Process Cheese: Scientific and Technological Aspects—A Review Rohit Kapoor and Lloyd E. Metzger
ABSTRACT: Process cheese is produced by blending natural cheese in the presence of emulsifying salts and other dairy and nondairy ingredients followed by heating and continuous mixing to form a homogeneous product with an extended shelf life. Extensive research on the important physicochemical and functional properties associated with process cheese and the various physicochemical, technological, and microbiological factors that influence these properties has resulted in process cheese being one of the most versatile dairy products with numerous end-use applications. The present review is an attempt to cover the scientific and technological aspects of process cheese and highlight and critique some of the important research findings associated with them. The 1st objective of this article is to extensively describe the physicochemical properties and microstructure, as well as the functional properties, of process cheese and highlight the various analytical techniques used to evaluate these properties. The 2nd objective is to describe the formulation parameters, ingredients, and various processing conditions that influence the functional properties of process cheese. This review is primarily targeted at process cheese manufacturers as well as students in the field of dairy and food science who may want to learn more about the scientific and technological aspects of process cheese. The review is limited to the relevant research associated with process cheeses as defined by the U.S. Code of Federal Regulations and does not cover imitation and substitute cheeses.
Introduction Process cheese is a dairy product which differs from natural cheese in the fact that process cheese is not made directly from milk. However, the main ingredient of process cheese is natural cheese. Process cheese is produced by blending natural cheese of different ages and degrees of maturity in the presence of emulsifying salts and other dairy and nondairy ingredients followed by heating and continuous mixing to form a homogeneous product with an extended shelf life (Meyer 1973; Thomas 1973; Caric and others 1985; Guinee and others 2004). The origin of process cheese dates back to the early 20th century (Meyer 1973). Contrary to the present status of process cheese, the initial idea of process cheese was to increase the shelf life of natural cheese and find alternative uses for natural cheese that was difficult to sell. Process cheese was invented in 1911, in Switzerland, by Walter Gerber and Fritz Stettler of Gerber and Co. who melted Swiss cheese using sodium citrate as the emulsifying salt to produce a smooth, homogeneous product. A few years later, in the United States, the development of process cheese was brought about by J. L. Kraft in 1916, when he preserved natural cheese in cans by heating and mixing it in order to increase its shelf life. The
development of process cheese with the use of phosphate-based emulsifying salts in the United States can be attributed to J. L. Kraft and the workers from Phenix Cheese Co. who were awarded numerous patents for their work on process cheese between 1916 and 1938, as reported by Zehren and Nusbaum (2000), who have extensively reviewed the history of the development of process cheese in the United States. Legal definition
In the United States, process cheese is a generic term used to describe various categories of cheese as defined by the Code of Federal Regulations (CFR). According to the CFR, these categories differ on the basis of the requirements for minimum fat content, maximum moisture content, and minimum final pH, as well as the quantity and the number of optional ingredients that can be used (21CFR133.169 to 133.180) (FDA 2006). The 3 major categories of process cheese, as described by the CFR, are pasteurized process cheese (PC), pasteurized process cheese food (PCF), and pasteurized process cheese spread (PCS). Table 1 summarizes the allowed ingredients and compositional specifications of PC, PCF, and PCS in the United States. In addition to the categories described by the CFR, there is another undefined category called pasteurized process cheese products. This category of proMS 20070698 Submitted 9/11/2007, Accepted 11/28/2007 . Authors are with cess cheese has a composition similar to the various categories of Midwest Dairy Foods Research Center, Dept. of Food Science and Nutrition, process cheese; however, ingredients such as milk protein conUniv. of Minnesota, St. Paul, MN 55108, U.S.A. Direct inquiries to author Metcentrate that are not allowed in PC, PCF, or PCS are utilized in zger (E-mail:
[email protected]). the formulation. 194
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
C
2008 Institute of Food Technologists
Process cheese—a review . . . Table 1 --- CFRa definition of the 3 major categories of process cheese in the United States. Category PCb
Major ingredients and other optional ingredients (and their permitted levels)
rCheese rEmulsifying agents (≤ 3% (w/w) of the final product) rAcidifying agent rCream, anhydrous milk fat, dehydrated cream (weight of the fat derived is ≤ 5%
Moisture (% w/w)
Fat (% w/w)
pH
≤40
≥30
≥5.3
≤44
≥23
≥5.0
44 to 60
≥20
≥4.0
(w/w) of the final product)
rWater, salt, colors, spices or flavorings, enzyme-modified cheese, mold inhibitors PCFc PCSd
(≤ 0.2% (w/w) or ≤ 0.3% (w/w) of the final product), antisticking agent (≤ 0.03% (w/w) of the final product) rCheese (≥ 51% (w/w) of the final product) rOther optional ingredients and their permitted levels include all of the ingredients allowed in PC in addition to milk, skim milk, buttermilk, and cheese whey rCheese (≥ 51% (w/w) of the final product) rOther optional ingredients and their permitted levels include all of the ingredients allowed in PCF in addition to food gums, sweetening agents, and nisin (≤ 250 ppm of the final product)
a CFR = Code of Federal Regulations (FDA 2006). b PC = pasteurized processed cheese (21CFR133.169). c PCF = pasteurized processed cheese food (21CFR133.173). d PCS = pasteurized processed cheese spread (21CFR133.179).
Production and market trends
The production of process cheese has remained relatively flat since 1990, and in 2005, total process cheese (PC, PCF, and PCS) production in the United States was approximately 1014 million kg (IDFA 2006). As a relative comparison, total natural cheese production was approximately 4149 million kg in 2005. Process cheese (243 million kg) was the leader in total supermarket cheese sales followed by cheddar (240 million kg) and mozzarella (120 million kg) (IDFA 2006). Supermarket sales of process cheese are primarily in the form of processed slices, and this form of process cheese accounts for 74% of the total supermarket sales. It is also interesting to note that the volume of low fat/light process cheese increased substantially (22.3%) between 2004 and 2005 (IDFA 2006). Although this is currently a small category, low fat/light process cheese appears to have the potential for substantial growth. Process cheese end-use applications and functional properties
The popularity of process cheese can be attributed to its numerous end-use applications. According to Sørensen (2001), process cheese is one of the leading cheese varieties in the world that is used as an ingredient in various food preparations (processed foods and food service). In the United States, process cheese is produced and sold in various forms such as loaves, slices, shreds, and spreads and is used as an ingredient in numerous products (Figure 1). The versatility of process cheese can be attributed to its unique functional properties. According to Guinee (2002), “the functional properties of a cheese (when used in a particular food) refer to the performance of the cheese during all stages of preparation and consumption of the food that would eventually contribute to the taste as well as the aesthetic appeal of that prepared-food.” Guinee (2002), in his review, extensively describes the various functional properties of natural cheese and process cheese and their potential applications. Depending on its end-use application, the desired functional properties of process cheese can be grouped into 2 major categories: unmelted texture and melted texture properties. Table 2 and 3 summarize the important unmelted textural and melted textural properties of process cheese, respectively. In addition to the individual functional properties, certain process cheese applications also require
Spread 4.5%
Cubed 0.3%
Other 0.1%
Shredded/Grated 1.1% Loaf 20.0%
Slice 74.0% Figure 1 --- Process cheese supermarket sales in the U.S.A. in 2005 based on form (Source: IDFA 2006).
an optimum interaction between both the melted and the unmelted textural properties. For example, the desired functional properties of process cheese used to make breaded cheese sticks would not only include high firmness and cohesiveness so that the process cheese can be easily cut or shredded when cold, but it would also need to have a normal melt and stretch (so that it softens during heating and stretches when consumed) and a high “melted” viscosity (so that the cheese does not ooze out of the breaded casing during cooking or consumption). Similarly, the appropriate process cheese slice for a toasted sandwich should not only have firmness, cohesiveness, and limited adhesiveness so that it has appropriate machineability during manufacture, but it should also have normal melt during toasting. Consequently, the required functional properties are unique for each process cheese product form and application. Various researchers have devised numerous empirical and instrumental techniques to evaluate and quantify the functional properties of process cheese and they are discussed in the section “Process Cheese Functional Properties.”
Vol. 7, 2008—COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
195
CRFSFS: Comprehensive Reviews in Food Science and Food Safety Process cheese manufacture
Figure 2 indicates a schematic flow chart of process cheese manufacture. Meyer (1973) and Zehren and Nusbaum (2000) have described in extensive detail the various steps of process cheese manufacture and the different equipment used in each step. The major steps in process cheese manufacture can be divided into 2 stages: 1 Ingredient selection and formulation: r Selection and grinding of natural cheese (on the basis of age, pH, flavor, and intact casein content) r Selection of appropriate emulsifying salt r Formulation and computation of other ingredients (in order to meet the targeted moisture, fat, salt, and pH values of the final product as per government regulations) 2 Process cheese processing and storage: r Cooking (heat and mixing) r Packaging, cooling, and storage Ingredient selection and formulation. The 1st stage of process cheese manufacture involves selection of ingredients and preparation of a formulation. As described in Table 1, in addition to natural cheese and emulsifying salts, there are various other dairy and nondairy (colors, flavors, spices, food gums, mold inhibitors, and so on) ingredients that are used in process cheese manufacture. Different ingredients affect the physicochemical properties, flavor, and the functional properties of process cheese in differ-
ent ways. Moreover, the appropriate selection of natural cheese and emulsifying salt is also very important in order to produce process cheese with desired final properties. Guinee and others (2004) have summarized the main function of certain optional ingredients on the final properties of process cheese. The effect of formulation parameters and ingredients on process cheese properties is discussed in detail in subsection “Formulation parameters and ingredients.” Process cheese processing and storage. Following the preparation of a desired formulation, the ingredient blend is processed using heat and mixing to produce a homogeneous mass, which is packaged and cooled. Although, the minimum cook temperature and time specified by CFR for process cheese is 65.5 ◦ C for 30 s (FDA 2006), process cheese manufacturers use various types of cookers with different designs and operating conditions to manufacture process cheese. These cookers differ on the basis of the mode of process cheese production (batch or continuous production), the type of mixing and agitating systems involved, and the type and mechanism of heating (indirect heating or direct steam injection) (Meyer 1973; Berger and others 1998; Zehren and Nusbaum 2000). Two common types of batch cookers use single/twin-screw augers (Blentech Cooker, Blentech Corp., Rohnert Park, Calif., U.S.A.) or highspeed cutting blades (Stephan Cooker, Sympak Inc., Mundelein, Ill., U.S.A.). The single/twin-screw auger cookers operate at low
Table 2 --- Major unmelted textural properties associated with process cheese along with their definitions, their importance in process cheese manufacture, and end-use ability and techniques commonly used to measure them. Descriptor terms Firmness
Definitiona Ability of the process cheese (at ambient or low temperatures) to show resistance to deformation when subjected to an external force.
Importance
Measurement techniques
1. Machineability during “slice-on-slice (SOS)” manufacture (chill belt) 2. Ability to maintain identity when shredded for preparing shredded cheese for retail and other food preparations 3. Slice identity for cold sandwich food preparations
1. “Thumb print” test 2. “Sliceability” test 3. Uniaxial compression test (texture profile analysis,b Instron): force at maximum compression 4. Penetrometry: force at maximum penetration 5. Low temperature dynamic stress rheometryc : elastic modulus
Brittleness/fractureability Tendency of the process cheese to fracture into pieces when subjected to an external force.
