Non-Thermal Processing of Skim Milk: Impact on Microbial Reduction

Non-thermal processing of skim milk: Impact on microbial reduction,

physico-chemical properties and quality of Brie type cheese

by Dipendra Khanal

A Thesis presented to The University of Guelph

In partial fulfillment of requirements for the degree of Master of Science in Food Science

Guelph, Ontario, Canada © Dipendra Khanal, May, 2014

ABSTRACT

NON-THERMAL PROCESSING OF SKIM MILK: IMPACT ON MICROBIAL REDUCTION, PHYSICO-CHEMICAL PROPERTIES AND QUALITY OF BRIE TYPE CHEESE

Dipendra Khanal Advisor: University of Guelph, 2014 Dr. Mansel W. Griffiths

Pulsed Electric Fields (PEF) and Microfiltration (MF) are emerging non-thermal technologies that are gaining in popularity, since they preserve food by inactivation of microorganisms and fulfill the demand for minimally processed and fresh-like foods with natural taste and extended shelf life. The present research investigated the effect of either stand-alone or a combination of PEF and MF treatments on microbial reduction, physicochemical properties in skim milk and its comparison with HTST treatments, and also explored the quality of Brie type cheese preparation and its comparison with raw milk cheese. Reduction in microbial load by non-thermal technologies were comparable or even better than HTST treatments, and effects on physico-chemical properties were varied depending on the use of different pore sized MFs.

Significant effects of non-thermal technologies were not observed on cheese quality and microbial analysis of ripened cheese did not detect the presence of pathogenic bacteria.

ACKNOWLEDGEMENTS

I wish to thank my supervisor Dr. Mansel W. Griffiths for his support, encouragement and enthusiasm during my graduate work. Dr. Griffiths has provided an opportunity to explore new world of science and technology by accepting me as a graduate student and welcomed by showing faith and provided the environment to update my knowledge in food science field, not only as a graduate student, but also as a researcher and problem shooter. I am grateful to Dr.

Griffiths for including me into his research group and I will always cherish those moments I spent while working in lab under his invaluable guidance and support throughout this project.

My sincere thanks also go to Dr. Lisa Duizer, Dr. Loong-Tak Lim, Dr. Milena Corredig and Dr.

Art Hill for their valuable advice, proper guidance and necessary support on my research.

My sincere gratitude to Dr. Markus Walkling-Ribeiro and Dr. Hany Anany who provided me with sound advice on scientific thinking and critical analysis as well as organising and assisting in conducting the experiments. My especial thanks goes to Ms. Anupam Chugh and all my friends at CRIFS.

I would like to thank Dairy Farmers of Ontario (DFO) for providing financial support for this project, Gay Lea Foods Co-operative for providing the milk throughout the project period, and Guelph Food Technology Center (GFTC) for providing the facility during cheese making. I would also like to thank Canadian Dairy Commission (CDC) for providing me the scholarship which also helped me to attend the international scientific conference to broaden my knowledge.

Last but not the least, heartfelt appreciation goes to my family, sisters, brother-in-laws who raised me, supported me, loved me, taught me the value of life and shed their sweats to make me, where I stands now, and finally I like to dedicate this thesis to my sisters, brother-in-laws and my wife Mrs. Pratima Rijal Khanal for their unconditional love to make my life better than the best.

iii

TABLE OF CONTENTS

Acknowledgements ...... iii

List of abbreviations ...... vii

List of tables...... ix

List of figures ...... x

Chapter 1: Introduction ...... 1

1.1 General introduction ...... 1 1.2 Research objectives ...... 5 Chapter 2: Literature review ...... 7

2.1 Milk processing ...... 7 2.2 PEF processing ...... 11 2.3 Factors affecting microbial inactivation ...... 17 2.3.1 Process parameters ...... 17 2.3.2 Product parameters ...... 22 2.3.3 Microbial parameters ...... 23 2.4 Mechanism of microbial inactivation in PEF processing...... 24 2.5 Milk processing through PEF ...... 27 2.6 Microfiltration processing of milk ...... 44 2.7 Hurdle technology ...... 54 2.8 Cheese making with PEF/MF treated milk ...... 57 Chapter 3: Impact of single and combined pulsed electric field and microfiltration treatments vs. heat pasteurization on microbial reduction and physico-chemical properties in skim milk ...... 62

3.1. Abstract ...... 62 3.2 Introduction ...... 63 3.3 Materials and methods ...... 67 3.3.1. Skim milk preparation and storage ...... 67 3.3.2. PEF treatment ...... 67

3.3.3. Microfiltration ...... 69 3.3.5. HTST treatment ...... 70 3.3.6. Microbiological analysis...... 70 3.3.7. Physico-chemical properties of milk ...... 71 3.3.8. Pathogen challenge study ...... 71 3.3.9 Statistical analysis...... 72 3.4 Results ...... 73 3.4.1. Suitability of PEF, MF, PEF/MF, and HTST for microbial reduction ...... 73 3.4.2. Effect of nonthermal and thermal processing on physico-chemical properties...... 74 3.4.3. Pathogen control by PEF and PEF/MF ...... 76 3.5. Discussion ...... 78 3.6 Conclusions ...... 86 Chapter 4: Color and flavor of Brie-type cheese made from high voltage pulsed and membrane microfiltered skim milk subjected to pathogen screening ...... 87

4.1 Abstract ...... 87 4.2 Introduction ...... 88 4.3 Materials and methods ...... 92 4.3.1 Skim milk preparation ...... 92 4.3.2 PEF treatment ...... 93 4.3.3 Microfiltration ...... 93 4.3.4 Brie cheese preparation ...... 93 4.3.5 Analysis of color attributes ...... 94 4.3.6 Analysis of volatile compounds ...... 95 4.3.7 Proximate analysis of cheese ...... 96 4.3.8 Listeria monocytogenes, Salmonella spp., and E. coli screening ...... 96 4.3.9 Statistical analysis...... 97 4.4 Results and discussion ...... 97 4.4.1 Physico-chemical quality of cheese ...... 97 4.5 Presence of pathogenic bacteria ...... 106 4.6 Conclusions ...... 108 Chapter 5. Conclusions and future outlook ...... 110

5.1 Thesis summary and general conclusions ...... 110 5.2 Future outlook ...... 113 References ...... 115

LIST OF ABBREVIATIONS

AC alternating current ANOVA analysis of variance °C degree Celsius CFU colony forming unit cm centimetre cP centipoises CRIFS Canadian Research Institute for Food Safety d day DC direct current E electric filed intensity e.g. exempli gratia (for example)

Eo critical electrical field strength ESL extended-shelf-life g grams h hours HV high voltage HTST high-temperature short-time Hz Hertz J Joule J flux kg kilo gram kJ kilo Joule kPa kilo Pascal kV kilo Volt LTLT low-temperature long-time m meter M molar MF microfiltration mg milligram

vii min minutes mL millilitre mS milliSiemens mTorr milliTorr NaCl sodium chloride P statistical probability PEF pulsed electric fields ppb parts-per-billion

RC filtration cake resistance, s seconds SD standard deviation SMUF simulated milk ultra-filtrate SNF solid-non-fat spp. species TFMF tangential flow microfiltration

Ti inlet temperature TMB Tris–maleate buffer

To outlet temperature

Tp processing temperature

Tt treatment time UHT ultra-high temperature. UV ultraviolet v Volt w/w weight by weight μm micrometer μs microsecond Ω ohm

viii

LIST OF TABLES

Table 2.1 Summary of microbial inactivation by PEF treatment...... 30

Table 2.2 Summary of physico-chemical properties of pulsed electric field (PEF) treated products...... 42

Table 2.3 Summary of microbial reduction by membrane microfiltration...... 50

Table 3.1 Physico-chemical properties of skim milk after high-temperature short-time pasteurisation (HTST-75 = 75 °C for 20 s; HTST-95 = 95 °C for 45 s), lab-scale tangential flow microfiltration (MF-0.65µm = 0.65µm, MF-1.2µm = 1.2 µm), pilot scale cross-flow microfiltration

Table 3.2 Pathogen bacteria cell recovery test results of skim milk after pulsed electric fields at low (PEF-L: 28 kV/cm; 2805 µs), and high (PEF-H: 40 kV/cm, 1571 µs) processing intensities and selected non-thermal combination treatment using PEF and subsequent lab-scale tangential flow microfiltration (MF-1.2µm = 1.2 µm) (PEF-H, PEF-L/MF-1.2µm, PEF-H/MF-1.2µm)……..77 Table 4.1 Proximate analysis of Brie cheese prepared from raw milk, milk treated with pulsed electric field (PEF; 40 kV/cm, 1571 µs) or a combination of PEF and cross-flow microfiltration

(CFMF1.4 = 1.4 µm) and ripened for 65 days (Data = mean ± S.D.; n =3) ……………………..98

Table 4. 2 Comparison of Hunter colour space attributes L* (Lightness), a* (redness) and b* (yellowness) measured in Brie cheese prepared from raw milk, pulsed electric field (PEF; 40 kV/cm, 1571 µs) and combination of PEF and pilot scale cross-flow microfiltration (CFMF1.4 = 1.4 µm) treated milk, after 65 days of ripening period (Data = mean ± S.D.; n = 3) ...... 101

Table 4.3 Retention of volatile flavour compounds in Brie cheese prepared from raw milk, pulsed electric field (PEF; 40 Kv/cm, 1571 µs) and a combination of PEF and cross-flow microfiltration (CFMF1.4 =1.4 µm) treated milk, after 65 days of ripening (Data = mean ± S.D.; n =3) ………………………………………………………………………………………...……105

LIST OF FIGURES

Figure 2.1 Simplified sketch of PEF with circuit for monopolar exponential decay pulse...... 12

Figure 2.2 Different types of pulse shape used in PEF treatment: (a) square pulse (b) exponential decay pulse, (c) bipolar square pulse, (d) bipolar exponential decay pulse...... 15

Figure 2.3 Configurations of treatment chamber for PEF processing...... 16

Figure 2.4 Electroporation of microbial cell when exposed under PEF treatment...... 26

Figure 2.5 Image of Saccharomyces cerevisiae suspended in apple juice as seen on transmission electron microscopy (TEM)...... 27

Figure 2.6 Different types of microfiltration ………………………………...... 45

Figure 3.1 Inactivation of native total aerobic microorganism in raw skim milk processed with HTST (HTST-75 = 75 °C for 20 s; HTST-95 = 95 °C for 45 s), lab-scale tangential flow microfiltration (MF-0.65µm = 0.65µm, MF-1.2µm = 1.2 µm), pilot scale cross-flow microfiltration

(MF-1.4µm = 1.4 µm), pulsed electric field at low (PEF-L: 28 kV/cm; 2805 µs), moderate (PEF- M: 40 kV/cm, 1122 µs), and high (PEF-H: 40 kV/cm, 1571 µs) processing intensities and selected nonthermal combination treatments using PEF and subsequent microfiltration. Initial microbial load in raw skim milk was 7.46 (±0.14) log10 CFU/mL. Same letters above the column stand for no statistical differences between two treatment means (n = 3)...... 73

Figure 3.2 Pathogen reduction obtained with pulsed electric fields at low (PEF-L: 28 kV/cm; 2805 µs), and high (PEF-H: 40 kV/cm, 1571 µs) processing intensities and selected non-thermal combination treatment using PEF and subsequent lab-scale tangential flow microfiltration (MF-

1.2µm = 1.2 µm) (PEF-H, PEF-L/MF-1.2µm, PEF-H/MF-1.2µm). Initial counts of Salmonella spp.

= 4.75 (±0.10), E. coli = 4.61 (±0.11) and Listeria monocytogenes = 4.60 (±0.15) log10 CFU/mL; columns without error bars (standard deviation) indicate that the detection limit (1.0 log10 CFU/mL) was reached (n = 3)………………………………………………………………….. 76

CHAPTER 1: INTRODUCTION

1.1 General introduction

Milk is the only food specifically “designed” for consumption by mammalian neonates, including humans. Bovine milk contains easily digestible proteins, fats, along with lactose, minerals, vitamins, and other components, which makes it nature’s single most complete food

(Jensen 1995). The high water activity (0.995), and water content (~88%), almost neutral pH and an ample supply of essential nutrients in milk make it a suitable growth medium for numerous spoilage and pathogenic microorganisms (Claeys and others 2013, Adams and Moss 2007).

These properties make raw milk a highly perishable food and easily subject to microbial contamination. Milk produced from the mammary glands of healthy animals is “initially sterile”

(Fernandes 2009), but subsequently may be contaminated with microorganisms from within the udder, exterior of the teats and udder, the milking environment, and handling and storage equipment (Adams and Moss 2007).

Consumption of raw milk, or products made from raw milk, has been associated with outbreaks of foodborne illness associated with several pathogens, including Escherichia coli

O157:H7, Listeria monocytogenes, Salmonella spp., Campylobacter spp., Staphylococcus aureus

, Bacillus cereus and Clostridium botulinum (Chye and others 2004, Claeys and others 2013).

Food spoilage and food poisoning has been a challenge since time immemorial, and civilizations, knowingly or unknowingly, have developed food preservation techniques such as drying, salting, pickling, smoking and heating (Hitzel and others 2013, Ho and Mittal 2000, Barat and others

2003).

The heat treatment of milk for preservation may be “as old as the cow and the use of fire”

(Holsinger and others 1997), and is arguably the most effective method for the destruction of microorganisms to produce microbiologically safe milk; thus ensuring public health and food safety (Walkling-Ribeiro and others 2011a, Steele 2000, Dutreux and others 2000). Thermal pasteurization also extends the shelf life of refrigerated milk by up to three weeks (Sepulveda and others 2009). Many countries have banned the sale of raw or unpasteurized milk, and made mandatory the thermal pasteurization of milk before consumption. Thermal processing also has some limitations, consumers have objected to the “cooked flavour” of milk treated at high temperatures and it also brings about undesirable changes to its physicochemical, nutritional and organoleptic quality, as well as resulting in denaturation of protein and loss of vitamins

(Wirjantoro and Lewis 1996, Pereda and others 2009) and destruction of natural antimicrobial milk components and immunoglobulins (Li-Chan and others 1995).

Consumer interest in minimally processed and “fresh-like” foods with natural taste and extended shelf life has triggered food scientists and technologists to look for alternatives to thermal processing, to avoid the adverse effects of heat, whilst ensuring food safety (Pereda and others 2009, Fernandez-Molina and others 2005). Non-thermal technologies could be an alternative to thermal technology to minimize the adverse effect of heat on food quality

(Barbosa-Cánovas and others 1999), food processed by non-thermal technologies are fresh-like and color, flavour and nutrients are less affected (Bermudez-Aguirre and others 2009).

Pulsed electric fields (PEF) is an emerging non-thermal technology, which is gaining in popularity, since it preserves food by inactivation of microorganisms (Barbosa-Cánovas and others 1999) using low energy, at ambient temperature and reduces the time for processing

(Barbosa-Cánovas and others 1999). PEF is a technology in which food is placed in between two electrodes under the influence of a high electric field and very short, high voltage pulses are applied, which brings about the destruction of microbial cells. The mechanism of microbial inactivation is known as electroporation, in which formation of pores takes place in cell membranes causing mechanical destruction of the cell, and ultimately leakage of the cellular contents into the environment; causing cell death (Toepfl and others 2007, Pothakamury and others 1997).

PEF has shown promise for liquid food processing by reducing/inactivating total counts

(Rodríguez-González and others 2011;Walkling-Ribeiro and others 2011a), L. monocytogenes

(Reina and others 1998), Salmonella spp. (Sensoy and others 1997, Floury and others 2006a), E. coli O157:H7 (Dutreux and others 2000, Evrendilek and Zhang 2005), and S. aureus (Sobrino-

Lopez and others 2006). Qin and others (1995) did not observe any significant change in physical and chemical properties of PEF treated skim milk and the product was fresh-like with natural nutritional characteristics. PEF treated food also showed satisfactory shelf life. The antimicrobial effect of PEF can be increased by combination of electric field strength and moderate temperature (Heinz and others 2003, Floury and others 2006b), since the increased temperature reduces the “critical potential of membrane electrical breakdown”, resulting in increased sensitivity of microorganisms (Floury and others 2006b).

Membrane microfiltration (MF) is another non-thermal technology that has been used in the food industry for some time as a separation process. MF is a pressure-driven physical separation process, which uses membranes with a pore size in the range of 0.10 - 5 μm. This makes it suitable for separation of particles, somatic cells, bacteria and spores from liquid foods (Huisman

2000, Ripperger and Altmann 2002). Membrane microfiltration in combination with thermal treatment has been in use in several countries to produce extended-shelf-life (ESL) milk (Elwell and Barbano 2006). Kumar and others (2013) have reported that MF processing is more efficient than bactofugation (centrifugal separation used to remove bacteria and spores in milk) for removal of bacteria and spores. Several researchers have investigated the efficiency of MF to reduce microbial counts in skim milk and have achieved reductions in counts of up to 5.4 log10 cycles coupled with removal of all somatic cells (Daufin and others 2001, Gosch and others

2014, Saboya and Maubois 2000). Maubois (2002) has reported a spore reduction of more than

4.5 log10 cycles in skim milk by MF using a 1.4 µm pore size membrane, whereas Madec and others (1992) reported the reductions in count of Salmonella spp. and Listeria spp. in skim milk of 2.5 and 1.9 log10 cycles, respectively. Membrane microfiltration has also been used to produce cheese with comparable quality to traditionally produced cheese (Gcsan-Guiziou 2010).

Amornkul and Henning (2007) suggested the use of MF as a promising alternative to heat pasteurization of milk for cheese making and claimed that cheese made with MF treated milk was superior in terms of E. coli load and overall sensory acceptability.

Raw milk cheeses are best known for their unique flavour; not obtainable in cheese made with heat pasteurized milk. Heat pasteurization of milk changes the chemical and microbiological quality of milk, which ultimately affects cheese quality along with sensory characteristics (Amornkul and Henning 2007). However, the risk of illness associated with raw milk cheese is generally higher than for cheese made with heat-treated milk. Brie and

Camembert are among popular soft cheeses made with raw milk. Little and Knochel (1994) have described Brie and Camembert related outbreaks of salmonellosis, shigellosis, botulism, listeriosis and infection with both enteropathogenic and enterotoxigenic E. coli.

Some researchers have reported the promising results in terms of texture and flavour on cheese made with PEF treated milk in comparison with cheese made with raw milk and heat pasteurized milk (Sepulveda and others, 2000, Yu and others 2009). Non-thermal technology may be an answer to consumers and cheese manufacturers looking for alternative technology to produce safe cheese with desired sensory characteristics. PEF processing has not been successful in the destruction of bacterial spores (Raso and others 1998), whereas, MF treatment has proven its effectiveness in removing bacterial spores (Olesen and Jensen 1989). Thus, it may be possible to combine PEF and MF as a non-thermal hurdle approach (Leistner 2000) to achieve effective microbial inactivation with no effect on physical and chemical properties of food material with minimal energy expenditure.

1.2 Research objectives

There have been several studies conducted on the effect of PEF on milk. However very few have determined the effect of PEF performed at sub-lethal temperatures, nor its effect on the physico-chemical properties of milk. MF is used in the cheese industry, but there is a gap in our knowledge regarding the effect of membrane pore size on the quality of milk and the soft cheese made from it. This study was conducted to determine the effect of stand-alone PEF, MF and a hurdle approach combining PEF with MF on the microbial load and physico-chemical properties of milk and on the quality of a soft cheese, Brie, made from the milk. The results were compared to those obtained with milk that had undergone high-temperature short-time (HTST) pasteurization.

The main objectives were:

1) To determine the efficacy of stand-alone PEF, MF and a combination of PEF/MF for microbial inactivation in raw skim milk in comparison with HTST pasteurization;

2) To determine whether stand-alone PEF or the combination of PEF and MF result in the inactivation of E. coli, L. monocytogenes and Salmonella spp. in skim milk;

3) To study the effect of stand-alone PEF, MF and combination of PEF/MF on the physico- chemical properties of skim milk in comparison with HTST pasteurization;

4) To evaluate the quality of Brie cheese prepared from milk processed using PEF or a combination of PEF and MF and compare it with cheese made from raw milk.

CHAPTER 2: LITERATURE REVIEW

2.1 Milk processing

Milk has formed a part of the human diet for more than 9000 years, and the history of cow domestication as a source of milk and other products can be traced back 6000 years (McGuffey and Shirley 2011). Milk is an important food and is consumed throughout life. It contains many essential nutrients and this, combined with its high water content and almost neutral pH makes it an ideal growth medium for spoilage and pathogenic microorganisms (Nada and others 2012).

Consequently, milk has a relatively short shelf-life. To overcome the perishable nature of food many preservation methods have been adopted (Ho and Mittal 2000), and, in the case of milk, early civilizations used to preserve it by using heat, and this practice originated along with the domestication of dairy animals and the discovery of fire (Holsinger and others 1997).

