavor or oxidative stability of liquid whey, however, culture strain influenced lipid oxidation of fluid whey. PRACTICAL APPLICATION Lipid oxidation products are primary contributors to whey ingredient off-flavors. Flavor plays a critical and limiting role in widespread use of dried whey ingredients, and enhanced understanding of flavor and flavor formation in fluid whey are industrially relevant. Results from this study demonstrate that oxidation occurs in milk prior to cheesemaking but that starter type and starter strain influence also oxidative stability and lipid oxidation off flavors in fluid whey.

Effect of temperature and bleaching agent on bleaching of liquid Cheddar whey.

The use of whey protein as an ingredient in foods and beverages is increasing, and thus demand for colorless and mild-tasting whey protein is rising. Bleaching is commonly applied to fluid colored cheese whey to decrease color, and different temperatures and bleach concentrations are used. The objectives of this study were to compare the effects of hot and cold bleaching, the point of bleaching (before or after fat separation), and bleaching agent on bleaching efficacy and volatile components of liquid colored and uncolored Cheddar whey. First, Cheddar whey was manufactured, pasteurized, fat-separated, and subjected to one of a number of hot (68°C) or cold (4°C) bleaching applications [hydrogen peroxide (HP) 50 to 500 mg/kg; benzoyl peroxide (BP) 25 to 100 mg/kg] followed by measurement of residual norbixin and color by reflectance. Bleaching agent concentrations were then selected for the second trial. Liquid colored Cheddar whey was manufactured in triplicate and pasteurized. Part of the whey was collected (no separation, NSE) and the rest was subjected to fat separation (FSE). The NSE and FSE wheys were then subdivided and bleaching treatments (BP 50 or 100 mg/kg and HP 250 or 500 mg/kg) at 68°C for 30 min or 4°C for 16 h were applied. Control NSE and FSE with no added bleach were also subjected to each time-temperature combination. Volatile compounds from wheys were evaluated by gas chromatography-mass spectrometry, and norbixin (annatto) was extracted and quantified to compare bleaching efficacy. Proximate analysis, including total solids, protein, and fat contents, was also conducted. Liquid whey subjected to hot bleaching at both concentrations of HP or at 100mg/kg BP had greater lipid oxidation products (aldehydes) compared with unbleached wheys, 50mg/kg BP hot-bleached whey, or cold-bleached wheys. No effect was detected between NSE and FSE liquid Cheddar whey on the relative abundance of volatile lipid oxidation products. Wheys bleached with BP had lower norbixin content compared with wheys bleached with HP. Bleaching efficacy of HP was decreased at 4°C compared with 68°C, whereas that of BP was not affected by temperature. These results suggest that fat separation of liquid Cheddar whey has no effect on bleaching efficacy or lipid oxidation and that hot bleaching may result in increased lipid oxidation in fluid whey.

Flavor and stability of milk proteins

A greater understanding of the nature and source of dried milk protein ingredient flavor(s) is required to characterize flavor stability and identify the sources of flavors. The objective of this study was to characterize the flavor and flavor chemistry of milk protein concentrates (MPC 70, 80, 85), isolates (MPI), acid and rennet caseins, and micellar casein concentrate (MCC) and to determine the effect of storage on flavor and functionality of milk protein concentrates using instrumental and sensory techniques. Spray-dried milk protein ingredients (MPC, MPI, caseins, MCC) were collected in duplicate from 5 commercial suppliers or manufactured at North Carolina State University. Powders were rehydrated and evaluated in duplicate by descriptive sensory analysis. Volatile compounds were extracted by solid phase microextraction followed by gas chromatography-mass spectrometry (GC-MS) and gas chromatography-olfactometry. Compounds were identified by comparison of retention indices, odor properties, and mass spectra against reference standards. A subset of samples was selected for further analysis using direct solvent extraction with solvent-assisted flavor extraction, and aroma extract dilution analysis. External standard curves were created to quantify select volatile compounds. Pilot plant manufactured MPC were stored at 3, 25, and 40°C (44% relative humidity). Solubility, furosine, sensory properties, and volatile compound analyses were performed at 0, 1, 3, 6, and 12 mo. Milk proteins and caseins were diverse in flavor and exhibited sweet aromatic and cooked/milky flavors as well as cardboard, brothy, tortilla, soapy, and fatty flavors. Key aroma active compounds in milk proteins and caseins were 2-aminoacetophenone, nonanal, 1-octen-3-one, dimethyl trisulfide, 2-acetyl-1-pyrroline, heptanal, methional, 1-hexen-3-one, hexanal, dimethyl disulfide, butanoic acid, and acetic acid. Stored milk proteins developed animal and burnt sugar flavors over time. Solubility of MPC decreased and furosine concentration increased with storage time and temperature. Solubility of MPC 80 was reduced more than that of MPC 45, but time and temperature adversely affected solubility of both proteins, with storage temperature having the greatest effect. Flavor and shelf stability of milk proteins provide a foundation of knowledge to improve the flavor and shelf-life of milk proteins.

Short communication: Development of a novel method for the extraction of norbixin from whey and its subsequent quantification via high performance liquid chromatography

Norbixin is the primary carotenoid in annatto coloring, which imparts the desired orange color in Cheddar cheese. However, a portion of the colorant remains in the cheese whey and is undesirable; therefore, a bleaching step is often applied. Restrictions exist for norbixin concentrations in products destined for infant formula. As such, evaluation of norbixin concentrations in whey and whey ingredients is desirable. Current extraction methods are laborious and require solvents that are banned in many countries. The objective of this study was to develop a fast and inexpensive norbixin extraction and quantitation technique using approved solvents with similar sensitivity to current established methods. Instead of solvent extraction and column purification, acetonitrile was added directly to fluid wheys, retentates, and rehydrated whey protein concentrates. An isocratic mobile phase [70% acetonitrile and 30% water with 0.1% (wt/vol) formic acid] was used and, to increase sensitivity, a large volume (50 μL) was injected onto the column. The column used was a C18 column with a particle size of 2.6 μm and column length of 10 cm. The column inner diameter was 4.6mm and the pore size was 100 Ǻ. All of the previously described conditions allowed the run time to be only 4 min. The sample was sent through a photodiode array detector and quantified at 482 nm. Norbixin was quantified using external standard curves. The developed method had a >90% norbixin recovery in both milk and whey (9.39 μg/L-2.35 mg/L). The limit of detection of norbixin in fluid whey was 2.7 μg/kg and the limit of quantitation was 3.5 μg/kg, both of which are significantly lower than in previously described methods. The extracts were stable over 30 min at 21°C and stable over 24h at 4°C. Repeatability and precision of the method had relative standard deviations of less than 13%. The developed method provides time and cost savings for evaluation of norbixin concentration in whey and whey products.