1. Ability to maintain identity when grated for cheese preparations for sprinkling on foods 2. Optimum “crumbliness” for salad preparations
1. Uniaxial compression test (texture profile analysis,b Instron): force at the 1st peak (point of fracture) during compression
Springiness/resilience
Tendency of the process cheese to recover to its original dimensions upon removal of the applied force.
1. Machineability during SOS manufacture (chill belt) 2. Ability to maintain identity when shredded for preparing shredded cheese for retail and other food preparations
1. Uniaxial compression test (texture profile analysis,b Instron) 2. “Roll” test: a representative sample of a slice immediately after it comes off the chill belt/casting rollers is rolled both perpendicular and parallel to the direction of the movement of the chill belt and the point at which it fractures is noted.
Adhesiveness/stickiness
Tendency of the process cheese to resist separation from a material it contacts.
1. Machineability during SOS 1. Uniaxial compression test (texture profile analysisb ) manufacture (chill belt) 2. Slice separation from other slices 2. “Slice separation” test: individual in “slice-on-slice” type product and slices are separated (from adjacent from the wrapper in “individually slices in the case of a SOS product wrapped slices (IWS)” type product and from the wrapper in the case 3. Stickability to foods in cold cheese of an IWS product) by hand and dips and cheese spreads their stickiness is evaluated.
a Adapted from Guinee b Breene (1975). c Marchesseau
196
(2002). The definitions correspond to the properties of process cheese at ambient and/or cold temperatures.
ˇ etina and others (1997), Drake and others (1999), and Piska and St ˇ (2003).
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
Process cheese—a review . . . mixing speeds ranging from 50 to 150 rpm with product temperatures ranging from 70 to 90 ◦ C with manufacturing times of 3 to 7 min. The high-speed cutting blade-type cookers operate at 1500 to 3000 rpm at 95 to > 100 ◦ C for 2 to 5 min. Recently, an additional cooker called the Rota Therm continuous cooker (Gold Peg Intl. Pty Ltd., Victoria, Australia) has been developed and is being used extensively for process cheese manufacture. This cooker operates at a high mixing speed (600 to 1000 rpm) with temperatures above 90 ◦ C and a residence time of approximately 30 to 40 s. Recently, another popular process to manufacture pro-
cess cheese involves sterilizing the premixes to 130 to 145 ◦ C for 2 to 3 s (Berger and others 1998). The primary method of heating utilized in most of the cookers is direct steam injection. Research has indicated that processing conditions such as cook time, temperature of cooking, extent of agitation (mixing) during cooking, and the rate at which the cooked process cheese is cooled have a significant effect on the functional properties of process cheese. The effect of processing conditions on process cheese properties is discussed in detail in the subsection “Processing conditions.”
Table 3 --- Major melted textural properties associated with process cheese along with their definitions, their importance in process cheese manufacture, and end-use ability and techniques commonly used to measure them. Definitiona
Descriptor terms Meltability Melt
Importance
Tendency of the process cheese to soften upon heating.
Viscosity/flow
1. Toasted sandwiches, burgers, and so on. Ability to maintain a uniform softening with minimal oiling-off when used on toasted sandwiches and other heated food preparations 2. Shredded cheese on pizza, in breaded cheese sticks, cheese insets in bratwurst, and burger patties 1. Cooker “drop-down” viscosity after manufacture 2. Pumpability during manufacture 3. Optimum hot-fill ability into loaves during packaging 4. Restricted flow during toasting (food preparation)
Tendency of the process cheese to spread and flow when completely melted.
Stretching ability
Tendency of the heated process cheese to form strings when extended.
a Adapted from Guinee (2002). b For the details of the various tests
1. Shredded cheese on pizza, in breaded cheese sticks
Measurement techniquesb 1. Arnott test 2. Schreiber melt test 3. Dynamic stress rheometry (DSR) 4. Melt profile analysis (UW Meltmeter) 5. Rapid visco analyzer (RVA)
1 Arnott test 2. Schreiber melt test 3. Tube melt test 4. Dynamic stress rheometry (DSR) 5. Melt profile analysis (UW Meltmeter) 6. Rapid visco analyzer (RVA) 1. Pizza “fork” test: The process cheese to be tested is shredded and baked on a pizza. The melted cheese after the pizza bake is extended/pulled into a string vertically off the pizza using a fork and the length of stretch before the string breaks is evaluated.
that measure the melted textural properties of process cheese, please refer to subsection “Techniques for measuring process cheese melted textural
properties” and Table 5.
Addition of other ingredients - Dairy protein ingredients - Dairy fat ingredients - Preservatives - Coloring agents - Flavoring agents - Water - Salt -Acidulants
Selection of natural cheese
Figure 2 --- Schematic flow chart of the basic steps involved in process cheese manufacture.
Grinding Blending Processing Packaging
Addition of emulsifying salts
Cooling Storage Vol. 7, 2008—COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
197
CRFSFS: Comprehensive Reviews in Food Science and Food Safety Due to an array of options in ingredients and formulations, and processing conditions, manufacturers have numerous possibilities for producing process cheese with different physicochemical properties which leads to a variety of flavor, functional properties, and end-use applications as desired by consumers. Therefore, appropriate selection of ingredient and processing conditions during process cheese manufacture is very important to produce process cheese with targeted functional properties. An additional, critical, component of process cheese manufacture is the utilization of rapid, at-line techniques for determining the fat and moisture content of process cheese. Fat and moisture testing are performed to ensure that the composition criteria identified in the CFR are met. The most common techniques utilized include near infrared spectroscopy and microwave-based methodology.
Process Cheese Physicochemical Properties and Microstructure Casein and natural cheese
Caseins are the major group of proteins present in milk and constitute approximately 2.3% to 3.0% of bovine milk (Eigel and others 1984; Swaisgood 1992). The 4 major casein molecules are: α s1 -casein, α s2 -casein, β-casein, and κ-casein and they are present in milk in a ratio of 4:1:4:1, respectively (Walstra 1990; Swaisgood 1992; Wong and others 1996). Caseins, like most proteins, have hydrophobic sections and hydrophilic sections. Caseins are unique in that they contain covalently attached phosphate groups and have a flexible hydrated secondary structure (Swaisgood 1992; Wong and others 1996; Farrell and others 2002). In their native form, caseins exist in the form of casein micelles. The micellar structure of caseins has been extensively reviewed (Farrell 1973; Payens and Veerman 1982; Schmidt 1982; McMahon and Brown 1984; Ruetimann and Ladish 1987; Walstra 1990; Holt 1992; Holt and Horne 1996; Horne 1998). Although the debate on the micellar configuration of caseins in milk is still in progress, most of the widely accepted models highlight the same fundamental configuration (Walstra 1990; Holt 1992). A casein micelle is 15 to 20 nm in diameter and is composed of around 10000 polypeptide chains with α s1 -, α s2 -, and β-casein present within the micelle. They are stabilized by protein–protein hydrophobic interactions and colloidal calcium phosphate-mediated cross-links. The κ-casein is mainly present on the surface of the micelle with its hydrophobic region embedded in the micelle and the glycosylated hydrophilic tail protruding outside (Holt 1992; Horne 1998). The glycosylated tail of κ-casein is negatively charged, thereby causing the micelles to repel each other. This phenomenon provides stability to the casein micelle, consequently protecting α s (α s1 and α s2 ) and β-casein components from being exposed to the environment. During the manufacture of natural cheese, rennet is used to cleave κ-casein at the Phe 105 and Met 106 position, thereby dislodging the glycosylated hydrophilic region (glycomacro peptide). Due to this phenomenon, the casein micelles lose their stability and α s - and β-caseins are exposed to the environment. The phosphoserine residues present on α s -and β-caseins take part in calcium-mediated cross-links, thereby forming a rigid, waterinsoluble, cross-linked calcium–paracaseinate phosphate complex commonly known as curd (Holt 1992). The fat phase is suspended in this calcium–paracaseinate phosphate complex. According to Shimp (1985), fat in natural cheese is underemulsified and the fat phase, as well as the water phase, is supported by a network of water-insoluble calcium–paracaseinate phosphate complex (Figure 3A). 198
Process cheese
In contrast to natural cheese, process cheese can be described as a stable oil-in-water emulsion (Palmer and Sly 1943; Shimp 1985; Zehren and Nusbaum 2000). The use of emulsifying salts such as disodium phosphate and trisodium citrate in process cheese manufacture aids in improving the emulsification properties of caseins by displacing the calcium phosphate complexes in the insoluble calcium–paracaseinate phosphate network present in natural cheese (Ellinger 1972; Gupta and others 1984; Caric and others 1985). This displacement of the calcium phosphate complex disrupts the major molecular force that cross-links the various monomers of casein in the network. This disruption of the calcium phosphate complex in conjunction with heating and mixing leads to hydration and partial dispersion of the calcium– paracaseinate phosphate network. In addition to being hydrated, the partially dispersed calcium–paracaseinate complex interacts with fat via hydrophobic interactions. After manufacture and during the cooling stage, the partially dispersed caseinate matrix forms “flocs” and the flocs subsequently interact to form a uniform, closely knit gel network (Zhong and Daubert 2004). This phenomenon gives rise to fat emulsified by a uniform closely knit protein gel network (Heertje and others 1981; Marchesseau and Cuq 1995; Ennis and others 1998; Lee and others 2003; Zhong and others 2004). Therefore, process cheese structure essentially consists of a fat phase evenly dispersed (in the form of fat globules, approximately < 1 to about 5 µm in diameter) in a partially dispersed casein gel network (Figure 3B). Various techniques have been developed to study and evaluate the physicochemical properties and microstructure of process cheese. Some of these techniques are discussed subsequently. Techniques for measuring process cheese physicochemical properties and microstructure Microscopic techniques. As a popular saying goes, “Seeing is
believing”; consequently, over the years, various microscopic techniques have been successfully utilized to study the structure of numerous food and dairy products (Heertje 1993; Kalab 1993; Kalab and others 1995). As for the microstructure of process cheese, Caric and others (1985) have extensively reviewed the microstructural changes that occur when natural cheese is converted into process cheese. Microstructural studies have described the changes in fat globule size and distribution (Rayan and others 1980) as well as rearrangement of the paracaseinate network (Heertje and others 1981; Heertje 1993; Lee and others 2003) in natural cheese during process cheese manufacture. Moreover, microscopic techniques have been utilized to study the size and distribution of fat globules, the type of crystals present (a defect in process cheese which will be discussed in a later section), and the characteristics of the proteinbased structural network of process cheese. Researchers have used both low-resolution microscopic techniques such as light microscopy, confocal microscopy, confocal laser-scanning microscopy, and fluorescence microscopy (Modler and others 1989; Sutheerawattananonda and others 1997; Bowland and Foegeding 2001; Awad and others 2002; Lee and others 2003) and high-resolution techniques such as transmission and scanning electron microscopy (Rayan and others 1980; Taneya and others 1980; Heertje and others 1981; Lee and others 1981, 2003; Caric and others 1985; Kalab and others 1987; Modler and others 1989; Savello and Ernstrom 1989; Tamime and others 1990; Awad and others 2002) to study the effects of various factors such as pH, ingredients, emulsifying salts, and processing conditions on fat globule distribution and the microstructure of process cheese.