Raw milk has been associated with several pathogenic bacteria such as Campylobacter jejuni, Shiga toxin-producing E. coli, L. monocytogenes, Salmonella spp. and Yersinia enterocolitica (Jayarao and others 2006) and the risk of infection increases if milk is consumed raw. In the USA, during 1993-2006, 56 fluid milk associated foodborne disease outbreaks were recorded and of the 56 cases, 82% were associated with non-pasteurized milk (Langer and others

2012). These data show that consumption of raw milk poses a public health risk.

To minimize the health risk due to milk consumption, in 1824 William Dewes found that the chances of milk spoilage was reduced when the milk was heated to boiling point, and he recommended that milk be boiled before feeding it to infants (Steele 2000). However, the credit for identifying the cause of spoilage as microorganisms and verifying their destruction by heat

7 goes to Louis Pasteur. Boiling milk before consumption is still practiced in many parts of the world.

Some of the diseases such as typhoid fever, scarlet fever, tuberculosis, foot and mouth disease, and anthrax were identified as milk borne diseases even before 1900, but the heat treatments required for the destruction of the causative pathogens to make milk safe for consumption were not fully understood and it took until the 1930’s when for pasteurization standards were widely adopted in the US (Westhoff 1978).

In the 1870s, the Pasteur process was applied to milk in Denmark and commercial pasteurization of milk was started in Denmark and Sweden in 1885. In the USA, Cincinnati was the first city to adopt pasteurization in 1897 and, in 1909; Chicago became the first city to pass an ordinance to make pasteurization compulsory for milk to be sold within the city (Steele 2000).

Pasteurization of milk has had a significant impact on the burden of disease, worldwide. For example, in the USA before 1938, about 25% of all foodborne and waterborne disease outbreaks were related to milk but currently that figure is only about 1% (Claeys and others 2013).

The definition of pasteurization according to the International Dairy Federation (IDF) is:

“Pasteurization is a process applied to a product with the objective of minimizing possible health hazards arising from pathogenic microorganisms associated with milk by heat treatment which is consistent with minimal chemical, physical and organoleptic changes in the product” (Varnam and Sutherland 1994).

The most commonly used method for pasteurization of milk is 63 °C for 30 min and 71.7 °C for 15 s for batch pasteurization (low-temperature long-time; LTLT) and high-temperature short–

8 time (HTST) pasteurization, respectively (Motarjemi and others 2014). Batch pasteurization systems were commonly used before the development of the continuous system used for HTST pasteurization, which was developed between 1920 and 1927 and now almost all large-scale dairies use continuous pasteurization (Griffiths 2010).

HTST pasteurization has been used to process milk for decades to extend its shelf life and improve safety (Knorr and others 1994) by destroying pathogenic bacteria and lowering the microbial load of spoilage organisms. Though thermal processing is a low cost and simple technology, it also has some limitations; it often causes cooked flavours and brings about undesirable changes in nutritional quality, denaturation of protein and loss of vitamins

(Wirjantoro and Lewis 1996) and destruction of natural antimicrobial components of milk and immunoglobulins, which may have beneficial health effects (Li-Chan and others 1995).

The food industry, only by using heat processing technology, is unable to meet the consumer demand of safe foods with extended shelf life and fresh-like properties with natural taste and high nutritional and sensory quality (Pereda and others 2009). Therefore, the food industry as well as food scientists are looking for an alternative to traditional thermal technology, to process food and minimize the adverse effect of thermal technology on food processing (Knorr and others 1994). Non-thermal technologies could be an alternative to thermal treatment to minimize the adverse effect of heat on food quality, since this technology runs at low temperature

(Barbosa-Cánovas and others 1999). Bermudez-Aguirre and others (2009) reported that foods processed by non-thermal technologies are fresh-like and its color, flavour and nutrients are minimally affected. Any new technology to be placed in the category of an alternative technology should have a positive impact on the quality of food products, inactivate

9 microorganism to a safe level, while satisfying consumer needs and should operate at low cost

(Lado and Yousef 2002).

Destruction of microorganisms can be achieved at low temperature by the application of non-thermal technologies, so, energy requirement is less in comparison with traditional thermal technologies, and are less expensive and more environmentally friendly than traditional thermal technologies (Piyasena and others 2003). Different types of non-thermal technologies are being researched around the world and there are some merits and limitations for each novel technology. Some non-thermal technologies showing promising results include pulsed electric fields (PEF; Sharma and others 2014; Walkling-Ribeiro and others 2009, Walkling-Ribeiro and others 2011a, Bermúdez-Aguirre and others 2011), high hydrostatic pressure (HHP; Buzrul and others 2008, Boluda-Aguilar and others 2013, Rodiles-López and others 2010), microfiltration

(MF; Hoffmann and others 2006, Walkling-Ribeiro and others 2011a, Rodríguez-González and others 2011), ultrasound (Zenker and others 2003, Khanal and others 2014), and irradiation, or ultraviolet light irradiation (UV; Matak and others 2007, Tran and Farid 2004).

PEF is an emerging non-thermal technology that is gaining in popularity, since it preserves food by inactivation of microorganisms using low energy inputs and reduces the time for processing (Barbosa-Canovas and others 1999). This novel non-thermal technology can be used as a food preservation technology, which inactivates spoilage and pathogenic microorganism, and enzymes, and ensures extended shelf life and food safety (Castro and others 1993).

Research on PEF technology has shown that this non-thermal technology can be used as complementary or to replace traditional thermal processing, as a means to avoid the negative effect of thermal technology such as color alterations, flavor damage, and vitamin and nutritional

10 losses to produce minimally processed and fresh-like food products. Therefore, PEF processed foods are considered to be superior in terms of quality attributes, in comparison with those processed using traditional thermal technologies (Giner and others 2000, Barbosa-Canovas and others 2000).

Several researchers have worked on PEF with different food products such as skim milk

(Walkling-Ribeiro and others 2011a, Bermúdez-Aguirre and others 2011), whole milk

(Bermúdez-Aguirre and others 2011; Sharma and others 2014), apple juice (Walkling-Ribeiro and others 2008, Bi and others 2013, Aguilar-Rosas and others 2007), orange juice (Rivas and others 2006, Agcam and others 2014, Walkling-Ribeiro and others 2009a) grape juice (Huang and Wang 2009, Robert and others 2013), smoothie-type beverage (Walkling-Ribeiro and others

2008; Walkling-Ribeiro and others 2010), liquid whole egg (Monfort and others 2012,

Kazmierska and others 2012; Marco-Moles and others 2011), beer (Walkling-Ribeiro and others

2011b, Evrendilek and others 2004a), and wine (Puértolas and others 2009, Wall and others

2003). PEF and MF technologies can be considered as a potential alternative to traditional thermal technology and are also suitable for liquid foods like milk to improve quality.

2.2 PEF processing

The history of use of electricity dates back to the work of a German engineer, Heinz

Doevenspeck, in the 1960s, who showed that microbial inactivation in food can be achieved by the application of pulsed electric fields. After this, Sale and Hamilton in the late 1960s worked on PEF and found that “loss in the semi permeability properties” of cell membranes was the cause of microbial death (Barbosa-Canovas and Sepulveda 2005). Sale and Hamilton (1967) pioneered the study of microbial inactivation by PEF and showed that electric field strength and

11 treatment time are the important factors for microbial inactivation, and the inactivation is a result of generated electrical pulses. They also reported that there is no role of increase in temperature or chemical products of electrolysis in microbial inactivation in PEF processing. In PEF processing, food is placed between two electrodes and very short pulses (of 1-10 μs duration) of a several thousand volts electric field are applied across (Barbosa-Canovas and others 2000, Ho and Mittal 2000). The electric field developed due to this applied high voltage results in microbial inactivation.

In general, a PEF system (Figure 2.1) consists of a high voltage generator, capacitor to store energy, high voltage control unit (HV switch) and treatment chamber. The system also consists of a cooling and/or heating system, temperature measurement and recording devices, voltage recording instrument and a pump (for continuous system), through which the product is transported to the treatment chamber (Huang and Wang 2009).

HV switch Temperature control

Resistance (R) Treatment

chamber Raw milk Capacitor

DC power Treated supply milk

Water circulation

Figure 2.1. Simplified sketch of PEF system with circuit for monopolar exponential decay pulse. Modified from (Zhang and others 1995).

The main purpose of the pulse generator is to receive low voltage as an alternating current

(AC) from a power supply and to convert it to high voltage (HV), which is then supplied to PEF chambers (Zhang and others 1995). Capacitors are charged by direct current (DC) power obtained from an amplified and rectified AC source. The low level voltage is collected and stored in a capacitor and this releases the energy to the food material through a HV discharge switch (Min and others 2007). The devices normally used as switches include Ignitron, Gas

Spark Gap, Trigatron, or Thyratron (Ho and Mittal 2000).

More than one capacitor can be combined to store more energy. According to Zhang and others (1995), the energy stored in the capacitor can be expressed as:

where, C = capacitance (F),

V = voltage (V)

The electric field strength (E; kV/cm) developed in the treatment chamber by supplying the electric energy stored in capacitor can be expressed according to Barbosa-Canovas and

Sepulveda (2005) as:

For parallel plate electrodes:

For coaxial electrodes:

where, V = electric potential difference between two points (V),

D = distance between two electrodes (cm)

r = radius of electric field measurement (cm)

R1 and R2 = radii of the inner and outer electrodes (cm)

Two types of pulse shape (i.e., exponential decay and square wave) are normally used in

PEF processing. Bipolar forms are also in use depending on circuit design. A square wave pulse is generally regarded as superior over exponential decay pulse for microbial inactivation (Qin and others 1994). In exponential decay pulse, the voltage is unidirectional and increases rapidly to a maximum level then decreases to ground level slowly, whereas in the case of a square pulse, the voltage increases and decreases quickly but remains constant for some time. A bipolar pulse consists of one positive and another negative pulse, while monopolar has either all positive or negative pulses. Exponential decay pulses are easy to generate, whereas a square wave pulse needs a pulse-forming network consisting of inducers and capacitors and is more complex and challenging to generate than an exponential decay pulse (Evrendilek and Zhang 2005). The types of pulses generated during PEF are illustrated in Figure 2.2.

Bipolar pulses are more efficient for microbial inactivation than mono polar pulses (square or exponential decay), due to alternate movement of charged groups of the cell membrane causing extra stress, which results in greater susceptibility to electric field strength (Ho and others 1995, Barbosa-Canovas and Sepulveda 2005).

a b

Voltage (V) Voltage Voltage (V) Voltage

Time (µs) Time (µs)

c d

Voltage (V) Voltage (V) Voltage

Time (µs) Time (µs)

Figure 2.2. Different types of pulse shape used in PEF treatment: (a) square pulse (b) exponential decay pulse, (c) bipolar square pulse, (d) bipolar exponential decay pulse.

Two types of chamber are used in PEF: static and continuous. In the static system, a unit block of food material is placed in the chamber, where it is treated with applied electric field strength and is replaced with another food block after a definite time, while in the continuous system the food is pumped through the chamber and is cooled before collection (Huang and

Wang 2009). The continuous system is an updated and modified form of the static system and is suitable for pumpable liquid foods. Due to greater treatment uniformity, the continuous system is more effective for microbial inactivation than the static system (Martin and others 1997).

In the treatment chamber, at least two electrodes are encased inside the insulating material, one is high voltage and the other is ground. Parallel and coaxial chamber configurations are

15 commonly used for processing foods by PEF (Motarjemi and others 2014) and are described in

Figure 2.3. The applied high voltage results in an electric field that causes microbial inactivation.

Parallel plates are normally used with a batch system, whereas coaxial systems are used for continuous operation (Toepfl and others 2005). Parallel plate configuration is easy to design and produces uniform electric field distribution; however, electric intensity is greatly reduced in the wall of the electrodes. The coaxial plate is best known for smooth and uniform product flow and is suitable for large industrial applications (Jeyamkondan and others 1999).

Electrodes used for food treatment should be of food grade material and sterilizable, and made up of electro-chemically inert materials. The material for constructions of electrodes is usually stainless steel but gold, platinum, carbon and metal oxides can also be used as an alternative (Toepfl and others 2005, Bushnell and others 1996).

Product flow Product flow (parallel plate) (coaxial plate) Figure 2.3. Configurations of treatment chamber for PEF processing. Adapted from (Toepfl and others 2005).

2.3 Factors affecting microbial inactivation

The success of PEF technology to pasteurize food largely depends on its ability to inactivate different types of microorganisms, including spoilage and pathogenic bacteria. To determine the efficiency of its operation, we need to know the critical factors that affect the rate of microbial inactivation. Wouters and others (2001) have summarized the different factors affecting PEF efficiency as: process parameters, product parameters and microbial parameters.

2.3.1 Process parameters

There are numerous treatment factors that can affect microbial inactivation, including electric-field strength, PEF treatment time, temperature (inlet and outlet), number of pulses, pulse width and pulse shape, flow rate, frequency, chamber configurations etc. (Wouters, and others 2001, Evrendilek and Zhang 2005). Microbial inactivation increases when electric field strength is increased above a certain value, the critical electric field strength (Qin and others

1998).This critical electric field strength depends on microbial types and morphology, and most of the microorganisms are not inactivated by electric field strengths of less than 4-8 kV/cm

(Peleg 1995).

Martin and others (1997) observed a higher degree of inactivation of E. coli in skim milk when the electric field strength was increased but the same pulse width was maintained. A higher degree of inactivation was observed with longer pulse width than shorter when the milk was treated at the same electric field strength. The increased microbial inactivation with increase in electrical field intensity is related to the electroporation theory of microbial inactivation, which

17 states that the induced potential difference across the cell membrane is proportional to the applied electric field intensity (Pothakamury and others 1997).

2.3.1.1 Microbial inactivation kinetics based on process parameters

Mathematical expression of microbial inactivation is essential to predict the desired level of microbial destruction by PEF processing. Lack of knowledge of inactivation kinetics fails to establish appropriate PEF processing conditions, which ultimately results either in under or over processing, leading to food safety problems and economic loss (Wouters and others 2001). Some of the kinetic models used in PEF processing are described below.

After observing the linear destruction of microorganisms beyond the limit of applied electric field strength, Hulsheger and others (1981) modeled microbial inactivation based on first order kinetics with respect to electric field strength (2.4), and treatment time (2.5), and this can be expressed as:

where, N = survivor fraction,

No = initial population,

E = applied electric field (kV/cm),

Ec = critical electric field strength (no inactivation below this value),

t= treatment time (pulse number × time constant; µs),

tc = critical treatment time to reach Ec (no inactivation below this value),

bE = coefficient of regression, depends on E and n,

n = number of pulses

After combining (2.4) and (2.5), Hulsheger and others (1981) derived the following model for microbial inactivation:

where, k = specific rate constant, value of Ec, tc, and k depends on microbial species.

Critical electric field strength (Ec) is related to cell size, the bigger the cell, the smaller the threshold strength required for inactivation (Saccharomyces cerevisiae: Ec = 4.7 kV/cm; E. coli:

Ec =13.7 kV/cm) (Grahl and Markl 1996). The electric field strength required to inactivate gram- negative bacteria (7.2 kV/cm) is generally lower for than for gram-positive bacteria (11.5 kV/cm). Critical treatment time (tc) is also lower for gram-negative bacteria (27 µs) than gram- positive bacteria (61 µs) (Hulsheger and others 1983).

Zhang and others (1994) reported linear inactivation of E. coli, and S. cerevisiae suspended in simulated milk ultra-filtrate (SMUF) and apple juice, by 3-4 log10 with respect to electric field strength and number of pulses. Inactivation of E. coli in skim milk (Martin and others 1997) and

E. coli, Lactobacillus brevis, and Pseudomonas fuorescens in sodium-alginate solution and UHT milk (Grahl and Markl 1996) also followed Hulsheger’s inactivation model.

With the advancement of technology and knowledge it became apparent that a linear relationship of microbial inactivation with respect to applied electric field and treatment time

19 does not always apply and the relation also differs with types of microorganism. Various critical factors, large range of operational conditions and different types of equipment used in PEF processing makes the task even more complex to compare the result of different types of microbial inactivation (Wouters and others 2001).

On the basis of the Fermi equation, Peleg (1995) proposed a double logarithmic model to describe the sigmoidal shaped curve of microbial inactivation with respect to applied electric field:

where, S = percent surviving microorganisms,

E = applied electric field (kV/cm),

Ec = critical electric field strength (value of E, when S = 50%),

k = steepness of the survival curve around Ec (kV/cm),

n = number of pulses

Some of the other kinetic models generally used in PEF inactivation includes log-logistic kinetic model and a Weibull distribution based model (Min and others 2007).

Treatment time is another factor which affects microbial inactivation, and is the effective time during which product is subjected to the field strength. It is calculated as a product of number of pulses and the width of the pulses applied. Longer pulse width results in higher inactivation and increased pulse width gives longer treatment time (Min and others 2007).

Relation of pulse width and microbial inactivation is governed by threshold field strength (Ve), in

20 which the longer the pulse width, the greater the decrease in threshold field strength (Ec) (Martin and others 1997).

Square-wave pulses are more efficient in microbial inactivation than exponentially decaying pulses (Pothakamury and others 1996) and bipolar pulses are more lethal than monopolar (Ho and others 1995). Zhang and others (1994) along with Qin and others (1994) have reported 60% more inactivation of S. cerevisiae by square wave pulses than exponential decay pulse, under the same conditions of electric field strength and treatment time.

Temperature control is also important during PEF treatment to achieve effective microbial reduction. Proper cooling of the product either after or in between processing is important to correctly evaluate the inactivation due to PEF treatment. Vega Mercado and others (1996b) and

Shamsi and others (2008) observed higher microbial inactivation when inlet and/or outlet temperature of product was increased in PEF treatment.

Zhang and others (1995) have reported an increase in inactivation of E. coli suspended in

SMUF when the treatment temperature was increased from 7 to 20 °C. Dunn and Pearlman

(1987) claimed that a combination of PEF and heat energy is more efficient for microbial reduction than either technology alone. Microbial inactivation in milk was increased using a combination of 55 °C and PEF. Increase in temperature reduced the trans-membrane breakdown potential of cell membrane, which caused an increase in microbial inactivation at high temperature (Zimmermann 1986).

2.3.2 Product parameters

PEF treatment has been applied to different types of food products and physical and chemical properties of different products have been shown to influence microbial inactivation.

Huang and Wang (2009) have summarized the product parameters which affect the microbial inactivation as electrical conductivity, water activity, density, viscosity and pH.

Electrical conductivity of any medium is its ability to allow flow of electrons. According to

Zhang and others (1995), electrical conductivity (S/m) can be expressed as:

Where, d = gap between the two electrodes (m),

R = resistance of the medium (Ω),

A = electrode surface area (m2)

The electrical conductivity of a food depends on its composition and presence of ions, salts and acids. According to Singh and Heldman (2014), electrical conductivity increases with increase in temperature, and their relation can be expressed as:

where, σo = electrical conductivity at 0 °C,

α = constant (1/ °C),

T = temperature (°C)

In a review by Wouters and others (2001), a higher inactivation of Lactobacillus plantarum was achieved in low conductivity medium (4 mS/cm buffer solution) than in higher conductivity medium (16 mS/cm buffer medium) when all other parameters were kept constant.

Wouters and others (1999) examined the effect of PEF on Listeria innocua suspended in phosphate buffer at pH 4.0 and 6.0. More than a 6 log10 reduction in count was found at the lower pH, while only 0.6 log10 reduction in count was observed for cells suspended in higher pH buffer. Vega Mercado and others (1996a) also found higher inactivation of E. coli in SMUF at pH 5.7 than at pH 6.8. However, Wouters and others (2001) reported in a review that some studies did not find any effect of pH on microbial inactivation and concluded that the effect of pH depends on microbial type. Low water activity food makes microorganisms more resistant to

PEF, resulting in lower microbial inactivation (Pagan and others 2005). Fat content also affects microbial inactivation. Higher fat protects the microorganisms against the electric field resulting in lower rates of inactivation (Bermúdez-Aguirre and others 2011, Grahl and Markl 1996).

2.3.3 Microbial parameters

Microbial cell size, species, strain, Gram type, and growth stage have been found to affect the inactivation of microorganisms in food by PEF treatment (Huang and Wang 2009). Larger microorganisms are more sensitive to electric field strength than those of smaller size. Small cells develop lower trans-membrane potential in an electric field than larger cells and develop greater resistance, so, larger cells are more easily permeabilized than small cells (Pagan and others 2005). Shamsi and others (2008) also reported that yeasts are more sensitive to applied electric field than bacteria due to their larger size.