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
Process cheese—a review . . . Spectroscopic techniques. Fluorescence spectroscopy has been successfully utilized to evaluate molecular-level interactions between fat and proteins in various food-based emulsions, as well as monitor structural changes in cheese and Maillard browning in milk and dairy products (Genot and others 1992; Dufour and Riaublanc 1997; Herbert and others 1999; Dufour and others 2000; Mazerolles and others 2001). Caseins in cheese contain the amino acid tryptophan, which is a naturally occurring fluorescent substance. The fluorescent properties of tryptophan in a hydrophobic environment are different from its fluorescent properties when it is in a hydrophilic environment (Lakowicz 1983). Consequently, researchers have utilized fluorescence spectroscopy to measure the spectra of tryptophan and predict the microstructure of natural cheese as well as process cheese (Karoui and others 2003; Garimella Purna and others 2005). On another
note, fluorescence spectroscopy has also been used to evaluate Maillard browning and oxidative stability of process cheese during storage (Christensen and others 2003). Analytical techniques involving wet chemistry. Recently, wet chemistry-based techniques have been developed and utilized to evaluate the physicochemical properties of process cheese (Lee and others 1979; Dimitreli and others 2005). These techniques have been used to study the protein-based interactions that contribute to the microstructure of process cheese (Marchesseau and Cuq 1995; Marchesseau and others 1997), to study the effect of different emulsifying salts on the mechanism of calcium chelation, and to evaluate the state of calcium in process cheese (Mizuno and Lucey 2005; Shirashoji and others 2006b). The major protein-based interactions that control the process cheese microstructure are hydrophobic interactions (between the various
(a)
Casein molecules
Figure 3 --- Schematic microstructure of (A) natural cheese and (B) process cheese.
Calcium phosphate microgranules (crosslinks)
Fat Solubilized casein molecules
Flocs (b)
Vol. 7, 2008—COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
199
CRFSFS: Comprehensive Reviews in Food Science and Food Safety caseins as well as between caseins and fat), hydrogen bonds, and calcium-mediated electrostatic bonds among the caseins. The basic principle behind these techniques is utilization of various chemical dissociating agents to selectively disrupt a particular kind of protein-based interaction in process cheese. This results in solubilization of the protein that was involved in the interaction. The amount of soluble protein is quantified with and without the addition of the dissociating agent and is used as an index of the relative importance of each protein-based interaction. The various chemical dissociating agents destabilize the protein-based interactions in different ways. Hydrophobic interactions are destabilized by sodium dodecyl sulfate. Hydrogen bonds are disrupted by urea, whereas ethylenediaminetetraacetate (EDTA) aids in breaking calcium–mediated electrostatic bonds (Marchesseau and Cuq 1995; Lefebvre-Cases and others 1998; Keim and Hinrichs 2004). Recently, an acid-base titration technique has been used to evaluate the calcium chelation ability of different emulsifying salts as well as the final state of calcium in process cheeses manufactured using different emulsifying salts (Mizuno and Lucey 2005; Shirashoji and others 2006b). The basic principle involved in this technique is that dilute solutions of process cheese are titrated using an acid followed by a base (at a constant rate) and the buffering capacity of the process cheese solution is measured at different pH values (Hassan and others 2004). Different calcium-based complexes and salts present in process cheese affect the buffering capacity of the solution in a variety of ways and the data collected are used to identify the state of calcium in the sample (Mizuno and Lucey 2005; Shirashoji and others 2006b).
Process Cheese Functional Properties The definitions, the 2 major types, and the importance of the various functional properties of process cheese have been covered previously in this article (subsection “Process cheese end-use applications and functional properties”). From a material science standpoint, process cheese can be described as a viscoelastic material since it is neither truly elastic (like an ideal solid) nor truly viscous (like an ideal liquid) (Gunasekaran and Ak 2003). Consequently, according to a rheologist, the functional properties of a process cheese are defined as properties that control its deformation and flow behavior when subjected to external forces. Gunasekaran and Ak (2003), in their book, have provided an extensive description of various rheological and textural properties of different cheeses as well as different instrumental techniques utilized to measure these properties. There have been numerous empirical and instrumental techniques utilized to evaluate the functional properties of cheese (Gunasekaran and Ak 2003). Our aim here is only to highlight and compare the various empirical and instrumental techniques that have been utilized over the years to evaluate the functional properties of process cheese in order to make it easier for readers to interpret the results when they come across these techniques in process cheese research reports. Once again, depending on its end-use application, process cheese functional properties can be grouped into 2 major categories: unmelted textural properties and melted textural properties.
and others 1984), as well as standard instrumental techniques such as textural profile analysis (Gupta and others 1984; Drake and others 1999; Kapoor and Metzger 2004, 2005) and lowtemperature dynamic rheological analysis (Drake and others 1999) have been utilized to evaluate process cheese hardness, fracturability, cohesiveness, adhesiveness, gumminess, chewiness, slicing ability, and elastic and viscous properties at low temperatures. Moreover, different process cheese manufacturers throughout the United States employ various customized techniques to measure the unmelted texture of process cheese such as firmness, slicing ability, and stickiness/adhesiveness depending on the resources available at their facilities. One of the earliest studies on the effect of various parameters on the firmness of process cheese was performed by Templeton and Sommer (1930). They compressed a standard sample of process cheese to a specified height and measured the force exerted by the process cheese in grams, which was indicated as the firmness of process cheese. Researchers have also used different penetrometry techniques to measure the firmness of process cheese (Thomas and others 1970a; Kalab and others 1987; Tamime and others 1990). Thomas and others (1970a) measured the firmness of cheese using a modified ball and a cone penetrometer and indicated the firmness of cheese as the distance traveled by the ball or the cone. In the same study, Thomas and others (1970a) also developed instruments to measure slicing ability, fracturability, and stickiness of process cheese. One of the more popular techniques to measure the unmelted texture of process cheese is texture profile analysis (TPA). TPA works by sending a crosshead down a vertical column, causing a “flat” plate to deform a specimen that has been placed on a lower plate. The constant crosshead speed leads to both force– time and force–distance curves and the work done to deform the cheese can be calculated (Breene 1975). Various types of equipment have been used to perform TPA on process cheese, such as the Instron Universal Testing machine (Harvey and others 1982; Gupta and others 1984), General Foods Texturometer, and the TA.XT2i texture analyzer (Texture Technologies Corp., Scarsdale, N.Y., U.S.A.) (Drake and others 1999; Awad and others 2002; Kapoor and Metzger 2004, 2005; Prow and Metzger 2005). TPA enables the measurement of a variety of unmelted textural properties of process cheese such as hardness, fracturability, adhesiveness, springiness, cohesiveness, and gumminess. Breene (1975), Peleg (1976), and Gunasekaran and Ak (2003) have provided detailed descriptions of TPA and how these parameters are measured. Recently, a low-temperature dynamic rheological analysis technique called dynamic stress rheometry (DSR) has also been used to measure and evaluate the unmelted viscoelastic properties of process cheese (Marchesseau and others 1997; Drake ˇ etina 2003). DSR measures the visand others 1999; Piska and Stˇ coelastic properties of process cheese. This device determines the storage modulus (G ), which is a measure of the elastic properties of process cheese, the loss modulus (G ), which is a measure of the viscous properties of process cheese, and the tan δ (which is G /G ) (Gunasekaran and Ak 2003). Drake and others (1999) in their study found a good correlation between G , G , and TPA hardness. Techniques for measuring process cheese melted textural properties
According to Gunasekaran and Ak (2003), melted textural properties or meltability of cheese refers to the “ease and extent to Various empirical techniques using customized instruments which the cheese will melt and spread/flow upon heating.” Vari(Templeton and Sommer 1930; Thomas and others 1970a; Gupta ous empirical and instrumental techniques have been developed
Techniques for measuring process cheese unmelted textural properties
200
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
Process cheese—a review . . . to measure and evaluate the melted texture of process cheese. The earliest method to measure process cheese melted texture was the Arnott Test (Arnott and others 1957). In this method, process cheese cylinders of specific dimensions are heated in an oven at a specific temperature for a specific time and the percent decrease in the height of the cylinder was reported as the melt of process cheese (Arnott and others 1957; Park and Rosenau 1984). Olson and Price (1958) developed the Tube Melt Test mainly to measure the melted texture and flow properties of process cheese spreads. In this test, a process cheese sample of specific weight and dimensions is placed in a glass tube that is heated horizontally in an oven at a specific temperature for a specific time and the extent of flow is measured to quantify the melted texture of the process cheese. The most popular empirical melt test for process cheese is the L.D. Schreiber Melt Test that was developed by Schreiber Foods of Green Bay, Wis., U.S.A. (previously known as L.D. Schreiber Co.) (Kosikowski and Mistry 1997). Over the years, researchers have modified the Schreiber Melt Test to overcome some of its shortcomings (Bogenrief and Olson 1995; Muthukumarappan and others 1999a). However, the basic principle of the test has not changed. In this test, process cheese discs of specific dimensions are heated in an oven at a specific temperature for a specific time and the final diameter or area of process cheese after melting is reported as the melt of the process cheese (Harvey and others 1982; Park and Rosenau 1984; Muthukumarappan and others 1999a). There have been various modifications to the above-mentioned melt tests in terms of the sample dimensions and testing conditions (Park and Rosenau 1984; Gunasekaran and Ak 2003). Instrumental, rheologicalbased techniques to measure the melted texture of process cheese include DSR and squeeze flow rheometry. As mentioned previously, DSR is a rapid test to measure the viscoelastic properties of process cheese (Drake and others 1999). Sutheerawattananonda and Bastian (1998) developed a DSR-based method for process cheese that heats the sample and utilizes DSR to measure the G , G , and the melting temperature (temperature at which G = G or tan δ = 1) of process cheese. Recently, our laboratory used DSR and the Tube Melt Test to evaluate the melted texture of process cheese spreads. We found a good correlation between G at 85 ◦ C from DSR and extent of flow as measured from the Tube Melt Test (Prow and Metzger 2005). Squeeze flow rheometry has also been utilized to measure the various melt properties of process cheese (Campanella and others 1987). One of the examples of the squeeze flow rheological technique is the UW Meltmeter and Melt Profile Analysis developed at the Univ. of Wisconsin, Madison, Wis., U.S.A. (Wang and others 1998; Muthukumarappan and others 1999b; Gunasekaran and Ak 2003). Melt profile analysis measures the softening point (softening time and softening temperature) that defines the ease with which the cheese melts (Muthukumarappan and others 1999b) and the melting point (melting time and melting temperature) that indicates both the ease of melting and extent of flow of a cheese (Gunasekaran and Ak 2003). Recently, another instrument was optimized to measure the melted textural properties of process cheese in our laboratory. This instrument is known as a rapid visco analyzer (RVA). The RVA is a computer-integrated instrument developed by Newport Scientific (Warriewood, Australia) to determine the viscous properties of cooked starch, grain, batters, and other foods. The RVA can measure apparent viscosity over variable conditions of shear and temperature as defined by the operator. Prow (2004) and Prow and Metzger (2005) used the RVA to evaluate the melted textural properties of process cheese and process cheese spreads. Prow (2004) developed a methodology to continuously measure the apparent viscosity of process cheese during a heating, hold-
ing, and cooling profile using the RVA and measured the minimum apparent viscosity (hot apparent viscosity) of the process cheese at the highest temperature as well as the time required for the process cheese to reach an apparent viscosity of 5000 cP (time at 5000 cP) during the cooling stage. According to Prow (2004), the hot apparent viscosity is a measure of how well a cheese flows when completely melted; and the time at 5000 cP is a measure of how quickly a melted cheese thickens during cooling. There was a good correlation between both the hot apparent viscosity and time at 5000 cP with process cheese melt properties as determined by DSR, the Schreiber Melt Test, and the Tube Melt Test. Since there are various melt tests, and each melt test measures a set of different properties (ease of melt, extent of flow, and so on), it becomes difficult to compare results obtained using different tests. Table 4 is a simplified chart, indicating how the values obtained using various melt tests are related. This will help readers to quickly comprehend the melted texture of process cheese when measured using different melt tests. As mentioned previously, numerous techniques have been developed to evaluate the functional properties of process cheese. However, one of the major limitations encountered is a poor correlation among some of these techniques (Park and Rosenau 1984; Gunasekaran and Ak 2003). Consequently, there is still a need for further research to develop standardized techniques to measure the unmelted textural and the melted textural properties of process cheese. When this is accomplished, manufacturers can evaluate the functional properties of their product and effectively communicate the properties of their process cheese to end users.