According to Nuumann (1992), trans-membrane potential (TMP) and cell size are related and the relation can be presented by the following equation:

(2.10)

where, k = form factor (3/2 for spherical cells)

E = electric field strength

r = cell radius

φ = angle between the electric field and the point of interest on the membrane (1 at the

poles)

Devlieghere and others (2004) have reported in a review paper that gram-negative bacteria are more easily inactivated by PEF treatment than gram-positive bacteria, and spores are most resistant. While Marquez and others (1997) observed more than 3.43 and 5.0 log10 reduction of

Bacillus subtilis and B. cereus spores in salt solution at 50 kV/cm and 30-50 pulses. Raso and others (1998) also reported 3.4 - 4.2 log10 destruction of heat resistant ascospores of

Zygosaccharomyces bailii suspended in different fruit juices when treated by PEF at around 30 kV/cm and 2 μs pulse width at 20 °C. Pagan and others (1998) did not observe destruction of

Bacillus cereus spores with PEF treatment of 60 kV/cm and 75 pulses. Pothakamury and others

(1996) treated E. coli suspended in SMUF by PEF at 36 kV/cm and 2-4 pulses and observed greater destruction of cells in the logarithmic phase than for cells in the stationary and lag phases.

2.4 Mechanism of microbial inactivation in PEF processing

Formation of pores (termed electroporation) in the cell membrane and leakage of cell content into the environment is the widely accepted theory to explain the mechanism of cell

24 death. This pore formation is due to structural changes in the microbial cells and membranes of microorganisms, developed by applied electric field (Pothakamury and others 1997). Sale and

Hamilton (1967) reported that mechanical damage to the cell membrane is responsible for cell death.

Several theories have been proposed to explain electroporation and microbial inactivation.

The theory proposed by Zimmermann and others (1974) has been accepted for the mechanism of electropermeabilisation and microbial inactivation, and is illustrated in Figure 2.4. This theory assumes the microbial cell is a capacitor filled with dielectric material of low dielectric constant, surrounded by an environment of high electrical conductivity. The applied electric field increases trans-membrane potential, due to accumulation of free charges with opposite polarity on both sides of the membrane surface. These charges are the result of differences in dielectric constant, which attract each other resulting in cell compression. When the applied electric force equals the critical electric field (electric field at which membrane breakdown occurs), reversible pore

(electroporation) formation in the membrane occurs, and a further increase in applied electric field results in electropermeabilisation, i.e., leakage of cellular contents into the extracellular environment, which brings about cell death. The formation of pores in the cell membrane leads to reversible (electrical) or irreversible (mechanical) breakdown leading to cell death and depends on the magnitude of the trans-membrane potential (Pagan and others 2005; Wouters and others 2001).

Figure 2.4. Electroporation of microbial cell when exposed under PEF treatment. E = electric field strength, Ec = critical electric field strength. Redrawn form (Pagan and others 2005).

The formation of large pores in the membrane results in an irreversible breakdown, and pore size can be enlarged by increasing the applied electric field strength and treatment intensity.

Nevertheless, greater trans-membrane potential is induced in large cells compared to small cells and in both of these cases, makes the cells more susceptible to PEF treatment, resulting in mechanical destruction leading to cell death (Toepfl and others 2005). Electroporation is a process of temporary destabilization of protein and the lipid bilayer in a microbial cell under PEF treatment (Castro and others 1993).

Harrison and others (1997) studied inactivation mechanisms of S. cerevisiae suspended in apple juice and found disruption of intracellular organelles and other structural changes to be the cause of inactivation (Figure 2.5). Microscopic examination of Staphylococcus aureus cells

26 suspended in SMUF by Pothakamury and others (1997) showed the appearance of rough surfaces of treated cells, leading to cell lysis.

Untreated cell PEF treated cell (40 kV/cm)

Figure 2.5. Image of Saccharomyces cerevisiae suspended in apple juice as seen on transmission electron microscopy (TEM). Adapted from (Harrison and others 1997).

2.5 Milk processing through PEF

Several studies have been conducted to evaluate the efficacy of PEF to inactivate microorganism in model systems as well as food. A summary of these findings is provided in

Table 2.1. A significant number of studies of PEF treatment have been focused on evaluation of the effect of PEF on milk and milk products, and most of the research has been done to study the effect of PEF on microbial inactivation (Bendicho and others 2002). The definitive research on microbial inactivation induced by PEF was performed by Sale and Hamilton (1967), who showed the destruction of vegetative bacteria and yeasts cells suspended in NaCl solution and

27 confirmed that the destruction is due to electric field strength and pulse applied rather than products of electrolysis and an increase in temperature. Several studies have reported that a minimum of 4 kV/cm need to be applied to inactivate larger cells like S. cerevisiae, whereas, for smaller cells like L. innocua, a minimum of 15 kV/cm is required in liquid medium

(theoretically more than 35 kV/cm), and electrical field intensity required to inactivate almost all other bacteria, which are associated with food preservation/food safety falls between that necessary to kill S. cerevisiae and L. innocua (Barbosa-Cánovas 1998, Sale and Hamilton 1967,

Toepfl and others 2007).

Early studies of microbial inactivation by PEF were performed on model food systems, such as distilled water, phosphate buffer, and SMUF. Hulsheger and Niemann (1980) observed about

4.0 log10 reduction in count of E. coli suspended in different types of electrolyte, under PEF at 20 kV/cm and 10 pulses. Hulsheger and others (1983) also found a 3.0 log10 reduction in count of L. monocytogenes suspended in phosphate buffer (pH 7.0) by PEF at 20.0 kV/cm and 30 pulses.

Walkling-Ribeiro and others (2011a) studied the reduction in mesophilic count in raw skim milk and showed that populations were reduced by 2.0, 2.1, 2.3 and 2.5 log10 CFU/mL when the skim milk was treated at 16, 20, 30 and 42 kV/cm with maximum processing temperature of 40, 45, 47 and 49 °C, respectively.

Bermúdez-Aguirre and others (2011) found 2.4 and 3.6 log10 CFU/mL reductions in count of mesophiles in skim milk (initial count slightly higher than 3.0 log10 CFU/mL) at 46.15 kV/cm and 30 pulses with processing temperature of 50 and 60 °C, respectively. Odriozola-Serrano and others (2006) investigated the effect of treatment time on PEF inactivation of total aerobic bacteria in whole milk. Less than 1.0 log10 reduction in count was observed at 35.5 kV/cm for

300 μs. When the total time was increased to 1000 μs under the same PEF conditions, more than a 1.0 log10 reduction in count was observed. In a study with whole milk, Sepulveda and others

(2009) achieved >1.5 log10 reduction in count of mesophilic bacteria by a PEF treatment at 35 kV/cm.

Walkling-Ribeiro and others (2009b) investigated the reduction of native microbiota in inoculated low fat UHT milk with an initial microbial count up to 10.0 log CFU/mL. A microbial reduction in count of 5.0, 5.0 and 6.4 log10 CFU/mL was observed when the milk was treated with PEF at 40, 40 and 50 kV/cm for 50, 60 and 33 μs, respectively. When the milk was PEF treated under the same conditions, after pre-heating to 30, 40 and 50 °C, a maximum reduction of

7.0 log10 CFU/mL was observed at 50 °C.

Sharma and others (2014) inoculated whole milk with pathogenic bacteria to achieve initial counts up to 8.0 log10 CFU/mL and determined the level of inactivation by PEF treatment under different conditions. Inactivation of 5 - 6 log10 cycles of E. coli and L. innocua was observed when the inlet temperature of milk was 55 °C and PEF treatment at 23-28 kV/cm was applied.

Table 2.1. Summary of microbial inactivation by PEF treatment.

Microorganism Treatment media PEF processing conditions log10 References reduction Mesophilic Skim milk E = 16 kV/cm, Tp= 2105 µs, Tt =40 °C 2 Walkling-Ribeiro and microorganism others (2011a) E =20 kV/cm, Tt =1454 µs, Tp = 45 °C 2.1

E=30 kV/cm, Tt = 983 µs, Tp = 47 °C 2.3

E= 42 kV/cm, Tt = 612 µs, Tp = 49 °C 2.5

Mesophilic Skim milk E=46.15 kV/cm, n = 30, Tp = 50 °C 2.4 Bermúdez-Aguirre and microorganism others (2011) E=46.15 kV/cm, n= 30, Tp = 60 °C 3.6

Mesophilic Whole milk E=35.5 kV/cm, Tt =300 μs <1 Odriozola-Serrano and microorganism others (2006) E=35.5 kV/cm, Tt = 1000 μs >1

Mesophilic bacteria Whole milk E=35 kV/cm, Tp = 65 °C >1.5 Sepulveda and others (2009) Native microorganism Low fat UHT milk E=40 kV/cm, Tt = 50 μs, Ti = 20 °C, Tt =55 5 Walkling-Ribeiro and °C others (2009b) E=40 kV/cm, Tt =60 μs, Ti= 20 °C, Tp=55 °C 5

E=50 kV/cm, Tt = 33 μs, Ti = 20 °C, Tt =55 6.4 °C Natural microflora Milk protein concentrate- E=35-40 kV/cm 5 Sampedro and others fortified soymilk (2005) Escherichia coli Whole milk E=23-28 kV/cm, Ti = 55 °C 5-6 Sharma and others (2014)

E. coli Skim milk E=45 kV/cm, n= 64, Tp= 15 °C 3 Martin and others (1997)

(continued on next page)

Table 2.1 (continued)

E. coli Whole milk E=30 kV/cm, Tt = 200 μs 3 Zhao and others (2013) E=30 kV/cm, Tt = 600 μs 5

E. coli Fat free milk E=41kV/cm, n = 63, Tp = 37 °C 5.7 Dutreux and others (2000) Phosphate buffer, pH 6.8 5.4

E. coli UHT milk E=23 kV/cm, n=20 >4 Grahl and Markl (1996)

E. coli SMUF E=70 kV/cm, n= 100, Tp = 20 °C 9 Zhang and others (1995)

E. coli SMUF (100% w/w) E=37.2 kV/cm, Ti = 10 °C 5.1 Alkhafaji and Farid (2007) E=43.4 kV/cm, Ti = 10 °C 6.3

E. coli SMUF (66.66% w/w) E=43.4 kV/cm, Tt= for 880 μs, Ti = 10 °C 5.29

E=43.4 kV/cm, Tt = for 880 μs,Ti = 17 °C 6.6

E. coli Milk protein concentrate- E=35-40 kV/cm 5 Sampedro and others fortified soymilk (2005) E. coli UHT full fat milk E=60 kV/cm, Tt = 210 μs, Tp = 50 °C 8 Shin and others (2007)

E. coli O157:H7 citrate–phosphate E= 30 kV/cm, Tt = 99 μs, Tp = 4 °C 0.93 Saldana and others (2012) McIlvaine buffer, pH 3.5 citrate–phosphate E=30 kV/cm, Tt = 99 μs, Tp =50 °C 3.76 McIlvaine buffer, pH 7 E. coli Soymilk E=20 kV/cm, Tt = 547 μs, Tp=25 °C 2.7 Li and others (2013)

E=40 kV/cm, Tt = 547 μs, Tp = 25 °C 5.2

(continued on next page)

Table 2.1 (continued)

Listeria monocytogenes Phosphate buffer, pH 7 E = 20.0 kV/cm, n =30 4 Hulsheger and others (1983) L. monocytogenes Whole milk E=30 kV/cm, Tt = 200 μs >3.5 Zhao and others (2013)

E=30 kV/cm, Tt = 600 μs >5

L. monocytogenes Whole milk (fat=3.5 %) E=30 kV/cm, Tt =600 μs, Tp =20 °C 3 Reina and others (1998) Skim milk (fat=2%) Skim milk (fat=0.2%) E=30 kV/cm, Tt =600 μs, Tp =50 °C 4

L. monocytogenes Distilled water E=20 kV/cm, n=10,000 4 Simpson and others (1999) L. monocytogenes McIlvaine buffer, pH 3.8 E =22 kV/cm, Tt = 800 μs, Tp =35 °C 5.1 Sampedro and others (2005) McIlvaine buffer, pH 5.4 3.9

McIlvaine buffer, pH 7 1.5

L. monocytogenes Skim milk gel E=30 kV/cm, Tt =32.5 µs, Tp = 35 °C 0.75 Fleischman and others (2004) E=20 kV/cm, Tt =165.5 µs, Tp = 35 °C 0.75

E=20 kV/cm, Tt =32.5 µs, Tp =55 °C 4.5

Listeria innocua Whole milk E=23-28 kV/cm, Ti = 55 °C 5-6 Sharma and others (2014)

L. innocua Fat free milk E=41kV/cm, n = 63, Tp = 37 °C 3.9 Dutreux and others (2000) Phosphate buffer , pH 6.8 3.5

L. innocua Milk E=20-80 kV/cm (not specified), Tp = 50 °C >6 Dunn (1996)

(continued on next page)

Table 2.1 (continued)

L. innocua Low-fat UHT milk E=40 kV/cm, Tt = 25 μs 2.5 Noci and others (2009)

E=40 kV/cm, Tt = 50 μs 3.3

L. innocua Skim milk E=50 kV/cm, n=32 2.4 Calderón-Miranda and others (1999) L. innocua Whole milk E=30 kV/cm, (Ti = 43 °C, n = 10), (Ti =33 4.3 Guerrero-Beltrán and °C, n =17.5), (Ti = 23 °C, n = 20), others (2010) (Ti = 13 °C, n = 25) E=40 kV/cm, n=15, Ti = 15 °C 5.5

Salmonella Dublin Skim milk E=35.0 kV/cm, Tt = 163.9 μs 4 Sensoy and others (1997)

S. Dublin Whole milk E=36.7 kV/cm 4 Dunn and others (1987)

Salmonella Enteritidis Skim milk E=47 kV/cm, Tt =500 ns 1.2 Floury and others (2006a)

E=47 kV/cm, Tt =500 ns + 62 °C for 38 s 2.3

S. Enteritidis Citrate-phosphate E=28 kV/cm, Tt = 1400 μs, Tp = 35 °C >6 Alvarez and others (2003) McIlvaine buffer, pH 7 Salmonella Typhimurium

S. Typhimurium Citrate–phosphate E= 30 kV/cm, Tt = 99 μs, Tp = 4 °C 1.86 Saldana and others (2012) McIlvaine buffer, pH 3.5 citrate–phosphate E=30 kV/cm, Tt = 99 μs, Tp =50 °C 5.18 McIlvaine buffer, pH 7 S. Typhimurium Distilled water E=20 kV/cm, n=10,000 6 Simpson and others (1999) Bacillus UHT full fat milk E=60 kV/cm, Tt = 210 μs, Tp = 50 °C 8 Shin and others (2007) stearothermophilus E = electric filed intensity, Ti = inlet temperature, To= outlet temperature, Tp = processing temperature, Tt = treatment time, n =pulse number SMUF= simulated milk ultra-filtrate, UHT = ultra-high temperature.

Zhao and others (2013) inoculated E. coli or L. monocytogenes in sterilized whole milk to an initial count of about 7.0 log10 CFU/mL and treated the milks with PEF under different conditions to investigate the effect of electric field strength and treatment time on microbial inactivation. The milk treated with 30 kV/cm for 200 μs resulted in 3.0 and >3.5 log10 reduction in count for E. coli and L. monocytogenes, respectively. When the milk was treated with 30 kV/cm for 600 μs, the inactivation reached 5.0 and >5.0 log10 CFU/mL for E. coli and L. monocytogenes, respectively.

In a review of microbial inactivation by PEF, Sampedro and others (2005) reported a 4.0 log10 reduction in count of S. Dublin in pasteurized, homogenized, inoculated milk. Sensoy and others (1997) inoculated skim milk with S. Dublin to achieve an initial count of 5.0 log10

CFU/mL and treated it with PEF to study the effect of different operational conditions on microbial inactivation. A maximum inactivation of 4.0 log10 CFU/mL was achieved with a PEF treatment at 35.0 kV/cm with 163.9 μs treatment time of 1 μs pulse duration, and concluded that increases in electric field strength, treatment time and temperature resulted in greater inactivation.

Floury and others (2006a) combined PEF and heat treatment to study the microbial inactivation of S. Enteritidis in skim milk. PEF treatment alone at 47 kV/cm, 500 μs and 60 Hz resulted in a 1.2 log10 reduction in count, whereas when the PEF treated milk was held at 62 °C for 38 s before cooling and collection of sample, the degree of inactivation was increased to 2.3 log10.

Dutreux and others (2000) investigated the inactivation of E. coli and L. innocua suspended in fat free milk and phosphate buffer having similar pH and electrical conductivity. PEF

34 treatment at 41 kV/cm and 63 pulses with 2.5 μs pulse width induced the reduction of 5.7 and 5.4 log10 CFU/mL for E. coli suspended in fat free milk and phosphate buffer, respectively; whereas a reduction in count of 3.9 and 3.5 log10 CFU/mL was observed for L. innocua suspended in fat free milk and phosphate buffer, respectively. They concluded that composition of the medium has very little or no effect on the reduction of microorganisms.

Martin and others (1997) conducted a study on the effect of different parameters of PEF on microbial inactivation in skim milk inoculated with E. coli at 9.0 log10 CFU/mL. They observed a nearly 3.0 log10 CFU/mL reduction in count when PEF was applied at 45 kV/cm, 64 pulses and

15 °C in a batch system. Grahl and Markl (1996) examined the lethal effect of PEF and specific critical electrical field strength on E. coli suspended in UHT milk. Inactivation of E. coli was found to be more than 4.0 log10 cycles at 23 kV/cm and the specific critical electrical field strength (Eo) was 13.7 kV/cm.

Zhang and others (1995) used a static PEF chamber to investigate the inactivation of E. coli suspended in SMUF with high strength, short duration pulsed electric fields. They successfully achieved a 9.0 log10 reduction in count of E. coli using a stepwise PEF treatment at electric field strength of 70 kV/cm, and 100 pulses with 2 μs pulse width at 20 °C.

Reina and others (1998) studied the efficiency of different parameters of PEF against L. monocytogenes suspended in milk with different composition. They used pasteurized milk with different fat content including 3.5 % fat milk, 2% fat milk and skim milk with 0.2% fat. About a

3.0 log10 reduction in count was observed for all samples when milk was treated at 30 kV/cm, pulse width 1.5 μs, for 600 μs treatment time at 20 °C. When they increased the treatment temperature to 50 °C, the reduction in count was increased to 4.0 log10 cycles. They concluded

35 that the fat content in milk does not affect the inactivation of E. coli but an increase in treatment temperature positively correlates with levels of inactivation achieved. The effect of PEF treatment time and electric field on lethality of L. innocua suspended in milk was investigated by

Noci and others (2009). When the milk was treated at 40 kV/cm for 25 μs, L. innocua was inactivated by 2.5 log10 CFU/mL, while on increasing the treatment time to 50 μs, the inactivation reached to 3.3 log10 cycles. This research concluded that increasing the electric field strength and treatment time resulted in an increase in inactivation rate.

Alkhafaji and Farid (2007) used SMUF to study the inactivation of E. coli with respect to

PEF electric field strength, treatment time and processing temperature. On increasing the electrical field strength from 37.2 to 43.4 kV/cm, E. coli inactivation also increased from 5.1 to

6.3 l log10. When they increased the inlet temperature of SMUF from 10 to 17 °C and treated at

43.4 kV/cm for 880 μs, the inactivation of E. coli was found to increase from 5.29 to 6.6 log10 reduction. They also reported an increase in lethal effect whenever there was an increase in treatment time.

In a review, Sampedro and others (2005) reported that a 5.0 log10 reduction of natural microflora and E. coli was achieved when milk protein concentrate-fortified soybean milk was treated by PEF at 35-40 kV/cm at ambient temperature. Dunn and others (1987) observed a 4.0 log10 reduction in count of S. Dublin in inoculated pasteurized homogenized milk treated by PEF at 36.7 kV/cm. Saldana and others (2012) investigated the effect of temperature, pH and other factors to evaluate the inactivation of S. Typhimurium and E. coli suspended in citrate–phosphate

McIlvaine buffer treated at PEF 30 kV/cm for 99 μs. When the pH and treatment temperature

36 was changed from 3.5 and 4 °C to 7.0 and 50 °C, the inactivation level was increased from 1.86 to 5.18 log10, and 0.93 to 3.76 log10 for S. Typhimurium and E. coli O157:H7, respectively.

Shin and others (2007) applied PEF treatment of 60 kV/cm with 1 μs pulse width, for 210 μs treatment time at 50 °C to UHT full fat milk inoculated with E. coli or Bacillus stearothermophilus. Reductions in count of 8 and 3 log10 CFU/mL were observed for E. coli and

B. stearothermophilus, respectively, with no change in pH even after PEF treatment. Li and others (2013) studied the effect of different parameters of PEF at 25 °C on E. coli inactivation in soy milk. When the electric field strength was increased from 20 to 40 kV/cm for 547 μs, the reduction of E. coli was increased from 2.7 to 5.2 log10 CFU/mL.

Simpson and others (1999) conducted a study on PEF inactivation of S. Typhimurium and L. monocytogenes suspended in distilled water, Tris–maleate buffer (TMB; pH=7.4), and a model beef broth (0.7% w/s). They did not find any significant effect of suspension medium on microbial destruction and inactivation of S. Typhimurium (6.0 log10 CFU/mL), which was more pronounced than for L. monocytogenes (4.0 log10 CFU/mL) when treated by PEF at an electric field strength of 20 kV/cm applied as 10,000 pulses.

Alvarez and others (2003) used citrate-phosphate McIlvaine buffer (pH=7.0) to investigate the inactivation of Salmonella spp. using a batch treatment PEF chamber. No linear relation was found between inactivation at specific electric field strength and treatment time. More than a 6.0 log10 reduction in count of S. Enteritidis and S. Typhimurium was observed when the bacterial suspensions were treated at 28 kV/cm for 1400 μs treatment time at 35 °C. Dunn (1996) investigated the destruction of L. innocua in milk using PEF at 55 °C for a few seconds and observed more than a 6.0 log10 cycle reduction in count with no change in physical and chemical

37 properties of the milk. In a study with skim milk, Calderón-Miranda and others (1999) used PEF alone and in combination with other preservation methods to analyse inactivation of inoculated

L. innocua. The highest inactivation (2.4 log10 CFU/mL) was found after PEF treatment at 50 kV/cm and 32 pulses.

Guerrero-Beltrán and others (2010) studied the inactivation of L. innocua in milk by the combination of varying inlet temperature and different operating conditions of PEF. A maximum reduction of 4.3 log10 was achieved when milk was treated with PEF at 30 kV/cm, with either of the combinations of 10, 17.5, 20 and 25 pulses, and inlet temperature of milk of 43, 33, 23 and

13 °C, respectively. In the same experiment, when the electric field strength was increased to 40 kV/cm, a maximum reduction of 5.5 log10 CFU/mL was observed at 15 pulses, with inlet temperature of 15 °C. They concluded that, by increasing the inlet temperature of milk and electric field intensity, a higher inactivation rate could be achieved even with a fewer number of pulses and this would be energy efficient.

Sampedro and others (2005) reviewed the PEF inactivation of L. monocytogenes suspended in McIlvaine buffer, high acid, low acid foods and milk with different fat content. They found a maximum reduction of 5.1 log10 CFU/mL in McIlvaine buffer (pH = 3.8), while McIlvaine buffer (pH = 7) resulted in only a 1.5 log10 reduction in count when samples were treated by PEF at 22 kV/cm for 800 μs with 2μs width pulse at 35 °C. In another experiment with PEF treatment, they found L. monocytogenes was more sensitive to PEF when present in high acid food (pH=3.7) compared with low acid food (pH=7). No protective effect of fat content of milk was observed, as milk with different fat content (1.5 and 3.5%) resulted in the same inactivation of L monocytogenes. After analysis of microbial inactivation results with different foods of the

38 same pH and electrical conductivity, they concluded that by adjusting the pH and electrical conductivity, suspension of cells in buffer could be used as a model to predict microbial inactivation of the organisms in different foods.

Fleischman and others (2004) used a static chamber to investigate the inactivation of L. monocytogenes in water-based gellan gum using different combinations of PEF parameters and thermal energy. Less than 1.0 log10 reduction in count was observed at electric field strength of

20-30 kV/cm and 5-50 pulses with 3.25 μs pulse width at or below 35 °C, while increasing the treatment temperature to 55 °C resulted in greater cell death. This study showed that L. innocua is difficult to inactivate using only PEF, and combinations of PEF with thermal energy produce higher reductions in count due to a synergistic effect.

Several researchers have suggested that application of a hurdle approach can enhance the microbial inactivation effect of PEF. Combination of PEF and thermal technology could be an approach to achieve higher microbial inactivation rate at low intensity of each technology, with no adverse effect on food quality (Devlieghere and others 2004, Buckow and others 2014, Heinz and others 2003). The synergistic effect of PEF and heat may increase the susceptibility of microorganisms due to a decrease in the “critical potential of membrane electrical breakdown”

(Floury and others 2006a).

In a review paper, Bendicho and others (2002) have concluded that PEF does not destroy the minor components of food, so, if inactivation of microorganisms is achieved to a safe level, this non-thermal technology could be regarded as a novel method to produce high quality food products. The effect of PEF processing on physico-chemical properties is summarised in Table

2.2.

In a PEF processing study, 2% fat milk was treated at 40 kV/cm, with pulse duration of 2 μs for 20 pulse number at maximum treatment temperature of 50 °C (Qin and others 1995). No significant change in physical and chemical properties was found. Moreover, sensory evaluation of heat pasteurized and PEF pasteurized milk did not show any significant differences in acceptability. In a similar study with fresh apple juice, no significant change in physical or chemical properties of fresh squeezed and PEF treated apple juice was observed and sensory evaluation also did not reveal any difference between the two products (Qin and others 1995).

Floury and others (2006b) investigated the effect of PEF processing at low temperature on different physical properties of skim milk, including viscosity and pH. PEF treatment of skim milk at 45-55 kV/cm (for 2.13-3.55 µs) did not produce any significant change in pH, whereas a decrease in viscosity was observed on increasing the electrical field intensity. Change in viscosity was probably due to “modifications of the hydrodynamic volume of casein micelles or of the mineral balance”.

Bermúdez-Aguirre and others (2011) studied the effect of PEF on physicochemical characteristics of skim and whole milk under different conditions of PEF (30.76-53.84 kV/cm at

20-40 °C). They observed a significant change in pH and electrical conductivity of skim and whole milk. Michalac and others (2003) conducted a study on skim milk and investigated the effect of PEF on physical properties. Physical properties of milk such as total solids, protein, pH, electrical conductivity, viscosity and density were not affected by PEF processing at 34.4 kV/cm with outlet temperature of 52 °C.

There are conflicting results regarding the effect of PEF treatment on the viscosity of milk.

Some researchers have shown a decrease in viscosity after batch PEF treatment, which may be

40 due to voluminosity of casein micelles and milk fat globules through aggregation. Results obtained with continuous PEF systems are more variable, with some studies showing that PEF results in a decrease in viscosity, whereas others observed no effect (Sharma and others 2014). In a review paper Sharma and others (2014) have described that PEF treatment of raw skim milk at

25-28 kV/cm (pulses of 2 µs at 200 and 400 Hz) at <45 °C did not affect the total solids, pH, and yield of cottage cheeses made from the milk and values were comparable to those obtained for product made from raw milk.

Mathys and others (2013) studied the effect of PEF and other parameters on physical properties of milk. PEF treatment at 5 - 45 kV/cm did not change the pH of milk, and electrical conductivity remained unchanged at lower electric field strength, whereas a significant change in electrical conductivity was observed at higher field strength.

Grahl and Markl (1996) did not find any significant difference in sensory evaluation of milk and orange juice after PEF processing. There was no any significant change observed for food components. They recommended the PEF treatment as a novel method of food processing to preserve sensitive food components.

Table 2.2. Summary of physico-chemical properties of pulsed electric field (PEF) treated products.

Treatment media PEF processing conditions Physico-chemical properties References Unchanged Changed

Skim milk E= 45-55 kV/cm, Tt = 2.13-3.55 µs, pH decrease in viscosity Floury and others

Ti = 10 °C, Tp= <50 °C (2006b) Skim and whole milk E=30.76-53.84 kV/cm, n = 12-30, pH and electrical Bermúdez-Aguirre

Tp= 20-40 °C conductivity and others (2011)

Skim milk E=35 kV/cm, Tt =188µs, total solids, pH, electrical Michalac and

Ti = 22 °C, To = 52 °C conductivity, viscosity others (2003)

Skim milk E=25-28 kV/cm , Tp = <45 °C, total solids, pH Sharma and others

To = <22 °C (2014)

Raw milk E = 5-45 kV/cm, Ti = 20-45 °C, To = color, pH, decrease in electrical Mathys and others 39-72 °C electrical conductivity (at conductivity (at (2013) lower electric field strength) higher field strength)

Yoghurt-based drink E= 30 kV/cm, Tt =32 µs pH and °Brix (total soluble Evrendilek and solids) others (2004b)

Raw skim milk, E=45 kV/cm, Tt =20µs, decrease in viscosity Hemar and others

Reconstituted skim milk Ti =25, To =30 °C (2011) (10% total solids), Concentrated skim milk, Milk protein concentrate (18% total solids)

Whole milk E=35.5 kV/cm, Tt =300-1000 µs, pH Odriozola-Serrano,

To= <40 °C and others (2006)

E = electric filed intensity, Ti = inlet temperature, To= outlet temperature, Tp = processing temperature, Tt = treatment time

Evrendilek and others (2004b) investigated the effect of a combination of heat and PEF treatment on microbial and physical properties in a yoghurt-based drink. No significant difference in pH and °Brix (total soluble solids) values was observed when the drink was first heated at 60 °C for 30 s and then cooled to 30 °C before PEF treatment at 30 kV/cm electric field strength with 1.4 µs pulse width for 32 µs.

In an investigation to study the effect of PEF treatment on viscosity, Hemar and others

(2011) found a decrease in viscosity of raw skim milk, reconstituted skim milk (10% total solids), concentrated skim milk, and milk protein concentrate (18% total solids) after PEF treatment at 45 kV/cm and 30 °C. A decrease in viscosity was again observed for the samples passed through the same system but without application of electric pulses. Moreover, when both samples were again analysed the following day after storing the sample at 4 °C, the viscosity increased and was equal to the value of the original sample. They assumed the decrease in viscosity was not due to the PEF treatment, but rather was due to the high shear effect and increase during storage was due to age-thickening. No change in physical properties of whole milk including pH has been reported following PEF treatment at 35.5 kV/cm and <40 °C

(Odriozola-Serrano, and others 2006).

Several studies described above show that PEF processing can be used to process liquid foods, which reduces the microbial load, thereby extending the shelf life of food, and destroying pathogenic microorganism to ensure food safety. This novel technology has also proven its efficiency to process food by retaining the natural taste and minimizing the damage to food components. The advantage of PEF processing is its energy efficiency, compared to thermal pasteurization. Barbosa-Cánovas and others (1999) found the energy consumption by HTST

43 processing was 364 J/mL to achieve microbial reduction in apple juice of 6.0 log10 cycles, whereas PEF processing required 28 J/mL to achieve the same degree of microbial inactivation.

Several PEF studies conducted with combination of heat treatment show that the efficiency of microbial inactivation can be enhanced by processing the food with hurdle approach at or below the non-lethal temperature. Therefore, PEF can be used as a novel non-thermal technology to process milk, either alone or in combination with another non-thermal technology, to replace the traditional thermal treatment.

2.6 Microfiltration processing of milk

Membrane filtration has been used in milk processing for a long time, and nowadays it is one of the most important processing techniques in the dairy industry. Membrane microfiltration

(MF) is a pressure-driven physical separation process suitable for separation of particles, somatic cells, bacteria and spores from liquid foods. It is one of the oldest membrane technologies; the first commercial membrane was available in the 1920s, and was mainly used for bacteriological analysis of water (Huisman 2000). According to its mode of application, MF can be classified into dead-end and cross-flow systems (Figure 2.6). Dead-end is one of the simplest modes, where feed flows perpendicular to the membrane and after some time, the permeate flux decrease, as a result of cake formation due to increase in volume of retained particles over the membrane (Cheryan 1998).

To improve the efficiency of MF filtration, in 1907 Bechhold found an increase in permeate flow when feed flowed parallel to the membrane, and this laid the foundation of cross-flow MF.

In cross-flow MF, feed flows parallel to the membrane surface, and the shear exerted by the feed solution flowing parallel to the membrane surface can sweep the deposited particles towards the

44 retentate side; thereby reducing the formation of the cake layer, and the cake remains relatively thin (Ripperger and Altmann 2002).

Dead-end MF Cross-flow MF

Feed • ο • • • ο ο ο ο• ο • Feed •ο • ο ο• •ο • • • • ο • •ο → ο ο • → Retentate • • ο ••ο → • • ο • • •ο ο • • ο ο • •ο ο ο • ο • ο ο •ο •ο ο ο ο ο • ο • •ο •• ο • • • •ο • • • ο • •ο • ●● ○○●○●●●○○•●●● ○●● ↓↓↓↓↓ ↓↓↓↓↓ Permeate Permeate

Time Time Figure 2.6. Different types of microfiltration (RC = cake resistance, J = flux).

Cheryan (1998) has described the use of MF in the dairy industry for fat separation, bacterial removal, spore removal and casein concentration (milk fractionation). This technology can be used in the food industry to clarify the food material and separate the suspended particles in the range of 0.10 - 5 μm (Cheryan 1998; Ripperger and Altmann 2002). Meersohn (1989) described

MF as a non-thermal technology, which can be used to extend the shelf life of milk by reducing the microbial load and remove spores from the milk, while keeping the organoleptic quality as before.

In MF the membrane has a microporous structure and separation of particles takes place according to particle size exclusion using a semi-permeable membrane. Berk (2013) has

45 described the basic requirement for the membrane as follows: the membrane should have high mechanical strength with chemical and thermal stability; have high resistivity towards different kinds of chemicals including detergents, acid, base used for cleaning purposes and also resistivity towards microbial action. Guerra and others (1997) pointed out the limitation of MF, as only applicable for skim milk processing due to clogging caused by fat globules.

In MF, solvent and smaller particles pass through the pores in the membrane due to a pressure gradient, leaving large molecules unfiltered. The portion passing through the membrane is known as permeate, while the unfiltered portion is rich in retained macromolecules, usually larger than the pore size and is smaller in volume compared with feed and permeate, thus it is known as retentate (Cheryan 1998).

MF processing can be carried out at low temperature to remove bacterial cells and spores and somatic cells, so, cooked flavour cause by thermal treatment can be avoided. Moreover, MF treated milk has high quality and extended shelf life, in comparison with heat processing, as, unlike MF, the latter does not remove the dead cell debris along with potentially active enzymes such as thermoduric proteases (Saboya and Maubois 2000, Olesen and Jensen 1989).

According to Berk (2013), membranes used for commercial application can be classified into organic polymer, which includes cellulose and its derivatives (mainly cellulose acetate), polyolefines, polysulfones, polyamides, chlorine and fluorine substituted hydrocarbons, while inorganic (ceramic) membranes are based on oxides of zirconium, titanium, silicium and aluminum. Membranes made of cellulose acetate are among the most popular in the dairy industry due to their low cost and low fouling characters. On the other hand, ceramic membranes

46 are well known for their longer operational life and resistance to cleaning chemicals and high temperature (Beckman and Barbano 2013, Cheryan 1998).

Membrane processing was first used in the dairy industry to separate milk components in the late 1960s, and now is widely used for whey and cheese processing (Gesan-Guiziou 2010).

Fernandez Garcia and others (2013) have recently reviewed the application of MF in the dairy industry and described that membrane processing was first used in the dairy industry to separate cream and skim milk using polymeric filters with pore sizes of 0.2 - 10 μm. Later on 2 μm ceramic membranes were successfully used to obtain skim milk virtually free of fat.

With the development of ceramic membranes in the 1980s, the use of MF became more widely used in the dairy industry and its major application was bacterial removal from skim milk. Membrane separation has wide potential application in the food industry including wine, juice, vegetables, brine, vinegar, gelatin, whey, beer as well as the dairy industry for separation of macromolecular solutes, for concentration and/or clarification (Saboya and Maubois 2000,

Vanderhorst and Hanemaaijer 1990). In a review paper, Kumar and others (2013) have reported that microorganism free milk is possible by “suitable modification in membrane structure, design, and composition” and Hoffmann and others (2006) found negligible (0.02 - 0.03%) loss of total protein in milk treated by MF. There is an economic benefit to using MF for cheese making as it increases production plant capacity and the amount of cheese production per day but without an increase in cheese yield (Papadatos and others 2003). In the process of cheese making, Awad and others (2010) used a ceramic membrane of 1.4 μm pore size to process raw milk at 50 °C and found a 5.18 log10 cycle reduction of total bacterial count.

The first commercial application of MF in the dairy industry was in Sweden, where it was used to remove spores of Bacillus spp. from milk. This resulted in an extension of shelf life to 16

- 21 days from 6 - 8 days, as well as producing milk with better quality than heat-treated product

(Daufin and others 2001).

Stack and Sillen (1998) along with Kumar and others (2013) have described the successful application of MF to remove bacteria and spores without affecting the quality of milk and removal was comparatively higher than that achieved by bactofugation (centrifugal separation used to remove bacteria and spores in milk). A summary of reports citing microbial removal by membrane microfiltration is described in Table 2.3. Membranes are already well established and are gaining popularity in the dairy industry (Daufin and others 2001), with the total MF area used in the dairy industry increasing to 15,000 m2 in 2010 from less than 750 m2 in the 1990s (Gesan-

Guiziou 2010). Membrane microfiltration has been used commercially in the dairy industry in

Canada in combination with other preservation methods (Elwell and Barbano 2006, Amornkul and Henning 2007). Several dairy companies in Europe are using MF as one of their unit operations in processing and it is becoming popular among consumers due to better flavour and longer shelf life.

Membrane microfiltration can reduce the microbial load in milk by 4 - 5 log10 cycles, but to meet the regulations governing the sale of milk in Canada, which states that microbial reduction must be from application of heat, MF can be only used along with thermal pasteurization to extend the shelf life of milk (Goff and Griffiths 2006). In the UK, the shelf life of milk has been extended up to 23 days by using MF and market share for this type of milk is up to 11% (Gesan-

Guiziou 2010). Some dairy companies in Canada, UK, and Scandinavia are using MF to process

48 skim milk obtained after separation of milk. The permeate is mixed with separately heated cream

(95 °C, 20 s), and finally this mixture is HTST pasteurized (72 °C, 15 s) before aseptic filling.

Consumer’s interest has increased towards this kind of processed milk due to absence of cooked flavour and long shelf life (up to 35 days) (Saboya and Maubois 2000).

Kosikowski and Mistry (1990) have described the successful removal of Clostridium tyrobutyricum spores in milk by 1.4 µm pore-sized MF and this was then used to prepare high quality cheese. Olesen and Jensen (1989) have reported the successful application of MF in milk processing to reduce the bacterial load without affecting quality (Elwell and Barbano 2006,

Maubois 2002). Membrane microfiltration efficiently removes bacteria and spores, but it does not ensure the removal of all pathogenic bacteria present in milk, so MF needs to be combined with thermal pasteurization (Rosenberg 1995).

Rodríguez-González and others (2011) observed a 2.1 log10 cycle reduction in counts of mesophilic microorganisms in skim milk, using cross-flow MF of 1.4 μm pore-size. Maubois

(2002) reported a reduction of vegetative cells in skim milk by >3.5 log10, when the milk was processed at 55 °C through MF of 1.4 μm pore size. The MF treated milk was free from somatic cells and spore reduction was more than 4.5 log10 cycles. Pafylias and others (1996) investigated the efficiency of 1.4 μm pore sized ceramic membrane MF to remove bacteria from inoculated reconstituted skim milk. An average reduction of 4.5 log10 in bacterial count was reported. They concluded that a substantial reduction in bacterial count in skim milk can be achieved without any significant change in milk composition.

Table 2.3. Summary of microbial reduction by membrane microfiltration.

Microorganism Treatment media Membrane processing conditions log10 References reduction

Total bacterial load Skim milk MF-1.4 μm, ceramic membrane, Tp= 50 °C 5.18 Awad and others (2010)

Total bacterial load Reconstituted skim MF- 1.4 μm, ceramic membrane, Tp= 50 °C 4.5 Pafylias and others (1996) milk

Total bacterial load Skim milk MF-1.4 μm, ceramic membrane, Tp= 55 °C >3.5 Maubois (2002)

Total bacterial load Skim milk MF- 1.4 μm, ceramic membrane, Tp= 6 °C No colony Fritsch and Moraru (2008) growth Total bacterial load Colostrum MF-0.8 μm, tubular ceramic ISOFLUX® membrane >5.4 Gosch and others (2014)

MF-1.4 μm, tubular ceramic ISOFLUX® membrane >3.5

Skim milk MF-0.8 μm, tubular ceramic ISOFLUX® membrane >2.3

MF-1.4 μm, tubular ceramic ISOFLUX® membrane >2.3

Total bacterial load Skim milk MF-1.4 μm, ceramic membrane, Tp=51 °C 4.13 Caplan and Barbano (2013) Total bacterial load Skim milk MF-1.4 μm 2.1-3.1 Daufin and others (2001)

Total bacterial load Skim milk MF-1.4μm, ceramic membrane 2-3 Gesan-Guiziou (2010)

MF-0.8µm, Sterilox® membrane 5-6

MF-1.4 µm, Sterilox® membrane 3-4

(continued on next page)

Table 2.3 (continued)

Total bacterial load Skim milk MF-1.4μm, ceramic membrane 3.79 Elwell and Barbano (2006)

Total bacterial load Skim milk MF-1.4μm, ceramic membrane >3.5 Saboya and Maubois (2000) Total bacterial load Skim milk MF-1.4μm 4 Olesen and Jensen (1989)

Spore forming Skim milk MF-1.4μm >4.5 Trouve and others (1991) bacteria

Spores Skim milk MF-1.4 μm, ceramic membrane, Tp= 55 °C >4.5 Maubois (2002)

Bacillus cereus Skim milk MF-1.4 μm, ceramic membrane 3.5 Olesen and Jensen spores (1989) Bacillus subtilis Simulated milk MF-0.5 μm (microsieve) 6.6 Brans and others (2004) ultrafiltrate

Tp= processing temperature

Fritsch and Moraru (2008) conducted a study to investigate the efficiency of MF to remove microorganisms, spores and somatic cells from skim milk at cold temperatures. They were unable to detect bacteria in permeate from skim milk having an initial count of 5.25 and 2.15 log10 CFU/mL of vegetative bacteria and spores, respectively following application of MF using a 1.4 μm pore size at 6 °C, and they also showed that the somatic cell count was reduced by 3.0 log10 cycles.