Factors Controlling Process Cheese Properties Formulation parameters and ingredients
As described in subsection “Process cheese manufacture,” the 1st step during process cheese manufacture involves the formulation of process cheese using various ingredients. The desired formulation of process cheese is achieved by appropriate selection of natural cheese and other ingredients as allowed by the CFR. In addition to natural cheese and other ingredients, manufacturers also select the emulsifying salts to be added to their process cheese formulation. The majority of the manufacturers in the United States also make appropriate formulation adjustments to their process cheese in order to incorporate rework (discussed in the subsection “Rework”) into their process cheese. Typically, process cheese formulations depend on the type of process cheese being manufactured and the type of end-use applications that the process cheese will be targeted for. Various chemical and compositional properties of process cheese affect the final functional properties of process cheese in a variety of ways. Therefore, while formulating a process cheese, manufacturers often try to control the final chemical properties of process cheese through appropriate selection of ingredients in order to achieve a process cheese formula that will have a specific functional property after it is manufactured. However, the availability of natural cheese (type and age), cost, availability of other ingredients, and presence or absence of rework varies from day to day. These are some of the constraints that manufacturers have to deal with while formulating their process cheeses in order to achieve a final product with consistent functional properties on a daily basis. As a result, there are numerous permutations and combinations involved in the selection of ingredients for the same process cheese formula. Traditionally, process cheese makers relied on their ex-
Vol. 7, 2008—COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
201
202
↑↑ ↓↓ ↓↓ Low meltability
a ↓,
Moderate meltability
↑, or NC indicates whether the indicated melt test value would decrease, increase, or remain the same, respectively, and the number of ↓ or ↑ is a rough indicator of the magnitude of decrease or increase.
↑↑ ↓↓ ↑↑ ↑↑ ↓↓ ↑↑ ↓↓ ↑↑ ↑↑ ↓↓ ↑ ↑↑ ↓ ↑↑ ↓↓ ↑ ↓↓ ↑↑ ↑ ↑ NC ↑ ↑ NC ↓ ↓ NC
↑ ↑ NC
↑↑ ↑↑ ↑ ↑ ↓↓ ↓ ↓↓ ↓ ↓↓ ↓ ↓↓ ↓ ↓↓ ↓↓ ↓ ↓ ↓↓ ↓↓ ↓ ↓ ↑↑ ↑ ↑↑ ↑ ↓↓ --↓↓ ↓ ↑↑ ↑↑ ↑↑ ↑ ↑↑ ↑↑ ↑↑ ↑ ↑↑ ↑↑ ↑↑ ↑ ↓ ↓a ↓↓ ↓↓ ↓
High melt, high flow High melt, moderate flow Moderate melt, high flow Moderate melt, moderate flow High melt, low flow Low melt, high flow Low melt, low flow High meltability
Arnott Tube melt Melt Hot Time Test Test temperature Softening Melting apparent at height length Diameter Area G high temp. G high temp. (tan δ = 1) temperature temperature viscosity 5000 cP Certain probable scenarios Melted texture (meltability) of process cheese
Rapid visco analyzer melt test Melt profile analysis (UW Meltmeter) Dynamic stress rheometry Schreiber Melt Test
Table 4 --- Selected scenarios pertaining to the melted textural properties of a process cheese and corresponding interpretation of the various melt tests used to measure the melted textural properties of process cheese.
CRFSFS: Comprehensive Reviews in Food Science and Food Safety perience to select the ingredient blend for a specific formula. Over the years, process cheese manufacturers have started utilizing various computer-based formulation programs where they are able to set the desired chemical properties of a process cheese formulation and the formulation software determines the ingredient blend that produces the least cost formulation. Least cost formulation simply utilizes the cost and composition of an ingredient as a criterion for developing a formulation. As an example, various whey- and milkfat-based ingredients (whey powder, dried permeate, whey protein concentrate, butter, butter oil, dried cream, and so on) can be used as ingredients in process cheese. However, each of these ingredients has a different composition and current market price. A least cost formulation program uses the cost and composition information from each available ingredient and selects the ingredients used in a formulation based on their impact on the overall cost of the formulation. One such formulation program that we use in our laboratory is TechwizardTM , which is an Excel-based formulation software program (Metzger 2003; Kapoor and Metzger 2004) provided by Owl Software (Columbia, Mo., U.S.A.). As mentioned previously, various chemical and compositional properties have an effect on the functional properties of process cheese. These include fat content (Hong 1990), moisture content (Hong 1989; Lee and others 2004), pH (Templeton and Sommer 1930, 1932a; Marchesseau and others 1997; Lee and Klostermeyer 2001), total calcium content (Cavalier-Salou and Cheftel 1991), intact casein content (Vakaleris and others 1962; Meyer ˇ etina 2003; Garimella Purna and others 2006), 1973; Piska and Stˇ lactose content (Templeton and Sommer 1932a, 1934; Thomas 1973; Berger and others 1998), and whey protein content (Gupta and Reuter 1992; Thapa and Gupta 1992a, 1992b). Due to the regulations set by the FDA (2006) for moisture and fat contents for PC, PCF, and PCS, manufacturers generally keep the moisture and fat contents in their product constant. However, due to the day-to-day variations in inventory of the natural cheese they use, the source and age of natural cheese tend to be different. These differences in the source and age of natural cheese on a daily basis lead to variations in the total calcium content, pH, and intact casein content of the process cheese and hence the functional properties of the process cheese. In addition to natural cheese variations that affect the total calcium content, pH, and intact casein content of the final process cheese, the type and amount of emulsifying salts that are added to process cheese influence the state of calcium in process cheese and the process cheese pH. Moreover, other ingredients (nonfat dried milk, dried whey, whey protein concentrate, and so on) influence the amount of whey protein and lactose in the final process cheese. The variations in the chemical properties of a process cheese that arise during a process cheese formulation significantly influence its functional properties. The type and amount of rework that is added to a process cheese formulation also have an effect on the final functional properties of process cheese. Therefore, it is important to control the formulation parameters of process cheese to achieve a product with consistent functional properties. From the discussions above, it is evident that just standardizing the moisture and fat content of a process cheese formulation does not ensure a product with the desired functional properties. It is very important for the process cheese manufacturers to control and monitor the total calcium content, intact casein content, pH, type and amount of emulsifying salts used, lactose content, whey protein content, and type and amount of rework added while formulating their process cheese in order to produce a process cheese with specific physicochemical and functional properties. The individual influence of these formulation parameters on the
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
Process cheese—a review . . . functional properties of process cheese is discussed subsequently. Total calcium content. The total calcium content of a process cheese not only plays a role during its manufacture but also influences its final functional properties. A high total calcium content in a process cheese formula leads to difficulty in manufacture of the corresponding process cheese, since more calcium needs to be sequestered from the natural cheese caseins by the emulsifying salts added during process cheese manufacture (Sood and others 1979; Caric and others 1985; Cavalier-Salou and Cheftel 1991; Zehren and Nusbaum 2000). In a study performed by CavalierSalou and Cheftel (1991) on cheese analogs using sodium caseinate, they found that as the calcium content of the cheese analogs increased, their firmness increased, and their meltability decreased. The major ingredient that contributes to the variations in the total calcium content in a process cheese formula is natural cheese. It has been observed that when a natural cheese with high total calcium content is used to make process cheese, the resulting process cheese is firm and less meltable (Olson and others 1958; Zehren and Nusbaum 2000). The effect of natural cheese on the total calcium content of process cheese and its functional properties is discussed in the subsection “Natural cheese.” Intact casein content. Meyer (1973) and Shimp (1985) highlighted the importance of total intact casein of a process cheese on the quality of process cheese. Once again, the major ingredient that contributes to the intact casein in a process cheese formula is type and age of natural cheese utilized in the formula. The intact casein content of natural cheese is inversely related to the age of the natural cheese. As a natural cheese is ripened, its intact casein content decreases (Arnott and others 1957; Vakaleris and others 1962; Meyer 1973; Garimella Purna and others 2006). This occurs because, as the natural cheese ages, the enzymes and residual starter or nonstarter lactic acid bacteria present in the cheese hydrolyze the proteins present in natural cheese into peptides, thereby reducing the amount of casein that is still present in the intact (unhydrolyzed) form. Various researchers have indicated the effect of the age of natural cheese on the functional properties such as body and texture of the resulting process cheese (Arnott and others 1957; Vakaleris and others 1962; Garimella Purna and others 2006). The effect of the age of natural cheese utilized on the functional properties of process cheese is discussed in the subsection “Natural cheese.” In a recent study performed in our laboratory, Garimella Purna and others (2006) manufactured PCF with the same cheddar cheese at 2, 4, 6, 12, and 18 wk of ripening with 2.0%, 2.5%, and 3.0% trisodium citrate as the emulsifying salt. The results indicated that as the intact casein content of the natural cheese base used for PCF manufacture decreased, its viscosity immediately after manufacture and its firmness decreased whereas its meltability increased. However, interestingly, we found that with a decrease in the intact casein content of the cheddar cheese, the flowability of the resulting PCF increased initially up to 12 wk of ripening, but in the case of PCF manufactured with the cheese at 18 wk of ripening, its flowability decreased as compared to 12 wk of ripening. We attribute this change in flow properties of the PCF with a low level of intact casein to a phenomenon called “overcreaming.” Previous researchers have reported that overcreaming occurs when the natural cheese used in a formulation is excessively ripened (Meyer 1973). The casein in natural cheese that is excessively ripened is hydrolyzed into small peptides that are easily hydrated and dispersed during process cheese manufacture; and under normal process cheese manufacturing conditions extensive protein-based interactions occur that lead to a strong protein network that has restricted flow properties (Meyer 1973; Gerimella Purna and others 2006). The concept of overcream-
ing is discussed in more detail in the “Rework” subsection of this article. pH. The final pH of a process cheese has been found to have a significant effect on the quality, microstructure, and the type of protein interactions in the resulting process cheese emulsion (Palmer and Sly 1943; Meyer 1973; Marchesseau and others 1997). Various researchers have indicated that the pH range of a good-quality process cheese should be between 5.4 and 5.8 (Palmer and Sly 1943; Marchesseau and others 1997). According to Palmer and Sly (1943), the stability of the process cheese emulsion is decreased when the pH of the process cheese is below 5.4 or above 5.8. Marchesseau and others (1997) evaluated the microstructure of process cheese manufactured with different final pH. They found that a process cheese with a lower pH (5.2) had increased protein–protein interactions since the proteins were closer to their isoelectric point, thereby promoting the aggregation of proteins leading to a weaker emulsification of the fat phase in the process cheese. At higher process cheese pH (6.1), they found that process cheese had an open structure and therefore a weaker emulsion. In their study, process cheese with a pH of 5.7 produced a uniform fat emulsion with a closely knit protein network. Consequently, the final pH of process cheese is an important factor controlling the final structure and therefore the final functional properties of the process cheese. A previous study has shown the effect of the final pH of process cheese on its firmness (Templeton and Sommer 1932b). They found that as the final pH of the process cheese increased from 5.0 to 6.2, its firmness initially increased up to approximately pH 5.8 (where it had the highest firmness); however, with a further increase in pH (5.8 to 6.2), the firmness began to decrease. Alternatively, previous work has also indicated a very low correlation between the process cheese pH and melting properties of process cheese (Arnott and others 1957). The type and level of emulsifying salts (subsection “Emulsifying salts”) (Gupta and others 1984; Shirashoji and others 2006a) and the type and age of natural cheese (subsection “Natural cheese”) used during process cheese manufacture have a marked influence on the final pH of the resulting process cheese. Although previous reports indicate the individual importance of total calcium, intact casein, and pH on the functional properties of process cheese, they do not indicate the combined effect of these factors on the process cheese properties. They also do not highlight the magnitude and nature by which these factors influence the final properties of process cheese. Currently, a study is being performed in our laboratory to evaluate the combined effect of these factors on the functional properties of process cheese. Emulsifying salts. As discussed previously, emulsifying salt is a major ingredient in process cheese manufacture. Emulsifying salts are ionic compounds made up of monovalent cations and polyvalent anions. The 2 primary functions of emulsifying salts in process cheese are “calcium sequestering” (to help disrupt the calcium–phosphate-linked protein network present in natural cheese during process cheese manufacture) and “pH adjustment.” Both of these functions help in hydrating the caseins present in natural cheese so that they can easily interact with the water and fat phases, thereby producing a homogeneous process cheese emulsion (Ellinger 1972; Meyer 1973; Caric and others 1985; Guinee and others 2004; Mizuno and Lucey 2005). According to the CFR, there are 13 emulsifying salts that are approved for use (either alone or in combination) in process cheese manufacture (21CFR133.169 to 133.180). These include mono-, di-, and trisodium phosphates, dipotassium phosphate, sodium hexametaphosphate, sodium acid pyrophosphate, tetrasodium pyrophosphate, sodium aluminum phosphate, sodium citrate, potassium citrate, calcium citrate, sodium tartrate, and
Vol. 7, 2008—COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
203
CRFSFS: Comprehensive Reviews in Food Science and Food Safety sodium potassium tartrate. The most common emulsifying salts used for process cheese manufacture in the United States are trisodium citrate and disodium phosphate. Trisodium citrate is the preferred emulsifying salt for slice-on-slice process cheese varieties, whereas disodium phosphate (or appropriate combinations of di- and trisodium phosphates) is used in loaf-type process cheese and process cheese spreads. Sometimes, low levels of sodium hexametaphosphate are also used along with these emulsifying salts (in certain applications). Another emulsifying salt that has become popular recently is sodium aluminum phosphate. It is commonly used in rennet casein-based mozzarella type imitation process cheese varieties since it provides desirable functional properties for imitation process cheese that is used to replace mozzarella on frozen pizzas. One of the most common questions that arise during the selection of emulsifying salts for process cheese manufacture is their calcium sequestering ability and their mechanisms of calcium removal from caseins. There have been various studies using different model systems to answer these questions (Cavalier-Salou and Cheftel 1991; Guinee and others 2004; Mizuno and Lucey 2005). However, the interaction of emulsifying salts with calcium and casein in process cheese is not fully understood. Important properties of different emulsifying salts and the influence of the type and amount of different emulsifying salts on process cheese properties have been extensively studied (Templeton and Sommer 1936; Ellinger 1972; Meyer 1973; Rayan and others 1980; Thomas and others 1980b; Gupta and others 1984; Caric and others 1985; Cavalier-Salou and Cheftel 1991; Molins 1991; Berger and others 1998; Awad and others 2002; Dimitreli and others 2005; Shirashoji and others 2005, 2006a, 2006b). However, due to the differences in experimental conditions among these studies, including type and age of natural cheese used, differences in process cheese formulation and composition, and differences in processing conditions used for process cheese manufacture, these studies show large variations in their results and, consequently, comparisons and interpretation of the results from these studies can be difficult. Detailed physicochemical properties of various emulsifying salts can be found in the literature (Ellinger 1972; Molins 1991; Berger and others 1998; Guinee and others 2004). Gupta and others (1984) extensively studied the effect of the type and amount of various emulsifying salts on process cheese pH, meltability, hardness, and body and texture. In Table 5, we have compiled some relevant data from previous research on the important properties of certain selected emulsifying salts along with
their influence on process cheese properties. With relevance to Table 5, Rayan and others (1980) manufactured PC (40% moisture) with 4 different emulsifying salts at a 2.5% level without pH adjustment. Their results were in agreement with those of Gupta and others (1984) with PC manufactured using trisodium citrate and PC manufactured using sodium aluminum phosphate indicating a higher meltability than the PC manufactured using disodium phosphate; however, the firmness values of the 3 PC productions were similar. Thomas and others (1980b) manufactured process cheese (45% moisture) using trisodium citrate, disodium phosphate, and sodium hexametaphosphate at a 3% addition level (without pH adjustment) and evaluated the emulsion strength (by measuring oil separation), firmness, and meltability of all the process cheeses. They found that the emulsion strengths of all the process cheeses were not significantly different. The meltability of the process cheeses manufactured using trisodium citrate and disodium phosphate were not significantly different; however, process cheese manufactured using sodium hexametaphosphate had a significantly lower meltability. They also found that the firmness of the process cheese manufactured using trisodium citrate was significantly lower than both the process cheese manufactured using disodium phosphate and sodium hexametaphosphate. A more recent study also showed the effect of the type and amount of 3 different emulsifying salts on the hardness and meltability of process cheese (Shirashoji and others 2005). Shirashoji and others (2005) manufactured PC (38% to 39% moisture and 33% fat) with 3 different emulsifying salts, trisodium citrate, disodium phosphate, and sodium hexametaphosphate, at 0.25%, 1.5%, and 2.75% of the final process cheese. All of the PC was adjusted to pH 5.6 to prevent any effect on process cheese functional properties due to pH. They found that as the concentration trisodium citrate, disodium phosphate, and sodium hexametaphosphate in PC increased its firmness increased and its meltability decreased. Also, at 2.75% emulsifying salt concentration, PC made using sodium hexametaphosphate was the most firm and the least meltable followed by PC made using disodium phosphate and PC made using trisodium citrate. In another study on the effect of type of emulsifying salts in sliced process cheese, Shirashoji and others (2006a) produced sliced process cheeses (46% moisture, 19% protein) using 4 emulsifying salts (trisodium citrate, disodium phosphate, sodium hexametaphosphate, and tetrasodium pyrophosphate, at 2.5% each) without pH adjustment. They evaluated process cheese pH, flowability, and meltability, and created a “texture map” (by performing a large deformation
Table 5 --- Physicochemical properties of some emulsifying salts and their influence on process cheese (PC) properties. Physicochemical propertiesa Emulsifying salt Trisodium citrate (dihydrate) Monosodium phosphate (monohydrate) Disodium phosphate (dihydrate) Trisodium phosphate (dodecahydrate) Dipotassium phosphate Sodium hexametaphosphate Sodium aluminum phosphate
Chemical formula NaH 2 C 6 H 5 O 7 .2H 2 O NaH 2 PO 4 .H 2 O Na 2 HPO 4 .2H 2 O Na 3 PO 4 .12H 2 O K 2 HPO 4 (NaPO 3 ) n (n = 10 to 15)
Influence on PC propertiesc
Formula Solubility, g/100 pH, 1% pH of Hardness Meltability (kg) (mm) weight, g/mol g H 2 O at 20 ◦ C solution the PC 294 138 178 380 174 (102) n
75 85b 80 11b 160 157
a Data compiled from Mollins (1991), Berger and others (1998), and Guinee and others (2004). b Solubility of the anhydrous salt. c Data compiled from Gupta and others (1984). (The values indicated were true for their experimental
8.6 4.5 9.1 11.9 8.9 6.6 9.2
5.9 5.1 5.8 7.3 5.9 5.2 5.9
32 27 32 26 29 33 33
131 NMd 70 70 76 NM 101
setup and should be treated as general reference, not as universal results.) rProcess cheese (PC) manufactured using 75% young cheddar cheese and 25% aged cheddar cheese with moisture ranging from 38.4% (for PC with sodium aluminum phosphate) to 40.7% (for PC with trisodium phosphate) and fat approximately 31.5%. rThe emulsifying salt level for all the PC were 2.2% except for PC with disodium phosphate (which was 2.1%) and PC with trisodium citrate (which was 2.3%). rHardness values are rounded to the nearest whole number. rMeltability was measured using tube melt test. d No melt observed.