Gosch and others (2014) used 0.8 and 1.4 μm MF (tubular ceramic ISOFLUX® membrane) to process colostrum and skim milk. Microbial removal with a 0.8 μm MF membrane was more efficient with more than a 5.4 log10 reduction in total viable count being achieved, while, more than 3.5 log10 reduction in count was observed using membrane of 1.4 μm pore size. On the other hand, both types of MF reduced total viable counts by more than 2.3 log10 CFU/mL in skim milk.

They also used 0.14 and 0.2 μm MF and reported that permeate from both of these membranes were almost free (<1.0 log10 CFU/mL) from microorganisms. Caplan and Barbano (2013) applied 1.4-μm ceramic membrane MF treatment to study the efficiency of membrane filtration to remove bacteria and spores from milk. Skim milk processed through 1.4 μm MF at 51 °C, reduced the bacterial count by 4.13 log10 cycles while the spore count was found to be less than

1.0 log10.

Daufin and others (2001) reported a microbial reduction of 2.1 - 3.1 log10 CFU/mL when milk was passed through 1.4 μm MF, depending on initial count and morphology of bacteria, while Gesan-Guiziou (2010) reported a 2 - 3 log10 reduction in count following processing using a ceramic membrane of 1.4 μm size. However, the efficiency of Sterilox® membranes (Pall-

Exekia Company) is much higher due to narrow pore distribution size and can reduce the

52 microbial load by 5 - 6 and 3 - 4 log10 CFU/mL using 0.8 and 1.4 μm MF, respectively. Elwell and Barbano (2006) investigated the quality and storage stability of skim milk following MF using ceramic membranes with a pore size of 1.4 μm and found a 3.79 log10 reduction in bacterial count and reported that spore count was reduced to an undetectable level from initial counts of 2.0 log10 CFU/mL in raw milk.

In a review paper, Saboya and Maubois (2000) have reported >3.5 log10 reduction of bacterial count, and retention of all somatic cells in skim milk, when filtered through 1.4 μm pore size membrane at 50 °C. On comparing the results with 0.5 μm membrane processing, they found the bacterial reduction was increased by 2 - 3 log10 cycles when the smaller pore size membrane was used. Trouve and others (1991) have observed >4.5 log10 reduction of spore- forming bacteria when skim milk was treated with a 1.4 μm membrane.

Madec and others (1992) studied the effect of temperature and initial concentration of microorganisms in skim milk on retention efficiency of 1.4 μm pore size MF membranes.

Retention of Salmonella Abortus ovis was higher, probably due to its larger volumetric size, than

L. innocua. Another experiment with lower initial concentration did not result in any significant reduction, while, increasing the processing temperature to 50 °C increased the rates of inactivation achieved for both organisms.

Olesen and Jensen (1989) have applied 1.4 μm pore sized MF to investigate removal of bacteria and spores from skim milk. They observed 4.0 and 3.5 log10 reduction in total bacteria count and B. cereus spores, respectively. In a review paper Brans and others (2004) have described the use of a 0.5 μm microsieve, an advanced type of membrane filter. This type of

53 membrane has a narrow pore size distribution and can work at a low trans-membrane pressure.

They achieved a 6.6 log10 reduction in count of B. subtilis that was inoculated in SMUF.

Very little research has been done in the area of MF processing of milk to investigate physico-chemical properties. Use of microfiltration also limits the application of heat treatment and results in less damage to the nutritional and physicochemical properties of the raw material, so MF can play an important role to preserve liquid foods with natural properties, whilst ensuring food safety (Gesan-Guiziou 2010). It has been observed that pH and acidity of permeate obtained after MF 1.4 μm was the same as raw skim milk (Debon and others 2012).

Rodríguez-González and others (2011) have used combinations of PEF and ceramic membrane cross-flow microfiltration (CFMF) and reported that electrical conductivity of skim milk changed from 0.5 to 0.47 S/m after PEF (32 kV/cm at 42.3 °C) + CFMF-1.4 μm. No data for stand-alone CFMF was shown. Pafylias and others (1996) did not find any change in milk composition after MF using a 1.4 μm filter except in fat and microbial load.

2.7 Hurdle technology

The combination of more than one preservation technique to create a barrier for the microorganism, which makes them unable to survive and prevents their growth, is known as hurdle technology (Leistner and Gorris 1995). In hurdle preservation, microbial stability and safety can be achieved at low intensity of each preservation technique, thus, minimizing the negative impact on quality of foods, resulting in improved nutritional quality and sensory properties (Leistner 2000).

PEF is a proven technology for microbial inactivation, but in some cases, very high electric field with large number of pulses are required to achieve the desired degree of inactivation, which may sometimes produce technical as well as quality problems (Barbosa-Canovas and

Sepulveda 2005), so combinations of PEF, temperature and MF can be used as hurdles to ensure food safety.

As a non-thermal technology, microbial inactivation by PEF can be achieved at low (non- lethal) temperature, however, studies have reported the synergistic effect of temperature and PEF for microbial inactivation, i.e., the rate of microbial inactivation was increased on increasing the temperature at constant PEF parameters (Toepfl and others 2007, Pothakamury and others 1996,

Sensoy and others 1997, Zhang and others 1995). Rodríguez-González and others (2011) also studied the effect of temperature on microbial inactivation in milk subjected to PEF treatment. A greater reduction of mesophilic microorganisms was found following PEF treatment at 40 kV/cm, when the inlet temperature of milk was increased from 5.9 °C (outlet = 43 °C; reduction in count of 1.5 log10 CFU/mL) to 34 °C (outlet = 53.5 °C; reduction in count of 3.0 log10

CFU/mL).

Guerrero-Beltrán and others (2010) studied inactivation of L. innocua suspended in whole milk by PEF at 40 kV/cm with 5 pulses under different temperatures. They observed an increase in inactivation from less than 0.5 to 4.9 log10 on increasing the temperature from 23 to 53 °C.

Toepfl and others (2007) have reported that inactivation of E. coli in Ringers solution was increased from 2.0 log10 to 6.0 log10 when the PEF treatment temperature was increased from 35 to 55 °C, at electric field strength 16 kV/cm and 60 kJ/kg specific energy.

Sharma and others (2014) have investigated the effect of temperature on PEF treatment on microbial inactivation in whole milk. When milk was treated by PEF at ≥20 kV/cm, no significant reduction of E. coli, Staphylococcus aureus or L. innocua was achieved with inlet temperature at 4 °C, whereas, on increasing the inlet temperature to 50 °C, the inactivation was increased to ~2.0 log10. A maximum reduction of 5 - 6 log10 was observed on increasing the inlet temperature to 55 °C. An increase in bacterial inactivation with increase in temperature was evident due to an increase in the susceptibility of bacterial cells to electroporation at higher temperatures. Electroporation is the proposed theory for microbial inactivation, and this is a result of dielectric breakdown of cells due to differences in electric potential across the membrane (Jeyamkondan and others 1999). It is likely that an increase in temperature during

PEF treatment reduces the electric breakdown potential of the microbial cell membrane, which results in a higher degree of microbial reduction (Coster and Zimmermann 1975). This reduction could be due to temperature-dependent phase transition of the phospholipid molecules from a gel to a liquid-crystalline phase, which results in reduction in bilayer thickness making the cell membrane more prone to rupture at higher temperature even at low intensity of electric field

(Stanley 1991).

Heinz and others (2003) have reported the synergistic effect of PEF and temperature in apple juice processing. They observed the reduction in count of E. coli following PEF treatment was increased from less than 0.5 to 6.0 log10 CFU/mL when the treatment temperature was increased from 35 to 65 °C at an electrical field strength of 36 kV/cm and 10 kJ/kg specific energy. Sensoy and others (1997) studied the effect of temperature on inactivation of S. Dublin suspended in skim milk by PEF treatment and found the inactivation was increased from 3.0 to

4.0 log10 on increasing the temperature from 25 to 50 °C at an electric field strength of 25 kV/cm with pulse duration of 1 µs, for 100 µs treatment time.

Walkling-Ribeiro and others (2011a) investigated the inactivation of microorganisms in milk by a hurdle approach consisting of PEF and 1.4 µm pore-sized ceramic membrane MF. The reduction in total mesophilic aerobic counts in skim milk was 2.5 log10 CFU/mL when the skim milk was treated by PEF at 42 kV/cm for 612 µs at 49 °C, whereas, the combination of PEF/MF increased the microbial reduction to 7.1 log10, showing clear evidence of the effectiveness of the hurdle treatment.

2.8 Cheese making with PEF/MF treated milk

Milk is normally heat treated before cheese making to destroy the spoilage and pathogenic microbial populations present in milk (Lau and others 1991), however, this affects the microbiological, biochemical and chemical composition of cheese (Buchin and others 1998;

Ortigosa and others 2005), interaction of whey proteins with caseins, inactivation of indigenous enzymes (Amornkul and Henning 2007) and the flavor, taste and sensory quality of the cheese

(Morgan and others 2001, Johnson and others 1990, Wirjantoro and Lewis 1996). Cheese consumers are interested in cheese with typical flavor and texture (“cheeses belonging to their cultural patrimony”), characteristic of a geographical area, as well they need a high quality product with low food safety risk (Saboya and Maubois 2000).

Raw milk cheese is an important proportion of traditional cheese, and has more intense flavour than the same cheeses made from heat pasteurized milk and are characterized my higher concentration of alcohols, fatty acids and sulfur compounds than pasteurized milk cheese. The

57 manufacturing process and raw milk quality, both affect the sensory quality of cheese (Buchin and others 1998). Hydrolysis of lipids and proteolysis of protein compounds during ripening produces the flavour compounds of cheese (Grappin and Beuvier 1997), and volatile flavour components, along with small peptides and free amino acids produced during ripening also affect the flavour of cheese (Berger and others 1999, Sable and Cottenceau 1999, Muir and others

1996).

Cheese made from raw milk is popular among a large number of consumers, due to enhanced organoleptic properties, especially flavour, which is not commonly found in cheese made with pasteurized milk (Yu and others 2009), however, several foodborne illness outbreaks caused by Salmonella, Listeria, Campylobacter, Staphylococcus aureus and E. coli O157: H7 have been associated with consumption of raw milk and cheese made from it (Little and Knochel

1994, Baylis 2009, Fox 2000). Consumption of raw milk soft cheese has always been a concern from a microbiological point of view. Consumption of unpasteurized milk and its products are also associated with human infection by Mycobacterium bovis (organism associated with tuberculosis) (Rowe and Donaghy 2008). Langer and others (2012) reported that out of the 65 cheese associated foodborne disease outbreaks recorded in the USA during 1993-2006, 27 (42%) were from unpasteurized milk cheese. These data show that consumption of raw milk and cheese made from raw milk presents a potential public health risk. Therefore, non-thermal processing of milk, for example, with PEF could be an alternative technology in cheese making (Yu and others

2009).

Cheese-related foodborne illness outbreak data collected in the USA, Canada and Europe shows that surface mould-ripened soft cheeses (high moisture content), such as Camembert and

Brie always pose a greater risk to human health (Little and Knochel 1994). However, information of the effect of PEF treatment of milk on cheese making is very limited. Dunn

(1996) investigated PEF treatment (20-80 kV/cm, pulse duration = 1-10 µs) of milk and less flavour degradation was reported compared to conventional time-temperature pasteurization conditions, and proposed the manufacture of cheese with PEF treated milk. Sepulveda-Ahumada and others (2000) investigated the texture and sensory quality of cheddar cheese prepared with

PEF (35 kV/cm at temperature <30°C) treated milk, and suggested a possible option for cheese making with PEF treated milk for better texture and sensory aspects.

Membrane microfiltration is an alternative non-thermal technology to heat pasteurization of milk as it removes bacteria and improves the safety of milk products with no modification of the physicochemical properties of milk (Awad and others 2010). Use of mould spores is common in some of the cheese variety (Penicillum camemberti spores in Camembert and Brie type cheese), where as bacterial spores present in milk affect the keeping quality of milk and cheese, and these spores cannot be destroyed by heat processing normally used in cheese making, and MF processing is a better option to remove these bacterial spores. Some spore-forming bacteria, such as B. cereus, can cause foodborne illness and is a food safety concern (Pafylias and others 1996) and C. tyrobutyricum is also of concern in cheese making. The main source of Clostridia in milk is feeding of poor-quality silage and is commonly found in milk and, is a causative agent for late blowing of cheese especially in hard and semi-hard type cheese resulting in off-flavors and excessive gas formation (Wiedmann and others 1999).

Saboya and Maubois (2000) claimed that the cheese made from MF treated milk may be safer than cheese made from heat pasteurized milk. McSweeney and others (1993) have

59 described the benefits of cheese making with MF treated milk. Awad and others (2010) prepared cheese with MF treated milk and compared it with heat pasteurized milk cheese. The body and texture was found to be better in cheese made with MF treated milk.

In a review, Sampedro and others (2005) described the application of PEF treatment to milk and cheese made with this milk was compared with raw milk cheese, no difference was observed for enzyme activity, cheese yield, casein structure, and protein integrity. Some researchers have pointed out the benefits of cheese made for MF treated milk, and suggest they are of superior quality due to the absence of pathogenic as well as spoilage microorganisms (Meersohn 1989,

Hansen 1988). Gcsan-Guiziou (2010) has also reported that application of MF to produce

Camembert type soft cheese resulted in a product comparable with traditionally prepared cheese in terms of flavour and texture

For assurance of microbial safety, in Canada cheese made from unpasteurized milk must be

“held at a temperature of 2 °C or more for a period of 60 days or more from the date of the beginning of the manufacturing process” (Government of Canada 2014). However, investigations on survival of pathogens in cheese after 60 days’ ripening show the ability of E. coli O157:H7 to survive in Cheddar cheese for >120 d when stored at 6 - 7 °C (Reitsma and

Henning 1996). Other organisms are able to survive longer than 60 d, for example, S.

Typhimurium on Cheddar cheese at 7 °C, and on Colby cheese stored at 6.1 - 8.9 °C (Park and others 1970), S. Enteritidis in cheddar cheese stored at storage at 8 °C (Modi and others 2001), and L. monocytogenes on Cheddar cheese stored at 6 °C (Ryser and Marth 1987). Ramsaran and others (1998) used raw milk inoculated with E. coli O157:H7 and L. monocytogenes (at 104

CFU/mL) to prepare Camembert cheese. The analysis of cheeses after 65 days ripening, stored at

2 ± 1 °C, showed the presence of both pathogenic bacteria at levels higher than the initial inoculum. These results also show that there is a need for an alternative technology that can be used in cheese making to ensure food safety as well as to maintain the desired sensory characteristics.

Amornkul and Henning (2007) applied MF treatment to milk before Cheddar cheese making and compared the overall quality including microbial quality with raw milk cheese. Cheese made with MF treated milk was found to contain fewer E. coli than raw milk cheese and MF treated cheese showed “greater overall sensory acceptability” in comparision with raw milk cheese.

Therefore, combined application of non-thermal technologies, such as PEF and MF in white mould ripened soft cheese making could be an alternative to replace heat treatment or use of raw milk to ensure food safety by minimizing the microbial load, and producing cheese with typical flavour and taste similar to raw milk cheese.

CHAPTER 3: Impact of single and combined pulsed electric field and microfiltration treatments vs. heat pasteurization on microbial reduction and physico-chemical properties in skim milk

3.1. Abstract

Heat processing of milk has been practised for generations to ensure public health and food safety, but it also impacts adversely on the nutritional and organoleptic quality of the product, such as possible development of a cooked flavour, denaturation of protein, loss of vitamins. Non- thermal technologies, such as pulsed electric field (PEF) and membrane microfiltration (MF) represent a potential alternative for the production of minimally processed foods. Skim milk was processed by means of PEF and MF either alone or in combination and compared with high- temperature short-time pasteurization. PEF strengths, treatment times and energy densities used were 28 kV/cm, 2805 µs, and 83 kJ/L for PEF–L, 40 kV/cm, 1122 µs, and157 kJ/L for PEF–M, and 40 kV/cm,1571 µs, and 198kJ/L for PEF–H, while lab scale (MF-0.65µm and MF-1.2µm) and pilot scale (MF-1.4µm) MF were applied at trans-membrane pressures of 0.3 - 0.6 and 106 Pa, respectively. Raw milk stored at 8 °C for 24 h was used to determine the reduction of total aerobic counts following the above treatments, whereas physico-chemical properties (pH, electrical conductivity, °Brix, and viscosity) were analysed before and after processing of fresh raw milk. The latter was also used for inoculation with strains of Escherichia coli, Listeria monocytogenes or Salmonella spp. Microbial reduction by PEF-H was highest among the PEF treatments tested, whereas the highest reductions in count following microfiltration were achieved with MF-0.65µm, and MF-1.4µm. Combination of PEF-H/MF-0.65µm and PEF-H/MF-

1.4µm produced the highest reduction in microbial load in the milk of 5.95 and 5.90 log10

CFU/mL, respectively. Reductions in microbial count of 4.96 and 5.08 log10 were observed by

HTST-75 and HTST-95, respectively, and these reductions were not significantly different

(P≥0.05) from the reductions in count obtained using PEF-H/MF-1.2µm (5.0 log10), PEF-M/MF-

1.4µm (5.05 log10), and PEF-M/MF-0.65µm (4.97 log10). Electrical conductivity and pH were unaffected (P ≥0.05) by any of the treatments, while °Brix and viscosity values were significantly decreased (P <0.05) in the milk treated by MF-0.65µm and other treatments in which

MF-0.65µm was used. Counts of all three pathogenic bacteria were reduced to undetectable levels

(<10 CFU/mL) in the milk treated by PEF-H alone and by PEF-H/MF-1.2µm. These results indicate that a “hurdle” treatment in which PEF is combined with MF could be an alternative to thermal processing for milk.

3.2 Introduction

Bovine milk is a nutritious food for human beings; however, it is also an ideal growth medium for several microorganisms, including spoilage and pathogenic bacteria (Pereda and others 2009). Consequently, milk has a relatively short shelf-life. For many years, heat processing has been used in the food industry to extend the shelf life of food (Knorr and others

1994) by destroying pathogenic bacteria and lowering the load of spoilage microorganisms, thereby increasing the safety of the food. Although thermal processing is a low cost and simple technology, it may also have some disadvantages such as the possible development of a cooked flavour, undesirable changes in nutritional quality, denaturation of protein, loss of vitamins

(Wirjantoro and Lewis 1996), and destruction of natural antimicrobial components and immunoglobulins (Li-Chan and others 1995). It has been suggested that the food industry, by using only heat processing technology, is unable to meet the consumers’ demand for safe foods with extended shelf life that still retain a fresh taste and a high nutritional quality (Pereda and

63 others 2009). Therefore, the food industry is seeking to identify an alternative to traditional thermal technology to process food and minimize the adverse effects of thermal technology on food (Knorr and others 1994). Non-thermal technologies could be an alternative to thermal technology to minimize the adverse effects of heat on food quality since this technology runs at a low temperature (Barbosa-Cánovas and others 1999). Several researchers around the world have studied different non-thermal technologies, which include pulsed electric fields (PEF) (Walkling-

Ribeiro and others 2009, Sharma and others 2014, Walkling-Ribeiro and others 2011a,

Bermúdez-Aguirre and others 2011), high hydrostatic pressure (HHP) (Buzrul and others 2008,

Boluda-Aguilar and others 2013, Chen and others 2012, Rodiles-López and others 2010) microfiltration (MF) (Hoffmann and others 2006, Walkling-Ribeiro and others 2011a,

Rodríguez-González and others 2011), ultrasound (Khanal and others 2014, Zenker and others

2003), and irradiation or ultra-violet light (UV) (Matak and others 2007). Due to a short treatment time and a low level of heat processing, PEF is among the most appealing non-thermal technologies (Ho and Mittal 2000,Ho and others 1995,Barbosa-Canovas and others 2000) as it does not result in colour alteration, flavour damage or nutrient losses (Giner and others 2000); resulting in a product that maintains a high quality throughout its shelf-life (Cortés and others

2008).

In PEF processing, food is placed between two electrodes and then very short pulses (1-10

μs) of a several thousand volt electric field is applied across the food (Toepfl and others 2007,

Barbosa-Canovas and others 2000, Ho and Mittal 2000). Electric field strength and treatment time are important parameters for microbial inactivation in PEF processing. Other processing conditions such as temperature, pulse shape, pulse frequency, product flow rate, product composition, and electrical conductivity of the product also affect microbial inactivation (Shamsi

64 and others 2008). This technology is gaining popularity in terms of commercial interest and is progressing from the lab and pilot scale to the commercial scale (Devlieghere and others 2004,

Toepfl 2011). Electroporation is generally thought to be the mechanism of microbial inactivation, as it explains why the electric field results in the creation of pores in cell membranes, the formation of which leads to an increase in membrane permeability, allowing loss of cell contents and cell death (Zimmermann 1986, Pothakamury and others 1997). PEF has been applied to several products, such as skim milk (Walkling-Ribeiro and others 2011a, Bermúdez-

Aguirre and others 2011), whole milk (Bermúdez-Aguirre and others 2011, Sharma and others

2014), apple juice (Walkling-Ribeiro and others 2008,Bi and others 2013, Aguilar-Rosas and others 2007, Moody and others 2014), orange juice (Rivas and others 2006, Agcam and others

2014, Walkling-Ribeiro and others 2009a), grape juice (Huang and others 2014), smoothie-type beverages (Walkling-Ribeiro and others 2010, Walkling-Ribeiro and others 2008), liquid whole egg (Monfort and others 2012), beer (Walkling-Ribeiro and others 2011b), and wine (Puértolas and others 2009). These studies include the effect of PEF on the reduction of total bacterial count and other groups of bacteria in milk (Odriozola-Serrano and others 2006, Bermúdez-Aguirre and others 2011, Walkling-Ribeiro and others 2011a). The inactivation of the native microbiota of milk (Michalac and others 2003, Fernandez-Molina and others 2005, Rodríguez-González and others 2011), as well as Escherichia coli (Rodriguez-Gonzalez and others 2011, Zhao and others

2013, Sharma and others 2014), Listeria monocytogenes (Sharma and others 2014, Zhao and others 2013, Picart and others 2002,Noci and others 2009) and Salmonella populations (Saldaña and others 2012,Sensoy and others 1997) have been studied in milk using PEF processing.

Previous studies (Walkling-Ribeiro and others 2011a, Shamsi and others 2008) have also shown

65 that the extent of microbial inactivation produced by PEF is increased when it is performed at temperatures higher than room temperature (for example, at around 55 °C).

Combinations of PEF and other non-thermal technologies have been investigated as a hurdle approach (Leistner 2000) to increase the efficiency of processing and to minimize the intensities of the treatments applied. These non-thermal hurdle-processing approaches include the combination of PEF and ultrasonication (Halpin and others 2013, El Darra and others 2013,

Palgan and others 2012, Huang, Mittal, Griffiths 2006), combination of PEF and HHP (Huang and others 2006, Heinz and Knorr 2000) and PEF and MF (Rodríguez-González and others 2011,

Walkling-Ribeiro and others 2011a). MF can be used in milk processing to reduce the bacterial load without affecting quality (Mcsweeney and others 1991, Olesen and Jensen 1989, Elwell and

Barbano 2006). MF is a pressure-driven separation process using membranes of pore size from

0.1 to 2.0 μm (Kumar and others 2013). Since MF can reduce microbial load by 4 - 5 log10 cycles

(Goff and Griffiths 2006), this can be used to replace heat processing (Maubois 2002) and improve overall quality and keeping quality of milk by, for example, removing spores (Kumar and others 2013), which cannot be destroyed during heat processing. MF treated milk has a high quality in comparison with heat treated milk, as the latter does not remove debris associated with dead cells and heat-resistant enzymes, such as proteases and lipases produced by psychrotrophic bacteria; both of which are removed during microfiltration (Saboya and Maubois 2000).

Membrane separation is an established technology within the dairy industry (Daufin and others

2001), and is widely used by dairy processors in Canada (Elwell and Barbano 2006).

Microfiltration has been successfully used in combination with PEF to reduce the microbial load of skim milk (Walkling-Ribeiro and others 2011a). A combination of the two technologies is

66 advantageous because, whereas PEF processing may not destroy bacterial spores (Raso and

Barbosa-Canovas 2003), they can be removed from milk by MF (Meersohn 1989).

Hurdle processing of skim milk by the combination of PEF and MF could be a non-thermal alternative to heat pasteurization of skim milk to ensure public health and improve keeping quality. However, research on the effects of PEF and MF with respect to the physical properties of skim milk is limited. Therefore, this study determined the influence of PEF and MF treatment of skim milk, alone and in combination, on total microbial load, physico-chemical properties and inactivation of relevant pathogenic bacteria as compared to high-temperature short-time (HTST) pasteurization.

3.3 Materials and methods

3.3.1. Skim milk preparation and storage

Whole milk was obtained from a local dairy (Gay Lea foods Co-operatives, Guelph, ON,

Canada). The collected milk was filtered before skimming to ensure removal of any suspended particles, warmed to 35 °C and skimmed using a cream separator (STsM-100-18, Motor Sich

JSC, Zaporozhye, Ukraine) to give a fat concentration of 0.05 - 0.1% w/v and then incubated at 8

°C for 24 h to promote the growth of the native microbiota to levels approaching 7.0 log10

CFU/mL. The resulting skim milk was further cooled and stored at 4 °C until used.

3.3.2. PEF treatment

For non-thermal treatment, fresh skim milk, skim milk incubated at 8 °C for 24 h and skim milk inoculated with the pathogens under study were processed using a generator (PPS 30,

University of Waterloo, Waterloo, ON, Canada), which generated an exponential decay pulse of

1.5 µs pulse width. A single coaxial treatment chamber with two electrodes was used for treatment of skim milk. The electrode diameter was 7.53 cm and the distance between the two electrodes was 0.21 cm. Voltage, pulse width, pulse shape, and frequency were recorded with a digital oscilloscope (TD 1012, Tektronix Inc., Beaverton, OR, USA). Water was circulated through the PEF chamber to maintain the desired temperature during processing. A cooling bath

(NESLAB RTE 7, Thermo Fisher Scientific Inc., Newlington, NH, USA) set at 2 °C was attached to the system, and treated milk was passed through this cooling bath before the sample was collected. Untreated as well as treated samples were kept at 4 °C until they were analysed.

Thermocouples (TMTSS-040G-6, Omega, Stamford, CT, USA) were connected to the system to monitor the temperature of milk at the inlet and outlet to the chamber and also of the water circulating through the chamber. The temperatures were recorded using a wireless temperature data logger (OM-SQ2020-2F8, Omega, Stamford, CT, USA). The skim milk was processed by

PEF under the following conditions of electric field strength, product flow rate, and processing temperature: 28 kV/cm, 20 mL/min and 37-40 °C for an energy density up to 83 kJ/L (PEF-L);

40 kV/cm, 35 mL/min and 50-55 °C for an energy density up to 157 kJ/L (PEF-M); and 40 kV/cm, 25 mL/min and 60-65 °C for an energy density up to 198 kJ/L (PEF-H), for treatment times of 2805, 1122 and 1571 μs, respectively. A peristaltic pump (Masterflex pump drive 7524-

40 and pump head 77201-60, Cole Parmer Instrument Co., Vernon Hills, IL, USA) was used to pump the skim milk to the PEF system using Masterflex silicone tubing (size L/S 16). The PEF settings were chosen based on the results of a previous study, which also provides details of the equipment used (Walkling-Ribeiro and others 2011).

3.3.3. Microfiltration

Three different pore-sized MF filters were used to process the skim milk. Two of these (0.65

µm pore-size, Supor, medium screen, 1.4 cm thick and 1.2 µm pore-size, Supor, suspended screen, 1.8 cm thick polyethersulfone membranes; Pall Corporation, Port Washington, NY) were used with a bench top tangential flow microfiltration (TFMF) system (part no. FS001K 10, Pall

Corporation, Port Washington, NY, USA) and the third (1.4 µm pore-size ceramic membrane,

1P19-40 SCT Membralox®, Tetra Pak Filtration Systems, Aarhus, Denmark) was used in a pilot- scale cross-flow microfiltration (CFMF) system (MFS-1, Tetra Pak Filtration Systems, Aarhus,

Denmark). Skim milk was processed through the bench top microfiltration unit at a flow rate of

300 mL/min with a peristaltic pump (Masterflex EW 77200-60 pump, Cole Parmer Instrument

Co., Vernon Hills, IL, USA), generating 0.3 - 0.6 kPa trans-membrane pressures. The permeate flow rate was 7 - 13 mL/min for milk passing through the 0.65 µm membrane (MF-0.65µm), whereas it was 37 - 53 mL/min for the 1.2 µm (MF-1.2µm) membrane. Permeate and retentate flow rate for the pilot scale MF (MF-1.4µm) system was maintained at 100 L/h and 10 L/h, respectively, and the trans-membrane pressure was 106 kPa. The surface area of the membranes used in the bench top and pilot scale units was 0.093 m2 and 0.2 m2, respectively and milk was processed at 35 °C for all MF treatments. A water bath (Isotemp 210, Thermo Fisher Scientific

Inc., Newlington, NH, USA) was used to preheat the skim milk before MF processing. The milk sample was cooled before and after processing, and kept at 4 °C until analysed. The TFMF membranes were cleaned/disinfected before and after use with 0.5 M sodium hydroxide solution followed sequentially by 0.1 M nitric acid and 2% bleach at 50 °C and a final rinse with sterile distilled water. The pilot scale CFMF ceramic membrane was cleaned/disinfected before and after processing by a commercially available phosphoric acid solution (Diversey, Divos2

(VM13), Diversey, Oakville, ON, Canada) and a caustic detergent (Diversey Liquid Bril Tak

(VC85) followed by a final rinse with water.

3.3.4. Combined PEF/MF treatments

PEF and MF were combined to process the milk. Skim milk was processed by PEF-L, PEF-

M or PEF-H, followed by MF processing, as described above. Due to limitations in processing a large volume of milk through PEF-L, the pilot scale CFMF (MF-1.4µm) was carried out only in combination with PEF-M or PEF-H.

3.3.5. HTST treatment

Skim milk was pasteurized using a conventional, pilot-scale, dual stage heat exchanger

(UHT/HTST Lab-25 EDH, Microthermics Inc., Raleigh, NC, USA) at the Guelph Food

Technology Centre (GFTC), Guelph, ON. Pasteurization of skim milk was carried out at 75 °C

(HTST-75) for 20 s (flow rate 0.9 L/min) or 95 °C (HTST-95) for 45 s (flow rate 0.4 L/min).

After treatment, milk was cooled to 10 °C and then further cooled to 4 °C, at which temperature it was kept until analysed.

3.3.6. Microbiological analysis

Untreated and treated milk samples were serially diluted with Ringers solution (BR0052G,

Oxoid Ltd., Basingstoke, Hampshire, UK) for the analysis of total aerobic counts. An aliquot of

100 μL was surface plated on a milk agar plate (MPCA; CM0681, Oxoid Ltd., Basingstoke,

Hampshire, UK) under aseptic conditions in a class 2 biosafety cabinet (Forma 4071-1200,

Thermo Fisher Scientific Inc., Newlington, NH, USA) for the analysis of the total aerobic counts

(TAC). Agar plates were incubated at 30 °C for 72 h (Walkling-Ribeiro and others 2011b).

3.3.7. Physico-chemical properties of milk

The pH of the milk samples before and after processing was determined by using a direct immersion pH meter (Accumet AB15, Thermo Fisher Scientific, Newlington, NH, USA).

Similarly, for the measurement of electrical conductivity of the milk sample, a direct immersion conductivity meter (Oaktron Con 11 Series, Thermo Fisher Scientific Inc., Newlington, NH,

USA) was used. A refractometer (Pocket Digital Refractometer, SPER Scientific, Model-

300053, China) was used to determine the °Brix of the milk before and after processing, and for viscosity, a falling ball viscometer (Gilmont Instruments Viscometer size # 1, GV-2100, Thermo

Fisher Scientific, USA) was used. All measurements were carried out at room temperature.

3.3.8. Pathogen challenge study

Three of each strain of E. coli, Salmonella spp. and L. monocytogenes were used for the challenge studies. The bacteria were obtained from the culture collection of the Canadian

Research Institute for Food Safety (CRIFS) University of Guelph, Guelph, ON, Canada. Three strains each of shiga toxin producing E. coli (E. coli O157:H7 (C 477), E. coli O103:H2 (C 469), and E. coli O145:NM (C 757)), Salmonella enterica (S. Typhimurium DT 104 (C 1061), S.

Enteritidis (C 721) and S. Newport (C 1158)), and L. monocytogenes (L. monocytogenes ½ a (C

1324), L. monocytogenes LO1-665 (C 890), and L. monocytogenes P 7 (C 1301)) were obtained in cryovials at -80 °C. For the preparation of stock cultures, 100 μL of these cryovial cultures was transferred into 10 mL of tryptic soy broth (TSB; Becton, Dickinson and Company, Sparks,

MD, USA) and incubated at 37 °C for 24 h in a shaking incubator. The cultures were subsequently streaked onto respective selective media and incubated at 37 °C for 24 h. A single colony was obtained from the plates and used to inoculate TSB, which was subsequently incubated at 37 °C for 24 h with shaking. This culture (100 μL) was used to inoculate 10 mL of fresh TSB and this suspension was incubated at 37 °C for 24 h with shaking. This culture was diluted in TSB and used to inoculate raw skim milk to obtain initial bacterial populations of approximately 4.0 log10 CFU/mL. The inoculated milk samples were processed using PEF-H, and PEF-H/MF-1.2µm, and PEF-L/MF-1.2µm.

To enumerate the bacteria, milk was serially diluted with Ringers solution and a 100 μL aliquot was surface plated onto the appropriate selective medium: Listeria monocytogenes blood agar for L. monocytogenes was prepared as described by Johansson (1998), ChromAgar O157

(Becton, Dickinson and Company, Sparks, MD, USA) for E. coli, and BrillianceTM Salmonella

Agar Base (Oxiod-CM1092, Oxoid Ltd) for Salmonella spp. The plates were incubated at 37 °C for 24 h. The ability of bacterial cells to survive following treatment at numbers that were not detectable by plate count was assessed by incubating 10 mL of treated milks at 37 °C for 24 h, followed by enumeration using the appropriate medium as described above.

3.3.9 Statistical analysis

The proximate composition, color, flavor, and microbial screening data obtained in the present study were analysed statistically using SPSS version 21.0 (IBM Corporation, Chicago,

USA). Three different product batches were processed (n = 3) and analyzed by analysis of variance (ANOVA) at a confidence interval of 95%. Upon occurrence of statistical differences,

72 treatment means were compared with the Tukey’s test for multiple pair wise comparisons. Each treatment was carried out in triplicate and analysis was performed in duplicate.

3.4 Results

3.4.1. Suitability of PEF, MF, PEF/MF, and HTST for microbial reduction

After incubation at 8 °C for 24 h, initial microbial load in raw skim milk was slightly higher than 7.0 log10 CFU/mL. Total microbial reductions in milk after stand-alone PEF, MF, combination of PEF/MF, and HTST are shown in Figure 3.1. Microbial loads in the milk processed by HTST-75 and HTST-95 were reduced by 4.96 and 5.08 log10 cycles, respectively.

Similarly, reductions in count of 4.97, 5.0 and 5.05 log10 CFU/mL were achieved following treatment by PEF-M/MF-0.65µm, PEF-H/MF-1.2µm and PEF-M/MF-1.4µm, respectively, and these were not significantly different (P ≥0.05) from those obtained with HTST processing. High intensity PEF processing (PEF-H) produced the highest level of reduction (P <0.05) of 4.13 log10, among all stand-alone PEF treatments. Moreover, the highest reductions in count obtained by MF were 3.13 and 3.08 log10 CFU/mL by MF-0.65µm and MF-1.4µm processing, respectively.

Following the MF-1.2µm treatment there was a reduction in count of 1.61 log10 CFU/mL in the milk. Hurdle processing of milk using a combination of a high intensity pulsed electric field

(PEF-H) and 0.65 μm pore sized MF (PEF-H/MF-0.65µm) and PEF-H/MF-1.4µm led to the highest reductions of native microorganisms in milk (P ≥0.05) when compared with all other treatments; being 5.95 and 5.90 log10 CFU/mL, respectively.

g g 6

f f f f f CFU/mL)

5 10 e e d,e 4 c,d c c b,c 3 a,b 2 a

1 Microbial inactivation(log Microbial

Figure 3.1. Inactivation of native total aerobic microorganism in raw skim milk processed with HTST (HTST-75 =

75 °C for 20 s; HTST-95 = 95 °C for 45 s), lab-scale tangential flow microfiltration (MF-0.65µm = 0.65µm, MF-

1.2µm = 1.2 µm), pilot scale cross-flow microfiltration (MF-1.4µm = 1.4 µm), pulsed electric field at low (PEF-L: 28 kV/cm; 2805 µs), moderate (PEF-M: 40 kV/cm, 1122 µs), and high (PEF-H: 40 kV/cm, 1571 µs) processing intensities and selected nonthermal combination treatments using PEF and subsequent microfiltration. Initial microbial load in raw skim milk was 7.46 (±0.14) log10 CFU/mL. Same letters above the column stand for no statistical differences between two treatment means (n = 3).

3.4.2. Effect of nonthermal and thermal processing on physico-chemical properties

The physico-chemical properties determined were pH, electrical conductivity, °Brix and viscosity. The results are presented in Table 3.1. The pH and electrical conductivity of fresh raw skim milk were 6.64 and 4.71 mS/cm, respectively. These values did not change significantly (P

≥0.05) after any of the treatments were applied to the milk. The °Brix of raw skim milk was

10.56, which was lowered significantly (P <0.05) to 8.14 after processing by stand-alone 0.65

µm pore sized microfiltration (MF-0.65µm). Similarly, reductions in °Brix values to 8.17, 8.09

74 and 8.21 were observed after hurdle processing using PEF-L/MF-0.65µm, PEF-M/MF-0.65µm and

PEF-H/MF-0.65µm, respectively. The remaining treatments did not result in a significant change

(P ≥0.05) in °Brix value of skim milk.

Changes in viscosity of the milk followed the same trend as °Brix. Raw skim milk viscosity was 1.94 cP, and decreased (P <0.05) to 1.83 cP after stand-alone MF-0.65µm processing.

Similarly, a decrease (P <0.05) in viscosity was observed after hurdle processing of PEF and

MF-0.65µm. The value for viscosity was 1.83, 1.81 and 1.81 cP after PEF-L/MF-0.65µm, PEF-

M/MF-0.65µm and PEF-H/MF-0.65µm, respectively. No significant change (P ≥0.05) in viscosity of skim milk was observed when the skim milk was processed with any of the other treatments.

Table 3.1. Physico-chemical properties of skim milk after high-temperature short-time pasteurisation (HTST-75 =

75 °C for 20 s; HTST-95 = 95 °C for 45 s), lab-scale tangential flow microfiltration (MF-0.65µm = 0.65µm, MF-

Conductivity Viscosity Treatment pH1 °Brix1 [mS/cm]1 [cP]1 Raw skim milk 6.64(±0.13)a 4.71(±0.27)a 10.56(±0.26)a 1.94(±0.02)a

a a b b MF-0.65µm 6.59(±0.03) 4.81(±0.09) 8.14(±0.20) 1.83(±0.01) a a a a MF-1.2µm 6.70(±0.12) 4.82(±0.16) 10.40(±0.08) 1.92(±0.02) a a a a MF-1.4µm 6.62(±0.05) 4.53(±0.02) 10.43(±0.04) 1.91(±0.02) PEF-L 6.71(±0.09)a 4.69(±0.33)a 10.69(±0.13)a 1.92(±0.03)a PEF-M 6.64(±0.10)a 4.57(±0.35)a 10.61(±0.25)a 1.91(±0.02)a PEF-H 6.65(±0.06)a 4.32(±0.04)a 10.59(±0.05)a 1.92(±0.03)a

a a b b PEF-L/MF-0.65µm 6.70(±0.18) 4.82(±0.28) 8.17(±0.08) 1.83(±0.03) a a b b PEF-M/MF-0.65µm 6.71(±0.11) 4.64(±0.29) 8.09(±0.09) 1.81(±0.03) a a b b PEF-H/MF-0.65µm 6.78(±0.13) 4.83(±0.26) 8.21(±0.03) 1.81(±0.01) a a a a PEF-L/MF-1.2µm 6.70(±0.08) 4.77(±0.15) 10.41(±0.05) 1.95(±0.05) a a a a PEF-M/MF-1.2µm 6.67(±0.04) 4.69(±0.25) 10.38(±0.13) 1.94(±0.04) a a a a PEF-H/MF-1.2µm 6.73(±0.09) 4.64(±0.19) 10.18(±0.15) 1.92(±0.04) a a a a PEF-M/MF-1.4µm 6.65(±0.04) 4.66(±0.04) 10.33(±0.04) 1.94(±0.02)

a a a a PEF-H/MF-1.4µm 6.71(±0.02) 4.69(±0.03) 10.30(±0.04) 1.92(±0.03) HTST-75 6.63(±0.02)a 4.58(±0.11)a 10.59(±0.44)a 1.92(±0.02)a HTST-95 6.58(±0.02)a 4.61(±0.08)a 10.62(±0.30)a 1.94(±0.02)a P2 NS NS * * SEM3 0.009 0.018 0.143 0.007 a,b Different superscripted letters in the same column indicate statistical significance between the two means. 1Values in parentheses following the mean values of physico-chemical property indicate the standard deviations. 2P stands for the statistical probability; NS indicates no statistical significance, * refers to a statistical significance of P < 0.05, 3SEM abbreviates the standard error of the mean.