204
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
Process cheese—a review . . . burst test using a ball-type probe) of all the process cheeses. The pH of the process cheese made using sodium hexametaphosphate was significantly lower (pH 5.3) than the other process cheeses (pH 5.9 to 6.0). The meltability and flowability results showed that process cheese manufactured using tetrasodium pyrophosphate had the least meltability and lowest flowability followed by process cheese made using sodium hexametaphosphate. The meltability and flowability of the process cheeses made using trisodium citrate and disodium phosphate were similar. The results of the “texture map” indicated that process cheeses made using disodium phosphate and sodium hexametaphosphate tended toward a mushy and crumbly texture, whereas tetrasodium pyrophosphate provided the process cheese with a tough and rubbery texture (Shirashoji and others 2006a). In a study on the influence of the level of trisodium citrate on PCF properties performed in our laboratory, Garimella Purna and others (2006) manufactured PCF (44% moisture and 25% fat) using cheddar cheese (at 2, 4, 6, 12, and 18 wk of ripening) and trisodium citrate (at 2.0%, 2.5%, and 3.0% levels). All the PC batches were processed at 85 ◦ C for 6 min at 2 different mixing speeds (450 and 1050 rpm). The viscosity was immediately measured after manufacture and the firmness, RVA hot apparent viscosity (a measure of the flowability of process cheese, Table 4), and RVA time at 5000 cP (a measure of the meltability of process cheese, Table 4) after PC manufacture. The results indicated a significant effect of the concentration of trisodium citrate on the RVA hot apparent viscosity and RVA time at 5000 cP. As the trisodium citrate level of the PCF increased, both the flow properties and the meltability of the PCF decreased; however, there was no significant effect of trisodium citrate concentration on the firmness of PCF. Lactose content. The lactose content of a process cheese is another critical formulation parameter that needs to be controlled in a process cheese formula since a high level of lactose in process cheese can lead to the formation of lactose crystals or Maillard browning in process cheese. Nonfat dried milk (NDM) and dried whey are the major ingredients that contribute lactose to a process cheese formula. The problem of lactose crystallization in process cheese due to the addition of NDM or whey powder has been addressed by various researchers (Templeton and Sommer 1932a, 1934; Thomas 1973; Berger and others 1998). Lactose crystallization in process cheese depends on the maximum concentration of lactose that is soluble in the water phase of process cheese (Templeton and Sommer 1932a; Thomas 1973). The maximum concentration of lactose that is soluble in water is 17% at 20 ◦ C (Templeton and Sommer 1932a; Harper 1992). Hence, as a general guideline, it is important to maintain the amount of lactose in the water phase of process cheese at less than 17% in order to avoid lactose crystallization. Therefore, when formulating a process cheese, the manufacturers should ensure that the final lactose content in the process cheese should not exceed 7.48% for PCF (44% moisture product) and 10.20% for PCS (60% moisture product). Another defect that tends to arise due to the addition of lactose-rich ingredients in process cheese is Maillard browning, which leads to objectionable color and flavor development (Thomas 1969). Thomas (1969) indicated that postmanufacture storage temperature and time as well as the pH of the process cheese significantly affected the browning of process cheeses. He suggested that process cheese should not be stored at temperatures greater than 35 ◦ C for more than 6 wk. Whey protein content. Whey proteins constitute approximately 20% of the total proteins in bovine milk (Eigel and others 1984). Approximately 80% of whey proteins in milk are made up of the 2 major whey proteins (β-lactoglobulin and α-lactalbumin). One of the important characteristics of β-lactoglobulin from a processing standpoint is the presence of a “reactive” free sulfhydryl group in
its primary structure (Wong and others 1996). Whey proteins are also highly susceptible to heat treatment and are found to denature between 60 and 70 ◦ C. This temperature-induced denaturation of β-lactoglobulin exposes the free sulfhydryl group, which has the capability of crosslinking with other β-lactoglobulin and κ-casein molecules via disulfide bonds (Sawyer and others 1963; Wong and others 1996). NDM and whey protein concentrate (WPC), if used in a process cheese formula, may contribute to increased levels of whey protein in the process cheese. Since whey proteins can crosslink among themselves as well as with caseins at high temperatures, a high level of whey proteins in a process cheese formula not only influences its sensory properties but also may lead to an increase in the firmness of the final process cheese and a decrease in its meltability. The influence of whey protein incorporation in process cheese on its functional and sensory properties has been extensively studied (Gupta and Reuter 1992; Thapa and Gupta 1992a; Ido and others 1993; Abd El-Salam and others 1996; Al-Khamy and others 1997; Fayed and Metwally 1999; Mleko and Foegeding 2000, 2001; Laye and others 2004). Gupta and Reuter (1992) ultrafiltered whey to produce a liquid concentrate (26% total solids) with 20% protein and 5.8% lactose which was utilized as an ingredient to replace 20% of the solids in a PCF formula. They determined that addition of up to approximately 8% whey protein in the final PCF with average moisture content of 47% did not have an effect on the overall acceptability score of process cheese. However, in another study, Thapa and Gupta (1992a) have indicated that PCF (42% to 43% moisture, 2.5% emulsifying salt) containing WPC (at approximately the same final whey protein level) was firmer than PCF with no added WPC. Abd-El-Salam and others (1996) studied the effect of liquid WPC (28% total solids, 15% whey protein) addition (0%, 20%, and 40% of the final blend) on the compositional and the rheological properties of processed cheese spread (57% moisture, 3% emulsifying salt). In their study, addition of WPC increased the moisture by 0.8%, lactose by 2.5%, and pH by 0.3 in the final cheese spreads made from 40% WPC (6% whey protein in the final product) as compared to cheese spreads with no WPC added. They found an improved meltability (possibly due to the increase in moisture content), flavor, and other sensory properties of the PCS as the amount of WPC in the cheese spreads increased. In a recent study with a rennet casein process cheese model system (17% casein, 24% fat, 2% emulsifying salt), Mleko and Foegeding (2000) found that up to a maximum of 2% casein can be replaced with whey protein. However, they did find a slight increase in the firmness and a decrease in the meltability of the process cheese. They proposed that the heat-induced disulfide interactions involving the free sulfhydryl groups of β-lactoglobulins have a marked influence on the firmness and the melting properties of the process cheeses. Additional work by these researchers involved replacement of casein proteins with polymerized whey proteins in a rennet casein-based model process cheese system (Mleko and Foegeding 2000, 2001). They produced polymerized cross-linked whey proteins by heating whey protein dispersions to induce the disulfide cross-links between the whey proteins and added these polymerized whey proteins to a model process cheese system. Their work showed that, as the level of polymerized whey proteins was increased, there was an increase in the firmness and a decrease in the meltability of the process cheese analogs produced. However, they did conclude that replacement of 4% rennet casein (in a 17% rennet casein-based process cheese analog system) with 2% whey protein polymers could still produce process cheese analogs of identical texture and meltability (Mleko and Foegeding 2001). Rework. “Rework” is a term used to describe process cheese produced in a manufacturing facility that cannot be sold for a variety of reasons. The type of rework ranges from process cheese
Vol. 7, 2008—COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
205
CRFSFS: Comprehensive Reviews in Food Science and Food Safety lost during production line changeovers; shavings and edge trimmings removed during slice line operations; residual process cheese that is removed from the cookers, lines, hoppers, and packaging machines, also referred to as “hot melt” (Kalab and others 1987); and process cheese that has been rejected by quality control due to improper weight, packaging, or on the basis of a quality defect (Kichline and Scharpf 1969; Zehren and Nusbaum 2000). Since rework poses economic challenges to manufacturers, it is incorporated into a fresh process cheese blend and reprocessed. According to Lauck (1972), the amount of rework produced in a process cheese manufacturing facility ranges from 2% to 15% of the total process cheese produced. Since rework has already undergone the emulsification process and also contains emulsifying salts, the addition of rework to a fresh blend during process cheese manufacture can cause difficulty during processing and also affects the final functional properties of process cheese (Kichline and Scharpf 1969; Lauck 1972; Kalab and others 1987). In general, the addition of rework tends to decrease the meltability and produce a firmer process cheese (Kalab and others 1987). Kichline and Sharpf (1969) specify that the maximum amount of rework that can be added to process cheese without any undesirable effects on its properties is 4% of the total cheese blend used for process cheese manufacture. Kalab and others (1987) found that the types of rework, as well as the amount used, have an effect on the final properties of PCF (43% moisture, 24% fat, pH 5.5 to 5.7, and 2.7% added emulsifying salt in the form of trisodium citrate). They studied the effect of type of rework on the apparent viscosity of PCF immediately after manufacture, its firmness, and its meltability. In their study, they used 3 sources of rework at 20% of the final PCF: fresh rework (quickly frozen PCF emulsion right after manufacture), regular PCF slices from older processing runs, and overcooked process cheese mass to simulate process cheese that had undergone excessive processing (which they referred to as “hot melt”). The “hot melt” was used at 2 levels: 10% and 20% of the final product. Their results indicated that when any kind of rework was used in PCF blends, the manufactured PCF was more viscous coming out of the cooker and had higher firmness and lower meltability relative to the control. They also found that the effect of the type of rework on the increase in apparent viscosity after manufacture and the firmness of the PCF was fresh rework < PCF slices from older process cheese runs < hot melt when used at 10% of the total blend < hot melt when used at 20% of the total blend. A similar trend was observed for the decrease in meltability of the PCF with the PCF with hot melt at 20% showing no melt. The microstructural results from their study showed that hot melt samples had a denser protein matrix and were overemulsified when compared to fresh rework samples (Kalab and others 1987). This phenomenon where excessive cooking of process cheese can increase the interactions among the caseins to such an extent that they attain a thick pudding-like consistency is also referred to as “overcreaming” (Meyer 1973). In Europe, “precooked cheese” or rework obtained at different times during process cheese manufacture is sometimes used for enhancing the cooking and functional properties of process cheese (Meyer 1973; Berger and others 1998). According to Meyer (1973), fresh rework, which has a weakly dispersed (hydrated) protein structure (also referred to as “long structure”), can be effectively used to stabilize a process cheese that might show a tendency to overcream under normal cooking conditions. Normal rework with optimum protein dispersion and emulsification can be used from 2% to 30% in process cheeses where the creaming action is desired. However, Meyer (1973) cautions that hot melt should not be used at more than 1% of the process cheese blend in order to avoid adverse effects on the functional properties of the process cheese. In the United States, it is a general practice to collect shavings and edge trimming from slice-on-slice 206