3.4.3. Pathogen control by PEF and PEF/MF

Raw skim milk was inoculated with pathogenic bacteria and the inactivation level was compared with stand-alone high intensity PEF (PEF-H) and hurdle combinations of PEF and

MF-1.2µm (Figure 3.2). The initial counts of pathogenic bacteria in raw skim milk after inoculation were 4.61, 4.60 and 4.75 log10 CFU/mL for E. coli, L. monocytogenes and

Salmonella spp., respectively. A low level of inactivation was observed after treatment with the combination of low intensity PEF and MF-1.2µm (PEF-L/MF-1.2µm), when reductions in count of

1.93, 2.78 and 1.15 log10 CFU/mL for E. coli, L. monocytogenes and Salmonella spp., respectively, were observed.

PEF-L/MF-1.2µm PEF-H PEF-H/MF-1.2µm

CFU/mL)

10 3

1 Bacterial inactivation (log inactivation Bacterial 0 Salmonella ssp. E. coli L. monocytogenes

Figure 3.2. Pathogen reduction obtained with pulsed electric fields at low (PEF-L: 28 kV/cm; 2805 µs), and high (PEF-H: 40 kV/cm, 1571 µs) processing intensities and selected non-thermal combination treatment using PEF and subsequent lab-scale tangential flow microfiltration (MF-1.2µm = 1.2 µm) (PEF-H, PEF-L/MF-1.2µm, PEF-H/MF-

1.2µm). Initial counts of Salmonella spp. = 4.75 (±0.10), E. coli = 4.61 (±0.11) and Listeria monocytogenes = 4.60

(±0.15) log10 CFU/mL; columns without error bars (standard deviation) indicate that the detection limit (1.0 log10 CFU/mL) was reached (n = 3).

Inactivation to levels below the limit of detection (< 10 CFU/mL) was observed for all the pathogenic bacteria in the sample treated with stand-alone high intensity PEF (PEF-H) and in combination with MF-1.2µm (PEF-H/MF-1.2µm), and the organisms could not be detected following subsequent incubation of the milk samples at 37 °C for 24 h (Table 3.2).

Table 3.2 Pathogen bacterial cell recovery from skim milk after pulsed electric fields at low (PEF-L: 28 kV/cm; 2805 µs), and high (PEF-H: 40 kV/cm, 1571 µs) processing intensities and selected non-thermal combination treatment using PEF and subsequent lab-scale tangential flow microfiltration (MF-1.2µm = 1.2 µm) (PEF-H, PEF-

L/MF-1.2µm, PEF-H/MF-1.2µm).

Treatment Cell recovery Salmonella spp. E. coli L. monocytogenes

PEF-L/MF-1.2µm N/A N/A N/A PEF-H No No No

PEF-H/MF-1.2µm No No No

3.5. Discussion

The microbial reduction in milk by HTST is in agreement with the findings of other researchers. Halpin and others (2013) and Rodríguez-González and others (2011) have reported

4.0 and 4.7 log10 cycle reductions in total microbial count for whole milk heated at 72 °C for 20 s and 72 °C for 15 s, respectively, whereas a count reduction of 6.1 log10 CFU/mL has been reported by Walkling-Ribeiro and others (2009b) following HTST processing at 72 °C for 26 s.

Differences in the results observed are probably due to differences in initial microbial counts and types of microbial groups present in the milk. The higher reduction obtained with PEF-H also coincides with the results of other researchers showing that a higher electric field (Knorr and others 1994) and an elevated temperature (Shamsi and others 2008, Rodríguez-González and others 2011) promote increased bacterial inactivation. The higher microbial inactivation at high intensity electric field (PEF-H) is attributed to the fact that high electric field strength produces larger trans-membrane potential, thereby increasing the chances of pore formation leading to cell death (Fox 2007). At higher temperatures the cell membrane turns into a liquid-crystalline phase from a gel structure, resulting in reduction in bilayer thickness, making the cell membrane more prone to rupture, and high temperature also causes a decrease in elasticity of the cell membrane, making it more fragile (Stanley 1991). Floury and others (2006b) have also described that PEF processing at a higher temperature causes a decrease in electrical breakdown potential of cell membranes, and the synergistic effect of PEF and temperature results in higher microbial inactivation. Bermúdez-Aguirre and others (2011) observed <1 log10 cycle reduction in count of mesophiles when skim milk was treated at 46.15 kV/cm (0.2 kHz, 2μs pulse width) and 30 °C, and the reduction increased to >3.5 log10 CFU/mL when the skim milk was treated at 60 °C, with the same PEF intensity. Fernández-Molina and others (2005) compared the microbial

78 inactivation in skim milk subjected to different PEF intensities. Microbial reduction by <1 log10

CFU/mL was observed at 28 kV/cm and temperature <32 °C (pulse width 2.8 µs and treatment time = 84 µs), whereas, the reduction increased to >2 log10 CFU/mL on increasing the PEF intensity to 36 kV/cm, keeping all other processing conditions the same.

Lower levels of microbial inactivation by PEF treatment have also been reported by other research groups, who cited reductions in total bacterial counts of skim milk and whole milk of >2 log10 following treatment at 35 kV/cm with pulse width of 2.3 µs (Sepulveda and others 2009) and 1.0 log10 cycle when processed at 35.5 kV for 1000 µs (Odriozola-Serrano and others 2006).

The lower reduction in microbial count observed in these experiments may be related to differences in equipment design and diversity and initial concentrations of microbial load in the milk. Contrasting results were found by Walkling-Ribeiro and others (2009b), who reported that reductions in total microbial load in milk of 5.0 and 6.4 log10 were achieved following PEF treatment at 40 kV/cm for 50 µs and 50 kV/cm for 33 µs, respectively. The difference in microbial reduction obtained in the present study may be explained by variations in the initial concentration and composition of the microbiota in the milk as well as different operating conditions and design of treatment chambers.

With regard to MF, possible reasons for achieving higher levels of microbial removal by

MF-1.4µm than MF-1.2µm include the use of different membrane materials (ceramic vs. Supor plastics, respectively) and membrane module layouts, and the increased surface area of the former MF system. The results obtained in this study are similar to others that used a ceramic membrane with 1.4µm pore-size, in which reduction in counts of 3.7 log10 CFU/mL (Walkling-

Ribeiro and others 2011a) and 3.5 log10 CFU/mL (Brans and others 2004) were achieved. Other

79 similar studies conducted on skim milk processed by MF through a 1.4µm pore-size ceramic membrane found a decrease of the microbial load by 2.7 log10 CFU/mL (inlet and outlet pressures of 90, and 20 kPa) (Rodríguez-González and others 2011), and 4-5 log10 CFU/mL when using a ceramic membrane at 50°C and 100 kPa (Pafylias and others 1996). Retention of specific bacteria also depends on their morphology and cellular volume, which could be the cause of higher microbial retention in the latter study (Pafylias and others 1996, Gesan-Guiziou and other 2010). In another experiment carried out with skimmed colostrum, reductions in counts of more than 4.0 and more than 6.0 log10 cycles were observed with MF-1.4µm and MF-0.8µm

(tubular ceramic ISOFLUX® membrane, TMP= 0.7 bar, temperature 30 °C) processing systems, respectively (Gosch and others 2014). Higher levels of microbial reduction in this study may be due to the use of different types of membrane systems and higher levels of initial microbial load.

With regard to PEF/MF hurdle processing, the results obtained in the present study are in agreement with our earlier work (Walkling-Ribeiro and others 2011a), in which we observed reductions between 4.9 and 7.1 log10 CFU/mL in microbial populations of skim milk following treatment with PEF and MF through a ceramic membrane of pore-size 1.4µm. The higher microbial inactivation in PEF/MF hurdle processing may be due to the increased cellular volume of microorganisms, either by alteration in steric arrangement of cell by formation of pores, or by making clusters of microorganism with each other or with milk components during PEF processing, which ultimately increases the filtration efficiency (Walkling-Ribeiro and others

2011).

For physico-chemical properties, no significant differences were observed in pH, electrical conductivity, °Brix or viscosity for skim milk treated with either stand-alone PEF, MF-1.2µm and

MF-1.4 µm or their respective combinations, and HTST processing (P ≥0.05). These results are in agreement with Michalac and others (2003), who used PEF at 35 kV/cm for 188 µs and HTST

(73°C for 30 s) pasteurization to investigate the effects on the physical properties of raw skim milk. They did not find any differences in pH, electrical conductivity, total solids, or viscosity regardless of the treatments. Evrendilek and others (2004b) also did not find any difference in pH and °Brix (total soluble solids) values of a yoghurt-based drink when the drink was first heated at

60°C for 30 s and then cooled to 30°C before PEF treatment at 30 kV/cm electric field strength with 1.4 µs pulse width for 32 µs. Similarly, Floury and others (2006b) also found no change in pH following PEF treatment of milk at 45-55 kV/cm and 10 °C; however, the viscosity decreased with increase in PEF intensity. Change in viscosity was probably due to

“modifications of the hydrodynamic volume of casein micelles or of the mineral balance”

(Floury and others 2006b). The physico-chemical properties of skim milk processed with PEF vary depending on the processing conditions applied. Bermúdez-Aguirre and others (2011) studied the effect of PEF on skim milk at different intensities and temperatures. According to them, pH, electrical conductivity and solids-not-fat (SNF) were decreased at 30.76 kV/cm (40

°C), were increased at 38.46 kV/cm (40 °C) and again were decreased at 46.15 kV/cm (40 °C).

However, the changes in pH and SNF were significantly different, whereas electrical conductivity was unaffected for most of the PEF conditions applied. There were no trends in physico-chemical changes and this could be due to an arcing problem in the system. Values of pH and electrical conductivity of milk have been reported between 6.22 - 6.77 and 4.2 - 4.8 mS/cm, respectively (Neville and Jensen 1995). Moreover, no significant difference (P ≥0.05) in pH was found between raw milk, HTST (75 °C for 15 s) and PEF (35.5 kV/cm for 300 and 1000

µs) treated whole milk (Odriozola-Serrano and others 2006). In a similar study with skim milk,

Debon and others (2012) did not find any differences in pH for raw and MF-1.4µm (100-300 kPa at 45 °C) treated skim milk.

Significant reductions (P <0.05) in °Brix and viscosity were observed in the skim milk treated by MF-0.65µm and respective combinations with PEF. The decrease in °Brix and viscosity is due to trapping of milk components by smaller pore size MF (Sachdev and others 1997,

Vivekanand and others 2004). Pafylias and others (1996) have also recommended the use of MF-

1.4 µm in the dairy industry, which makes the right balance by removing the microorganisms and keeping the long term flux while maintaining the milk quality with negligible or no retention of milk components. In the present study, physico-chemical properties were not significantly affected (P ≥0.05) for skim milk processed through MF-1.2µm and MF-1.4 µm. The reason may be the difference in the membrane system, as described above.

Counts of all three pathogenic bacteria were reduced to undetectable levels (<10 CFU/mL) in the milk treated by PEF-H alone and by PEF-H/MF-1.2µm, while reductions in counts of the pathogens by PEF-L/MF-1.2µm were 1.93, 2.78 and 1.5 log10 cycles for Salmonella spp., E. coli and L. monocytogenes, respectively. Woulters and others (2001), also observed < 1 log10 cycle reduction of L. monocytogenes, when McIlvaine buffer (pH 7, conductivity 2.0 mS/cm) was treated at 22 kV/cm (pulse width = 2 µs, energy density = 2.2 kJ/kg), while the reduction in counts of Saccharomyces cerevisiae, E. coli, and Salmonella spp. were more than 3.0 log10 cycles. These results support the finding that small sized cells (such as L. monocytogenes) need higher electric field intensity to develop higher trans-membrane pressure to cause cell destruction, since they are more resistant than larger bacterial cells at a given electric field intensity (Rastogi 2003, Pagan and others 2005).

Higher microbial reduction at high intensity electric field achieved in the present study supports the earlier findings, which describe that increased electric field intensity and high temperature induced higher levels of microbial inactivation, causing irreversible membrane damage and decrease in electrical breakdown potential of the cell membrane by phase change of the cell membrane from gel into liquid crystalline structure, resulting in the leakage of cell contents into the environment leading to cell death (Otunola and others 2008, Stanley 1991).

Several researchers have investigated the inactivation effect of PEF treatments on pathogenic bacteria in milk. The findings obtained in the present challenge study are in line with the outcomes of other studies, in which the researchers observed a reduction of 5.0 log10 CFU/mL for both E. coli and L. monocytogenes, when the milk was processed under 30 kV/cm for 600 µs treatment time (Zhao and others 2013), a reduction of S. Dublin by 4.0 log10 CFU/mL by PEF at

36.7 kV/cm in homogenized and pasteurized milk (Sampedro and others 2005), and reduction by

~ 4.0 log10 CFU/mL of S. Dublin at electric field intensity of 30 kV/cm for 120 µs in skim milk

(Sensoy and others 1997).

Inactivation of E. coli and L. innocua (the latter used as a non-pathogenic surrogate for L. monocytogenes) at the level of 5 - 6 log10 CFU/mL was also reported in reconstituted whole milk after PEF treatments at 23 - 28 kV/cm (20 µs pulse width) with inlet temperature of milk at 55

°C (Sharma and others 2014). Similarly, a reduction in count of E. coli by >4.0 log10 CFU/mL has been reported in UHT milk processed by a static PEF system at 23 kV/cm with 20 pulses

(Grahl and Markl 1996). Dutreux and others (2000) reported a reduction of E. coli and L. innocua by 5.7 and 3.9 log10 CFU/mL, respectively, after PEF processing of fat free milk at

41kV/cm and 65 pulses. Effects of high processing temperatures on microbial inactivation by

PEF treatment have also been studied. Reina and others (1998) observed inactivation of L.

83 monocytogenes by 2.5 log10 CFU/mL after PEF processing at 30 kV/cm, 25 °C for 600 µs treatment time, and the inactivation level was increased to >4.0 log10 CFU/mL when the treatment temperature was increased to 50 °C at the same PEF operating conditions. Similarly,

Noci and others (2009) found a reduction of 3.3 log10 CFU/mL of L. innocua in low fat UHT milk processed by PEF at 40 kV/cm for 50 µs. MF has been also used to remove pathogenic bacteria. Saboya and Maubois (2000) reviewed the use of MF in the dairy industry and cited that a 3.4 and 3.5 log10 reduction in count of L. monocytogenes and S. Typhimurium, respectively, was achieved by MF. Gosch and others (2014) compared the effectiveness of MF-1.4µm and MF-

® 0.8µm (tubular ceramic ISOFLUX membrane, TMP= 0.7 bar, temperature 30 °C) to remove the pathogenic bacteria in skimmed colostrum. They did not find any significant difference in removal of pathogenic bacteria by both of the MFs, with levels of both L. monocytogenes and E. coli being reduced by >3.0 log10 cycles. These results support the findings of Rosenberg (1995), who recommended the use of other efficient antimicrobial technology to combine with MF in milk processing, as membrane microfiltration efficiently removes bacteria and spores, but it does not ensure the removal of all pathogenic bacteria.

The success of PEF technology to pasteurize food largely depends on its ability to inactivate different types of microorganisms, including spoilage and pathogenic bacteria. All these different rates of inactivation may be due to diverse operating conditions and determined by varying efficacies of the PEF system designs. Wouters and others (2001) have summarized the different factors affecting PEF efficiency as: treatment parameters, product parameters and microbial parameters. Electric-field strength, PEF treatment time and temperature, number of pulses, pulse width and pulse shape affect the microbial inactivation rate (Evrendilek and Zhang 2005;

Wouters and others 2001). A square wave pulse is more efficient than exponential pulse

(Pothakamury and others 1996) and bipolar pulses are more lethal than mono polar pulses (Ho and others 1995). Regarding the chamber configurations, a continuous treatment chamber is more efficient compared to static chambers for microbial inactivation (Martin and others 1997).

Microbial cell size, species, strain, Gram type, and growth stage have all been found to affect the inactivation of microorganisms in food by PEF treatment (Huang and Wang 2009). Larger size microorganisms are more sensitive to electric field strength than smaller cells (Pagan and others

2005) and Gram positive bacteria are more sensitive to PEF treatment than Gram-negative bacteria (Shamsi and others 2008). Fairly large discrepancies between different PEF studies on milk are indicated. In a review by Sampedro and others (2005) inactivation of E. coli in milk by

8 log10 CFU/mL is reported following PEF treatment with T=50 °C, E=60 kV/cm, whereas

Floury and others (2006a) reported the inactivation of S. Enteritidis was only 2.3 log10 CFU/mL with an inlet temperature of skim milk at 62 °C, and PEF at 47.0 kV/cm. It is unlikely that the use of different microbial strains/species in both studies account for such a pronounced difference in the results but in combination with the slight difference in temperature, application of different PEF systems (60 and 47 kV/cm) and differences in the processing conditions, the divergence in treatment efficacy may be explained. The latter also emphasizes the need for a harmonization principle with regard to reporting scientific findings and processing parameters for the comparison of technologies used in other studies. Finally, the combined application of

PEF under different process parameters with different MF processing equipment (lab scale MF-

1.2µm and pilot scale MF-1.4µm) used in the present study showed a substantial amount of microbial reduction compared with HTST processing without any significant effect on processed milk quality compared to raw milk, indicating the suitability of hurdle processing combining

PEF and MF as an alternative to HTST processing.

3.6 Conclusions

When used alone, high intensity PEF (PEF-H), MF-0.65µm and MF-1.4µm resulted in the highest reduction in total microbial load in milk; whereas when the treatments were applied in combination, PEF-H/MF-0.65µm and PEF-H/1.4µm were most efficient in reducing the total microbial load and the microbial reduction by this combination was higher than that obtained using HTST processing. The reduction in microbial counts of milk obtained by PEF-H/MF-

1.2µm and PEF-M/MF-1.4µm was comparable with HTST processing. Changes in physico- chemical properties of milk (pH, electrical conductivity, °Brix, and viscosity) were not observed when the skim milk was processed by PEF and MF (MF-1.2μm and MF-1.4μm) alone and in combination. However, microfiltration through a small pore-sized membrane (MF-0.65μm) resulted in a decrease in °Brix and in viscosity of skim milk regardless of whether it was applied alone or in combination with PEF, indicating that the membrane is not suitable for milk processing. Significant inactivation (> 4 log10 CFU/mL) of all three pathogenic bacteria (E. coli,

Listeria monocytogenes or Salmonella spp.) used in this study was observed following treatment of milk by PEF-H and PEF-H/MF-1.2µm. The outcomes of this study indicate potential for the use of PEF and MF as a hurdle approach in the dairy industry to produce skim milk of high quality.

CHAPTER 4: Color and flavor of Brie-type cheese made from high voltage pulsed and membrane microfiltered skim milk subjected to pathogen screening

4.1 Abstract

Raw milk cheese is popular due to its different flavour profile than pasteurized milk cheese.

Brie cheese is a white mould surface ripened soft cheese that may pose a food safety risk due to its high water activity and pH. Pulsed electric fields (PEF) and membrane microfiltration (MF) were used to process milk prior to manufacture of Brie and the microbiological profile and sensory characteristics were compared to Brie made from raw milk. Skim milk was treated with

PEF at 40 kV/cm, 60-65 °C for an energy density up to 198 kJ/L and a treatment time of 1571µs alone and in combination with MF using a 1.4 μm pore sized membrane. Milk was standardized with HTST (85 °C for 15 s) treated cream to protein and fat ratio of 0.86 before cheese making and prepared cheeses were ripened for 65 days at 4°C before analysis. Proximate analysis of cheese did not reveal significant changes (P ≥0.05) in the compositional values of cheese manufactured from milk that had undergone the different treatments, and colour analysis showed very slight change in color between cheeses made from treated and raw milk. Seventeen volatile flavour compounds were analysed and only acetic acid concentration was decreased in PEF and

PEF/MF treated milk cheese compared to raw milk cheese. Escherichia coli O157:H7,

Salmonella spp. and Listeria monocytogenes were not detected in any of the cheeses following ripening for 65 days at 4 °C. These findings indicate that combined application of PEF and MF to milk as a hurdle technology could be a non-thermal alternative for the safe manufacture of cheese.