production lines and immediately use them as “fresh rework” at approximately 10% of the final blend. The importance of controlling various formulation factors in a process cheese formula has been described previously. Also discussed previously are some of the important ingredients that affect these formulation parameters in a process cheese formula (for example, natural cheese affects the total calcium, intact casein, and pH; nonfat dried milk and whey ingredients affect the lactose content and whey protein level). Since the manufacturers use these ingredients to control the above-mentioned process cheese formulation parameters, the effect of natural cheese and other ingredients on process cheese properties is discussed subsequently. Effect of ingredients. Natural cheese. Natural cheese is one of the most important ingredients used in process cheese. As discussed previously, natural cheese has a marked influence on total calcium, intact casein, and pH and, hence, the final functional properties of process cheese. Depending on the country of manufacture, availability, and market demand, various types of natural cheeses such as cheddar, Swiss, Gouda, and so on, are used to manufacture process cheese (Meyer 1973). However, in the United States, cheddar cheese is the major type of natural cheese used for process cheese manufacture. Depending on the type of process cheese manufactured, the amount of natural cheese in a process cheese formula varies from 51% to > 80% of the final process cheese (FDA 2006). Consequently, the characteristics of natural cheese utilized to manufacture process cheese have a major influence on process cheese characteristics and appropriate selection of natural cheese is critical in order to achieve a process cheese with the desired chemical and functional characteristics. The natural cheese used in a process cheese formulation is generally selected on the basis of type, flavor, maturity, consistency, texture, and pH (Zehren and Nusbaum 2000). Process cheese manufacturers, through years of experience, have realized and mastered the art of selecting the appropriate blend of young and aged natural cheese in order to achieve process cheese with desired flavor and textural properties (Meyer 1973; Thomas 1973). Numerous researchers have highlighted the importance of natural cheese characteristics on functional properties such as unmelted texture and meltability of process cheese (Barker 1947; Meyer 1973; Thomas 1973; Caric and others 1985; Shimp 1985; Zehren and Nusbaum 2000). The physicochemical characteristics of natural cheese that influence the process cheese properties include pH, calcium and phosphorus contents, and age, or the amount of intact casein present in the natural cheese (Templeton and Sommer 1930; Barker 1947; Olson and others 1958; Vakaleris and others 1962; Meyer 1973; Thomas 1973; Harvey and others 1982; Zehren and Nusbaum 2000; Kapoor and others 2007). The importance of natural cheese pH on process cheese properties has been highlighted in a study performed by Olson and others (1958) in which they manufactured cheddar cheeses with modified manufacturing protocol so as to produce 2 cheddar cheese treatments with different final pH levels. The 2 cheddar cheeses were then used to manufacture PCS (at 10, 30, 60, 90, and 150 d of ripening), which were analyzed for unmelted texture using penetrometry and meltability using the tube melt test. Their results indicated that even after the final pH of the PCS was adjusted to 5.4 to 5.5, the PCS batches made using cheddar cheese with higher pH were harder and less meltable when compared to the PCS made using cheddar cheese with normal pH at all stages of ripening. However, the observed effect of natural cheese pH on process cheese properties was also related to the changes in the state and amount of calcium and phosphorus caused by the differences in natural cheese pH (Olson and others 1958; Zehren and Nusbaum 2000). The effect of natural cheese pH on the state
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
Process cheese—a review . . . and amount of calcium and phosphorus in natural cheese is discussed subsequently. Minor changes in the manufacturing protocols during natural cheese manufacture, such as set pH, drain pH, and level of salting (salt-to-moisture ratio), can significantly change the state and amount of calcium and phosphorus in natural cheese (Dolby and others 1937; Czulak and others 1969; Upreti and Metzger 2006). In a study on low-fat mozzarella cheese, Metzger and others (2000) studied the effect of cheese milk set pH on the amount and state of calcium in the final cheese. It was found that as the pH of the cheese milk before rennet addition during cheese manufacture (set pH) was reduced to 6.0 or 5.8 (using acetic acid), the total calcium of the final cheeses decreased by 11% (set pH 6.0) or 23% (set pH 5.8), respectively, when compared to the control cheese that was set at a pH of 6.5. Czulak and others (1969) highlighted the effect of drain pH of cheddar cheese on the calcium content and the final pH of the cheese. They found that as the pH of the curd during whey drainage was decreased from 6.14 to 5.75, there was a 27% reduction in the total calcium content of the cheese curd at the time of whey separation. As discussed in the subsection “Intact casein content,” the intact casein content of natural cheese is inversely related to the age of the natural cheese. As a natural cheese is ripened, its intact casein content decreases (Fenelon and Guinee 2000; Garimella Purna and others 2006). Researchers have described the effect of the age of natural cheese on the functional properties of process cheese (Templeton and Sommer 1930; Arnott and others 1957; Olson and others 1958; Vakaleris and others 1962; Piska and ˇ etina 2003; Garimella Purna and others 2006). All these studStˇ ies consistently indicated that as the age of natural cheese used in process cheese manufacture increased, the unmelted firmness of the resulting process cheese decreased (Templeton and Sommer 1930; Olson and others 1958; Vakaleris and others 1962; ˇ etina 2003; Garimella Purna and others 2006) and Piska and Stˇ the meltability of the resulting process cheese increased (Olson and others 1958; Vakaleris and others 1962; Garimella Purna and others 2006). Currently, another major thrust in the natural cheese industry is the utilization of concentrated milk to manufacture natural cheeses in order to increase the throughput of cheese plants. The type of concentration technique and the extent to which milk has been concentrated also influence the final pH, calcium content, and degree of proteolysis in the natural cheese (Sutherland and Jameson 1981; Anderson and others 1993; Acharya and Mistry 2004; Nair and others 2004). Consequently, utilization of these natural cheeses (manufactured using concentrated milk) as an ingredient in process cheese can also influence the functional properties of the process cheese manufactured using this natural cheese as an ingredient (Acharya and Mistry 2005). Acharya and Mistry (2005) manufactured PC using 5 different cheddar cheese treatments that were manufactured from cheese milk concentrated with different concentration techniques at different levels. The treatments were cheddar cheeses manufactured using control (normal cheese milk), ultrafiltered milk (concentration factor of 1.5×), ultrafiltered milk (concentration factor of 2.0×), vacuumcondensed milk (concentration factor of 1.5×), and vacuumcondensed milk (concentration factor of 2.0×). They found that, as the concentration factor of the milk utilized to manufacture cheddar cheese was increased to 1.5×, the calcium content of the cheddar cheese manufactured increased by 10% when the milk was ultrafiltered and by 4% when the milk was vacuumcondensed. Moreover, when the concentration factor of the milk utilized to manufacture cheddar cheese was increased to 2.0×, the calcium content of the cheddar cheese manufactured increased by 18% when the milk was ultrafiltered and by 13% when the milk was vacuum-condensed. Moreover, they found
that the degree of protein hydrolysis in the cheddar cheeses was also affected by both the method and the level of concentration. When these natural cheeses were used to make PC, the hardness of PC increased and the melt and the flow properties of PC decreased, as the level of concentration of the cheese milk to manufacture the natural cheese increased (in both the concentration techniques). Consequently, utilization of natural cheeses manufactured using concentrated milk as an ingredient in process cheese can also influence the functional properties of the process cheese manufactured. Sometimes, it is also common for process cheese manufacturers to freeze natural cheese that will eventually be used for process cheese manufacture (Zehren and Nusbaum 2000). Thomas and others (1980a) studied the effect of frozen natural cheese on process cheese functional properties. They manufactured 3 process cheese treatments utilizing the same natural cheese (regular or frozen). The natural cheese was split into 3 parts. One part was ripened for 6.5 mo at 10 ◦ C, the 2nd part was ripened for 6.5 mo and then frozen at −20 ◦ C for 3 mo before being used to make process cheese, the 3rd part was ripened for 3.5 mo followed by freezing at −20 ◦ C for 3 mo and further repining for 3 mo. They found that when frozen natural cheese (in all the freezing treatments) was used to manufacture process cheese, the firmness of the resulting process cheese was higher and its meltability was lower when compared to the process cheese manufactured using the same natural cheese which was not frozen. Nonfat dried milk/dried whey/whey protein concentrate. The levels of dairy-based ingredients other than natural cheese used in PC are specifically defined by the CFR (Table 1). Ingredients such as NDM and whey-based dairy ingredients such as liquid whey, whey powder, and WPC can be used in PCF and PCS (FDA 2006). Since the addition of these ingredients to process cheese formulation helps to reduce the cost of the product, manufacturers often try to maximize the addition of NDM and wheybased dairy ingredients in their products. The amounts of these ingredients typically added to process cheese are not known to cause significant changes in process cheese properties. However, since commercial NDM and dried sweet whey have an approximate lactose concentration of 50% and 75%, respectively, and commercial NDM and WPC have a significant amount of whey proteins, 2 important formulation factors need to be taken into account when using NDM and other whey-based ingredients in PCF and PCS manufacture. These factors are the level of lactose and the level of whey protein (as discussed in subsections on lactose content and on whey protein content) provided by these ingredients in the final process cheese. Food gums/hydrocolloids. The CFR allows gums or hydrocolloids to be used in PCS at levels not exceeding 0.8% of the finished product (Table 1). These include carob bean gum, gum karaya, gum tragacanth, guar gum, gelatin, sodium carboxymethylcellulose (cellulose gum), carrageenan, oat gum, algin (sodium alginate), propylene glycol alginate, or xanthan gum singly or in combination. Gums do not directly affect any of the above-mentioned formulation parameters in the process cheese; however, since PCS has a high moisture content (up to 60%), the major function of gums in PCS is to bind water and to provide appropriate viscosity/thickening to the product and improve its mouthfeel. Therefore, gums in a PCS formula have an effect on melted textural properties. Zehren and Nusbaum (2000) indicate that the selection of gums depends on various factors. These include ease of dispersibility, solubility, hydration behavior, moisture holding ability, cook viscosity, compatibility with milk proteins and other compounds present in process cheese, and optimum working pH range. Phillips and Williams (2000) provide extensive information on the properties and uses of different gums used in the food industry. However, there is a lack of available literature dealing
Vol. 7, 2008—COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
207
CRFSFS: Comprehensive Reviews in Food Science and Food Safety with the effects of various gums on the final properties of PCS. Another important area where the use of gums is gaining popularity is in low-fat and reduced-fat process cheese and imitation process cheese varieties (Brummel and Lee 1990; Swenson and others 2000). As described in the subsection “Process cheese manufacture,” after the preparation of a desired formulation, the ingredient blend is processed using heat and mixing to produce a homogeneous mass, which is packaged and cooled. Although effective monitoring and control of the formulation parameters are critical to ensure the production of process cheese with specific functional properties, appropriate selection of the process cheese cook conditions is also very important, since differences in processing conditions during process cheese manufacture have a major influence on the functional properties of process cheese. The effect of processing conditions on process cheese functional properties is discussed subsequently. Processing conditions
Processing parameters such as cook temperature (Lee and others 1981; Hong 1989; Berger and others 1998; Glenn and others 2003), cook time (Rayan and others 1980; Bowland and Foegeding 2001; Glenn and others 2003; Shirashoji and others 2006b), the amount of mixing provided during manufacture (Glenn and others 2003; Garimella Purna and others 2006), and the rate ˇ etina 2003; at which the process cheese is cooled (Piska and Stˇ Zhong and others 2004) also play a major role in controlling the emulsion formation and the resulting functional properties of process cheese. Glenn and others (2003) at North Carolina State Univ. performed an extensive study to evaluate the effect of processing conditions on process cheese meltability. They used 5 mixing speed/ cook temperature combinations (50 rpm/74 ◦ C, 50 rpm/86 ◦ C, 100 rpm/80 ◦ C, 150 rpm/74 ◦ C, and 150 rpm/86 ◦ C) to manufacture processed cheddar cheese. Process cheese at each mixing speed/cook temperature combination was processed for 1, 5, 10, 15, 25, and 35 min. The Schreiber melt test was performed on all the process cheeses to measure their meltability. They calculated the time–temperature effect (thermal history) and the time– shear effect (strain history) for each of the 30 mixing speed–cook temperature–cook time combinations that were used to manufacture process cheeses (as discussed previously) and finally correlated the meltability of the process cheese with the thermal history and strain history. The thermal history of process cheeses ranged from 24 MJ·s/kg for the process cheese manufactured at 74 ◦ C for 1 min to 886 MJ·s/kg for the process cheese manufactured at 86 ◦ C for 35 min. The strain history of process cheeses ranged from 807 for the process cheese manufactured at 50 rpm for 1 min to 84776 for the process cheese manufactured at 150 rpm for 35 min. They found that as the thermal history and the strain history of process cheese increased its meltability decreased, which indicates that an increase in cook temperature, cook time, and mixing speed during manufacture produces process cheese with lower meltability. The individual effects of processing temperature, processing time, mixing speed during manufacture, and rate of cooling of process cheese after manufacture on the functional properties of process cheese are discussed subsequently. Processing temperature. Although the minimum cook temperature and time specified by CFR for process cheese is 65.5 ◦ C for 30 s (FDA 2006), process cheese manufacturers use various types of cookers with different designs and operating conditions to manufacture process cheese. Consequently, the cook temperatures utilized range from 70 to > 100 ◦ C depending on the cooker design and the variety of process cheese manufactured. Lee and others (1981) manufactured processed Emmental cheese 208
at 4 different cook temperatures (80, 100, 120, and 140 ◦ C) and subsequently analyzed the firmness (using penetrometry) and the microstructure of the product. They found that, as the cook temperature during process cheese manufacture increased, the firmness of the process cheese increased and strength of the process cheese emulsion increased. Processing time. Rayan and others (1980) manufactured PC (40% moisture) with 4 different emulsifying salts (trisodium citrate, disodium phosphate, tetrasodium pyrophosphate, and sodium aluminum phosphate) at a 2.5% level. They used a processing temperature of 82 ◦ C with processing times of 6, 11, 16, 26, and 46 min and subsequently measured the meltability, firmness, degree of elasticity, and microstructure of all the PC batches. They found that as the processing time of the PC increased, there was a significant increase in the firmness and degree of elasticity and a significant decrease in their meltability. The results held true for all 4 emulsifying salts. The microstructural results from their study showed that as the processing time of the PC increased, there was a decrease in the size of fat globules, thereby indicating a stronger emulsification in PC with increasing processing time. In another study on PC manufactured at 80 ◦ C for 10, 20, and 30 min using cheddar cheese and 2.75% trisodium citrate (38.5% to 40.1% moisture, 32% fat), Shirashoji and others (2006b) showed that as the processing time of the PC increased, its firmness increased and its meltability decreased. Mixing speed during processing. In a study performed in our laboratory, Garimella Purna and others (2006) manufactured PCF (44% moisture and 25% fat) using the same cheddar cheese base (at 2, 4, 6, 12, and 18 wk of ripening) and trisodium citrate (at 2.0%, 2.5%, and 3.0% levels). All the PC batches were processed at 85 ◦ C for 6 min at 2 mixing speeds (450 and 1050 rpm). We found a significant effect of the mixing speed on the viscosity immediately after manufacture, on firmness, on flow properties, and on the meltability of the PCF at all natural cheese ripening and all trisodium citrate levels. As the mixing speed during PCF manufactured increased, there was an increase in the viscosity immediately after manufacture and the firmness of the PCF, and there was a decrease in the flow properties and meltability of the PCF. In a subsequent study on the effect of mixing speed on the microstructure of PCF, we performed cryoscanning electron microscopy on the PCF (with 2.5% trisodium citrate) manufactured by Garimella Purna and others (2006). Figure 4 indicates the cryoscanning electron microscopy images and the distribution in fat globule diameters of the 2 PCF batches. Figure 5 shows the number of fat globules/100 µm2 and mean fat globule diameters of the 2 PCF batches. The results clearly indicate that at high mixing speed, the PCF showed a larger number of fat globules/100 µm2 , a lower mean fat globule diameter, and a more uniform distribution in fat globule diameter as compared to PCF manufactured at low mixing speeds. ˇ etina (2003) Rate of cooling after manufacture. Piska and Stˇ manufactured PCS-type products using a blend of Dutch-type hard and semihard cheeses. These PCS products cooled at 2 different rates: slow cooling (where the PCS reached a temperature of 20 ◦ C in approximately 50 h) and fast cooling (where the PCS reached a temperature of 20 ◦ C in less than 1 h and a temperature of 5 ◦ C in 2 h). They found that PCS that was cooled slowly was significantly firmer and had significantly higher adhesiveness and gumminess. In another study on commercial PC samples, Zhong and others (2004) continuously measured the G (storage modulus) of the PC which was subjected to different cooling rates (0.025, 0.05, 0.1, and 0.5 ◦ C/min). They found that as the rate of cooling of PC decreased, the G increased, indicating that PC cooled at a slower rate was firmer. In the same study, they cooled 5-pound PC loaves at the same rate using 2 types of
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
Process cheese—a review . . . coolers (free convection and forced convection) and measured the slicing ability and meltability of the PC. They theorized that since the surface of the loaf would cool faster (irrespective of the cooler type) when compared to the center, there should be differences in the functional properties of the PC samples from the surface and the center. However, when they calculated the cooling rate at different locations in the 5-pound loaf under their cooler conditions, they found that the cooling rates (under their conditions) of the surface and the center of the loaf were not very different. They also found no trend between the sampling location of the PC sample in the loaf and the slicing ability and the meltability of the PC. However, they found a significant effect of the type of cooler used on the functional properties of the PC after cooling. PC samples cooled in the forced convection cooler were less firm and more meltable than the PC cooled in the free convection cooler. Effective control of the formulation and processing parameters during the manufacture of process cheese ensures the production of a good-quality process cheese that is free of defects. However, there still are some occasions when certain chemical, textural, or microbiological defects arise in the final process cheese, which make the product unfit for sale. In the following 2 sections, we discuss some of the possible defects that are associated with process cheese.
Defects in Process Cheese Defects related to appearance Crystal development. Researchers have shown various in-
stances of different crystals in natural cheeses. These range from tyrosine, calcium lactate, and lactose to various salts of calcium phosphate and calcium citrate (reviewed by Caric and others 1985; Guinee and others 2004). Concentrating on process cheese, researchers over the years have shown instances and even identified different types of crystals using microscopy, Fourier transform infrared spectroscopy, and x-ray diffraction (Sommer 1930; Scharpf and Michnick 1967; Scharpf and Kichline 1968, 1969; Rayan and others 1980; Uhlmann and others 1983; Klostermeyer and others 1984; Caric and others 1985; Pom-
mert and others 1988). The types of crystals identified in process cheese ranged from salts of calcium tartrate (Sommer 1930; Leather 1947) (although tartrate crystals are not commonly found anymore since tartrate-based emulsifying salts are not typically used today), calcium citrate (Morris and others 1969; Scharpf and Kichline 1969), tertiary complexes of sodium and calcium citrates (Klostermeyer and others 1984), various salts of sodium and calcium phosphates (Scharpf and Kichline 1968; Pommert and others 1988), and, sometimes, calcium salts of free fatty acids (Bester and Venter 1986). Other types of crystals that can occur in process cheese are lactose crystals (discussed previously) (Templeton and Sommer 1932a). It has already been discussed previously (subsection “Lactose content”) that lactose crystallization in process cheese can be avoided by maintaining the level of lactose below its maximum solubility level in the water phase of process cheese. However, major sources of crystal development in process cheese are due to the use of emulsifying salts such as various salts of phosphates and citrates that lead to the formation of various complexes/salts that might have a lower solubility than the regular emulsifying salts. The solubility of these complexes/salts is further influenced by pH of the process cheese or storage conditions. Scharpf and Kichline (1968) developed a multiple regression model and an equation involving prediction of crystal formation in process cheese (40% to 42% moisture). They showed that crystal formation was directly dependent on the pH and the amount of total phosphorus in process cheese (measured as % P 2 O 5 content and involving the sum total of natural phosphorus present in the process cheese and phosphorus coming from the emulsifying salts). They showed that the chances of crystal formation increased with increasing pH as well as increasing phosphorus content. The equation has been shown to be effective in predicting crystal formation in process cheese made using various phosphate-based emulsifying salts (Zehren and Nusbaum 2000). Scharpf and Kichline (1968) also showed that crystal formation accelerated when the process cheese surface was directly exposed to cold air. Therefore, careful selection of emulsifying salts during process cheese manufacture and proper storage conditions of the final process cheese are important to prevent crystal development (Berger and others 1998). Manufacturing equipment such as casting lines and slice cooling Figure 4 --- Cryoscanning electron microscopy images of 2 PCF samples (with 2.5% trisodium citrate) manufactured at (a) 1050 rpm; and (b) 450 rpm with fat globules indicated as F and protein matrix indicated as P. Also indicated is the distribution in fat globule diameter (a1) PCF manufactured at 1050 rpm; (b1) PCF manufactured at 450 rpm.
Vol. 7, 2008—COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY
209
CRFSFS: Comprehensive Reviews in Food Science and Food Safety belts have also been associated with promoting crystal development in process cheese as they can act by providing nucleation sites for crystallization when not cleaned properly and regularly (Berger and others 1998). Color defects. Browning. As discussed previously (subsection “Lactose content”), various factors influence browning (Maillard reaction) in process cheese. Browning in process cheese is initiated when ingredients with high lactose content are used during process cheese manufacture (Thomas 1969, 1973). Additionally, high process cheese final pH (> 5.9) (Thomas 1969) and high storage temperatures (35 to 37 ◦ C) (Thomas 1969; Kristensen and others 2001) have been found to accelerate browning in process cheese. Various natural cheese compositional factors also have an influence on the browning of process cheese manufactured from it. Bley and others (1985) found a high correlation between the salt-to-moisture ratio as well as the levels of residual galactose and lactose of cheddar cheese on the browning of the process cheese manufactured from it. Therefore, careful selection of ingredients
and optimum final pH during process cheese manufacture, along with desirable postmanufacture storage conditions, needs to be adhered to in order to prevent browning defects in the process cheese. Pink discoloration. Zehren and Nusbaum (2000) indicated the presence of pink discoloration in process cheeses that were either artificially colored with annatto or used a natural cheese, which had annatto color added to it. They further indicated that use of alkaline extracts of annatto in process cheese showed a higher propensity to cause pink discoloration in process cheese. Shumaker and Wendorff (1998) evaluated the effect of processing temperature, type and amount of emulsifying salt, amount of colored natural cheese in the process cheese blends, and the type of annatto colorant used on pink discoloration of process cheese. They found that with an increase in the amount of colored natural cheese in the blend there was an increase in the pink discoloration of the resulting process cheese. They also found that when the ratio of aged cheese (uncolored) was increased in the
Mean fat globule diameter (µm)
3.50
Figure 5 --- Effect of mixing speed on the microstructural attributes of PCF manufactured with 2.5% trisodium citrate.
3.00 2.50 2.00 1.50 1.00 0.50 0.00 1050
450
Mixing speed (rpm)
Number of fat globules / 100 µm2
35
30
25
20
15
10
5
0
1050
450 Mixing speed (rpm)
210
COMPREHENSIVE REVIEWS IN FOOD SCIENCE AND FOOD SAFETY—Vol. 7, 2008
Process cheese—a review . . . natural cheese blend (uncolored) during process cheese manufacture with added annatto colorant, there was an increase in the pink discoloration. Shumaker and Wendorff (1998) also found that annatto emulsion-based colorants were more susceptible to pink discoloration than annatto suspension-based colorants.
process cheese are of the genus Clostridium (Kautter and others 1979; Sinha and Sinha 1988; Glass and Doyle 2005). A few outbreaks involving botulism from the consumption of canned PCS have been reported over the years (Glass and Doyle 2005). All the PCS associated with these outbreaks were found to have high water activity (about 0.96 to 0.97) and high pH (about 5.7 to 5.8), which might have led to the production of toxins produced by Functional defects and defects in body and texture Functional defects of a process cheese are not defects in the true Clostridium botulinum in the PCS during storage (Briozzo and sense. Functional defects in a process cheese can be referred to others 1983; Glass and Doyle 2005). as the inability of the process cheese to demonstrate and perform the desired functional properties and end-use behavior for which Factors that control microbiological spoilage Canned PCS-type products can be classified as low-acid it has been manufactured. These can range from very high to very low meltability, very high to very low firmness, presence of stick- canned foods (21CFR113). Therefore, anaerobic sporeformers iness, and so on. These properties can be adjusted to the desired such as Clostridium spp. are a major concern. According to the levels by appropriate control of various formulation and process- regulations set forth by the CFR, all low-acid canned foods need to ing parameters in process cheese (section “Factors controlling be subjected to a process to render it commercially sterile. Thereprocess cheese properties”). In addition to these functional de- fore, these foods either need to be subjected to heat-processing fects, process cheese is also prone to defects in body and texture, so that there is a 12-log reduction in the botulinal spores or there including brittle, crumbly, and grainy texture, and oil separation. needs to be appropriate formulation changes as well as adjustBrittle, crumbly, and grainy texture. This type of defect generally ments in pH and water activity in these products in order to inhibit arises when the final pH of the process cheese is too low (