4.2 Introduction

Raw milk cheese is popular among consumers due to its unique flavour, sensory characteristics and texture, compared to pasteurized milk cheese (Yu and others 2012).

Camembert and Brie are popular surface mould-ripened, soft cheeses and are produced in large quantities annually (Marcos and others 1990). Brie is a soft cheese made from raw cow milk

(Sanaa and others 2004) and is named after Brie, the French province from which it originated.

In Brie cheese, mould growth is the most important factor for the production of volatile compounds and texture modification during the ripening period (Karahadian and others 1985).

Ripening is found to be faster in cheese made from raw milk compared to heat pasteurized milk cheese (Gaya and others 1990).

However, consumption of raw milk cheese may pose a risk to public health and its consumption has been associated with foodborne illness (Fox 2000). Little and Knochel (1994) have reported several outbreaks caused by Salmonella spp., Listeria spp., Campylobacter spp.,

Staphylococcus aureus and Escherichia coli O157:H7 associated with the consumption of raw milk and the cheese made from it. Soft cheeses made from raw milk pose a particular risk, and several outbreaks of foodborne illness have been recorded in the USA in the past 20 years, resulting from consumption of raw milk cheese (Brooks and others 2012). For assurance of microbial safety, in Canada (except in Quebec), cheese made from unpasteurized milk must be

“held at a temperature of 2 °C or more for a period of 60 days or more from the date of the beginning of the manufacturing process” (Government of Canada, 2014).

However, studies have shown that pathogens can survive in cheese after 60 days of ripening, including E. coli O157:H7 in Cheddar cheese ripened at 6-7 °C (Reitsma and Henning 1996),

Salmonella Typhimurium in Cheddar cheese at 7 °C, and Colby cheese ripened at 6.1-8.9 °C

(Park and others 1970), and Listeria monocytogenes in Cheddar cheese ripened at 6 °C (Ryser and Marth 1987). Ramsaran and others (1998) used raw milk inoculated with E. coli O157:H7 and L. monocytogenes (at 104 CFU/mL) to prepare Camembert cheese. The analysis of cheeses after 65 days ripening, stored at 2 ± 1 °C, showed the presence of both pathogenic bacteria at levels higher than the initial inoculum. These results indicate that raw milk cheese ripened for 65 days at low temperature may also contain pathogenic microorganisms. In the USA, during 1993-

2006, 56 fluid milk associated foodborne disease outbreaks were recorded and of the 56 cases,

82% were associated with non-pasteurized milk, whereas, out of the 65 cheese associated foodborne disease outbreaks recorded in same the period, 27 (42%) were from unpasteurized milk cheese (Langer and others 2012). These data show that consumption of raw milk and cheese made from raw milk can pose a risk to public health. Therefore, there is a need for an alternative technology for treatment of milk used for cheese making to ensure food safety as well as to maintain the desired sensory characteristics of the product.

Heat pasteurization is commonly used to destroy pathogenic microorganisms present in milk prior to cheese making (Awad and others 2010). However, it may also adversely affect the biochemical composition, flavor, colour, taste and nutrients of the product, which could affect the quality of cheese (Chambers and others 2010; Sepulveda-Ahumada and others 2000;

Wirjantoro and Lewis 1996). The colour of milk is governed by its components: carotenoids, riboflavin, protein and fat concentration. Lightness (L) of milk is the result of reflection of light from casein micelle and fat globules. Concentration of riboflavin and carotenoids in milk determines the intensity of the red (a) and yellow (b) colours of milk (Noziere and others 2006).

The sensory quality of cheese is a result of the raw material quality and the manufacturing

89 process (Buchin and others 1998). Heat pasteurization is also found to destroy some species and strains of indigenous microflora of milk, which are responsible for production of the flavor of cheeses made with raw milk (Grappin and Beuvier 1997) as well as causing the denaturation of serum proteins and the modification of rennetability of milk (Grappin and Beuvier 1997). Heat treatment of milk also brings about changes in stability of whey protein, pH and calcium distribution in milk and these changes play a crucial role in cheese making (Garcia-Amezquita and others 2013). In the case of cheese, flavour is the important quality attribute, which determines the consumer’s perception and acceptance of the product (Young and others 2004).

Consumers’ interest in minimally processed and “fresh-like” foods with natural taste and extended shelf life has triggered food scientists and technologists to look for alternative technologies for food processing other than thermal treatments to avoid the adverse effects of heat, whilst still ensuring food safety (Fernandez-Molina and others 2005; Pereda and others

2009). Various non-thermal technologies have been proposed as alternatives to thermal technology to improve food quality (Barbosa-Cánovas and others 1999; Bermudez-Aguirre and others 2009).

Pulsed electric field (PEF) is an emerging non-thermal technology, which can effectively be used to process liquid food. Non-thermal technologies destroy microorganisms using low energy intensity treatments at ambient temperature with reduced time for processing, while retaining thermally sensitive food components (Barbosa-Cánovas and others 1999). During PEF food is placed between two electrodes under the influence of a high electric field and then very short, high voltage pulses are applied, which brings about the destruction of microbial cells

(Pothakamury and others 1997; Toepfl and others 2007). PEF has shown some promising effects

90 in liquid food processing by reducing/inactivating total microbial counts (Rodríguez-González and others 2011; Walkling-Ribeiro and others 2011), L. monocytogenes (Reina and others 1998),

Salmonella spp. (Floury and others 2006; Sensoy and others 1997), E. coli O157:H7 (Dutreux and others 2000, Evrendilek and Zhang 2005), and S. aureus (Sobrino-Lopez and others 2006).

Sale and Hamilton (1967) are among the first researchers to use PEF for microbial inactivation. They applied electric field intensity up to 25 kV/cm (10 pulses of 20 µs) and successfully demonstrated the lethal effect of electric field to inactivate yeasts and bacteria, including Saccharomyces cerevisiae, Candida utilis, E. coli, Clostridium welchii Micrococcus lysodeikticus, Bacillus subtilis, Bacillus cereus, and Bacillus megaterium.

Sepulveda-Ahumada and others (2000) prepared Cheddar cheeses from PEF treated milk and suggested PEF technology as an alternative to heat treatment to improve cheese quality. Yu and others (2009) also investigated the effectiveness of PEF treatment of milk for cheese manufacture and suggested a possible option for cheese making with PEF treated milk for better texture and sensory aspects.

Membrane microfiltration (MF) is another non-thermal technology, which has been in use in the food industry for a long time for separation processes. MF is a pressure-driven physical separation process, which uses membranes in the range of 0.10 - 5 μm pore-size for separation of particles, somatic cells, bacteria and spores from liquid foods (Huisman 2000; Ripperger and

Altmann 2002) without affecting their quality (Olesen and Jensen 1989). Maubois (2002) and

Saboya and Maubois (2000) have also reported that MF treatments can be used to remove somatic cells, microorganisms and microbial spores from milk. Membrane microfiltration does

91 not ensure the removal of all pathogenic bacteria present in milk, so MF needs to be combined with other processing technologies (Rosenberg 1995).

Therefore, combined application of non-thermal technologies, such as PEF and MF for the processing of milk to be used in soft cheese making could be an alternative to replace heat treatment or the use of raw milk. These technologies may help to ensure food safety by minimizing microbial load, and assist in the production of cheeses with typical flavour and taste similar to those associated with raw milk cheeses. To our knowledge, no detailed scientific study has been undertaken for the evaluation of Brie type soft cheese made from PEF and MF treated milk. This study was undertaken, to investigate the effects of stand-alone PEF and PEF in combination with MF treatment on Brie cheese preparation, to compare the quality of cheese with raw milk cheese and to screen the product for the presence of pathogenic bacteria.

4.3 Materials and methods

4.3.1 Skim milk preparation

Whole milk was obtained from a local dairy (Gay Lea foods Co-operatives, Guelph, ON,

Canada) and the collected milk was filtered to remove suspended particles, and then warmed in a water bath (Isotemp 210, Thermo Fisher Scientific Inc., Newlington, NH, USA) to 35 °C. The warmed milk was skimmed using a cream separator (STsM-100-18, Motor Sich JSC,

Zaporozhye, Ukraine) to obtain a final fat content of 0.05-0.1%.

4.3.2 PEF treatment

Fresh skim milk was processed by PEF in a chamber designed at the University of Guelph using a pulse generator (PPS 30, University of Waterloo, Waterloo, ON, Canada) to generate an exponential decay pulse of 1.5 µs pulse width. The skim milk was pumped at a flow rate of 25 mL/min to the PEF treatment chamber and treated at 40 kV/cm at 60-65 °C for an energy density up to 198 kJ/L for a treatment time of 1571 μs. Raw and PEF treated skim milks were further cooled and stored at 4 °C until used. Further details of the equipment are given by Walkling-

Ribeiro and others (2011).

4.3.3 Microfiltration

The PEF-treated skim milk was further processed with pilot-scale cross-flow microfiltration

(CFMF) system (MFS-1, Tetra Pak Filtration Systems, Aarhus, Denmark) equipped with a 1.4

µm pore-size ceramic membrane (1P19-40 SCT Membralox®, Tetra Pak Filtration Systems,

Aarhus, Denmark). Milk was warmed to 35 °C before MF processing and treated samples were further cooled and stored at 4 °C until used. Details of the equipment and processes are further described by Walkling-Ribeiro and others (2011).

4.3.4 Brie cheese preparation

Three replicates of Brie cheese were prepared from three types of milk: raw skim milk, PEF treated skim milk and PEF/MF treated skim milk. The following cheese making procedure was applied for each type of treated milk. Raw or treated skim milk was combined with pasteurized cream (80 °C/15 s) to produce 6.5 kg of milk with a protein: fat ratio of 0.86. Starter culture (45 mg/kg; Choozit, MA 19, Danisco France SAS, Sassenage, France) and mould (5.5 mg/kg;

Choozit, PC HP 6, Danisco France SAS, Sassenage, France) were added after the temperature of the milk reached 32 °C. The inoculated milk was held for 1 h with frequent stirring while keeping the temperature constant at 32 °C. Calcium chloride solution and liquid rennet (Chris-

Hansen, No. 3103600, USA) were added both at the rate of 77 mg/kg of milk. The milk was left to coagulate for 45 min after mixing. The set milk was cut into small pieces and curd was allowed to settle for 1 h. The whey was drained before pouring the curd into cylindrical Brie molds (8.5 × 10.5 cm), which were turned every 20 minutes for the first hour, then every hour for the next 4 hours. The cheese was turned very early the next morning and, after being removed from the molds and weighed, dry salt (3% of cheese weight) was rubbed on all sides of the cheese. The salted cheese was placed on a plastic mat in a plastic tub and stored at 12 °C and

85% relative humidity (RH) for 24 h, with the lid slightly open. The cheese was stored at the same temperature for a further 14 days with a raised relative humidity at 95%, being turned over every day. The cheese was then packed in waxed paper and stored at 4 °C until analysis (65 days).

4.3.5 Analysis of color attributes

Color parameters of lightness (L), redness (a) and yellowness (b) of the Hunter Lab color space were analysed using a CM 3500-d spectrophotometer (Konica Minolta Sensing Inc.,

Mahwah, NJ, USA) equipped with Spectra Magic NX CM-S 100 software. The L value is an indicator of luminosity, and a low number (0-50) indicates darkness and a high number (51-100) indicates lightness. Negative to positive values of a and b indicate the green–red and blue–yellow components, respectively. Reflectance measurements were acquired over a wavelength range from 400 nm to 700 nm. The Hunter Lab color of all cheeses made from milks that had

94 undergone the three different treatments were analysed in triplicate and colour differences (∆E), in comparison to raw milk Brie cheese was calculated according to following equation:

where, ∆ = difference in values between raw milk cheese and treated milk cheese

Cserhalmi and others (2006) have described the color difference (∆E) between treated and untreated samples as follows: not noticeable (0 - 0.5), slightly noticeable (0.5 - 1.5), noticeable

(1.5 - 3.0), well visible (3.0 - 6.0) and great (6.0 - 12.0). The colour of the cheese was directly measured at points along the cross-sectional surface of the cheese after cutting it in half.

4.3.6 Analysis of volatile compounds

The selected ion flow tube mass spectrometer (SIFT-MS; Voice200®, Syft Technologies

Ltd., Christchurch, New Zealand) was used to measure headspace volatile flavour compounds in the cheese. This is a direct mass spectrometer, which applies chemical ionization for the analysis of volatile flavour compounds, yielding product ion to identify and quantify the volatile compounds by comparison with a built-in database. The cheese (10 g) was placed into a 300 mL glass jar, and sealed with a septum. Another jar filled with air was used as a blank. A full mass

+ + + scan was performed with H3O , NO , or O2 as reagent ions, scan range 15-200 m/z, scan step

1.0 m/z, time limit 100 ms, carrier gas helium pressure 250 kPa, arm temperature 120 °C, and

ﬂow tube pressure 122.0 ± 1.3 mTorr. Volatile flavour compounds were quantified using selected ion mode with time limit 100 ms, count limit 10,000, ambient background time 30 s, and scan time 60 s.

4.3.7 Proximate analysis of cheese

The moisture, protein, fat, salt and total ash contents were determined after 65 days of ripening for all cheeses. The moisture content was measured by desiccation in a hot air oven (Fisher

Scientific, Isotemp oven, 903IN0042, USA) at 105 °C for 48 h and dried samples were placed in a desiccator to cool down to the ambient temperature and moisture content was determined after weighing the dried samples and calculated by percentage weight loss before and after drying. Fat content was determined using the Babcock method 15.8A as described by Marshall (1992).

Protein content in cheese was measured (Simonne and others 1997) by the Dumas method (Leco

Corp., FP528, St. Joseph, MI, USA). Salt content in cheese was measured as described by

Marshall (1992). Total ash content was determined by drying to constant weight in a muffle furnace (Fisher Scientific, 800 Series, Isotemp oven, 903IN0036, USA) at 550 °C.

4.3.8 Listeria monocytogenes, Salmonella spp., and E. coli screening

All cheeses stored at 4 °C for 65 days were tested for the presence of L. monocytogenes,

Salmonella spp., and E. coli. Cheese (10 g) was weighed in a stomacher bag, diluted with 90 mL of sterile Ringers solution (BR0052G, Oxoid Ltd., Basingstoke, Hampshire, UK) and stomached

(Seward, 400, 110 V, England) for 2 min. The prepared suspension of cheese was serially diluted with Ringers solution and was surface plated onto the appropriate selective agar medium:

Listeria monocytogenes blood agar (prepared as described by Johansson (1998)), ChromAgar

O157 (Becton, Dickinson and Company, Sparks, MD, USA) and BrillianceTM Salmonella Agar

Base (Oxoid-CM1092, Oxoid Ltd, Basinngstoke, Hants, England) for L. monocytogenes, E. coli and Salmonella spp., respectively. The plates were incubated at 37 °C for 24 h.

4.3.9 Statistical analysis

The proximate composition, color, flavor, and microbial screening data obtained in the present study were analysed statistically using SPSS version 21.0 (IBM Corporation, Chicago,

USA). Three different product batches were processed (n = 3) and analyzed by analysis of variance (ANOVA) at a confidence interval of 95%. Upon occurrence of statistical differences, treatment means were compared with the Tukey’s test for multiple pair wise comparisons. Each treatment was carried out in triplicate and analysis was performed in duplicate.

4.4 Results and discussion

4.4.1 Physico-chemical quality of cheese

4.4.1.1 Proximate analysis of cheese

The proximate composition of raw milk and treated milk cheeses are shown in Table 4.1.

The moisture content of raw milk cheese was 50.49% whereas those of cheeses produced from

PEF and PEF/MF-1.4µm treated milk were found to be slightly higher (52.54 and 51.88%, respectively). However, these differences were not statistically significant (P >0.05). Similarly, protein content of cheese made from PEF and PEF/MF-1.4µm treated milk was 17.87 and 17.91% respectively, which were slightly but not significantly lower than that of raw milk cheese

(18.59%). The salt and fat contents in cheese were similar for all cheeses (P >0.05). Total ash content of Brie made from PEF and PEF/MF-1.4µm treated milk was 2.04 and 2.28%, respectively, while that of raw milk cheese was 2.12%. Again, these differences were not statistically significant (P >0.05). In conclusion, neither PEF nor PEF/MF-1.4µm treatment of

97 milk used for cheese manufacture significantly changed its composition compared to raw milk cheese.

Since PEF treatment of milk does not result in changes in composition (Bendicho and others

2002), a change in proximate composition of cheeses made from raw and PEF treated milk is unlikely. In a review paper, Sharma and others (2014) have described that PEF treatment at 25-

28 kV/cm at a temperature >45°C did not affect the total solids, pH, and yield of cottage cheeses prepared from PEF treated milk.

Table 4.1. Proximate analysis of Brie cheese prepared from raw milk, milk treated with pulsed electric field (PEF; 40 kV/cm, 1571 µs) or a combination of PEF and cross-flow microfiltration (CFMF1.4 = 1.4 µm) and ripened for 65 days (Data = mean ± S.D.; n =3).

Treatment Moisture1 Protein1 Fat1 Salt1 Total Ash1 [%] [%] [%] [%] (%) Raw milk 50.49 (±0.81)a 18.59 (±1.08)a 26.83 (±0.41)a 2.48 (±0.10)a 2.12 (±0.31)a PEF 52.54 (±1.07)a 17.87 (±0.73)a 26.51 (±0.32)a 2.60 (±0.11)a 2.04 (±0.39)a

a a a a a PEF/MF-1.4µm 51.88 (±0.59) 17.91 (±1.59) 26.42 (±0.58) 2.57 (±0.15) 2.28 (±0.16) P2 NS NS NS NS NS SEM3 0.509 0.384 0.102 0.042 0.107 a Same superscripted letters in the same column indicate no statistical significance between the two means. 1Values in parentheses following the mean values of proximate analysis indicate the standard deviations. 2P stands for the statistical probability; NS indicates no statistical significance (P > 0.05). 3SEM abbreviates the standard error of the mean.

Michalac and others (2003) did not observe a change in the physical properties of milk, including total solids, protein, pH, electrical conductivity and viscosity when the milk was treated by PEF at 34.4 kV/cm with outlet temperature of 52 °C. A similar study in which milk

(2% fat) was subjected to PEF at 40 kV/cm, (pulse duration of 2 μs for 20 pulse number and maximum treatment temperature of 50 °C), showed that the treatment did not significantly affect

98 its physical and chemical properties. Moreover, sensory evaluation of heat pasteurized and PEF pasteurized milk did not identify any significant differences (Qin and others 1995).

Awad and others (2010) compared Domiati cheese (a variety of Egyptian cheese) made from

MF treated milk to that produced from milk that had been heat pasteurization and did not find any significant difference in moisture, fat and salt content after 60 days ripening. Similar results have been reported for Cheddar cheese prepared from raw milk, heat pasteurized milk and MF

(1.4 µm) treated milk (McSweeney and others 1993). Sensory results showed pasteurised and

MF cheeses to be of high and similar sensory quality in comparison with raw milk Cheddar cheese. Amornkul and Henning (2007) also did not find a significant difference in moisture content, fat, protein and ash content in Cheddar cheese produced from raw and MF (1.4µm) treated milk, however, salt content in MF treated cheese was found to be slightly higher than raw milk cheese. Yu and others (2012) conducted a proteolytic study of milk treated with PEF and recommended the PEF treatment as a better process for cheese making, compared to raw milk cheese making, due to low concentration of free amino acids, which is negatively correlated with quality of cheese. The extensive proteolysis during aging of ripened cheeses triggers the release of more amino acids, which ultimately increases the pH value and favors the growth of

Clostridium tyrobutyricum causing late blowing defect in cheese (Ledenbach and others 2009).

Dunn (1996) applied PEF (intensity = 20-80 kV/cm; pulse width = 1-10 μs; T = 55 °C) treatment to milk to investigate its effect on enzyme activity, fat integrity, starter growth, rennet clotting yield, cheese production, calcium distribution and flavour degradation. No significant differences were found for physical or chemical properties, rather, less flavour destruction was observed in PEF treated milk. They proposed that PEF treatment of milk for manufacturing of

99 cheese had potential. These results conclude that PEF or PEF/MF treatment does not cause significant change in milk composition and the treated milk can be used to prepare cheese of high quality.

4.4.1.2 Color attributes

The Hunter color attributes of lightness (L), redness (a) and yellowness (b) of Brie cheeses produced from raw milk or treated milk are shown in Table 4.2. Lightness index (L) of raw milk cheese was 67.04 and that of PEF, and PEF/MF-1.4µm treated milk cheese were not significantly different; being 66.94 and 67.10, respectively. The value of yellowness index (b) also showed the same trend and was 11.42, 11.16 and 10.92 for cheese made from raw, PEF treated and PEF/MF-

1.4µm treated milks, respectively. However these were not significantly different (P >0.05).

Finally, the value for redness (a) increased significantly (P <0.05), from its value for raw milk cheese (-0.19) to that for PEF (-0.13) and PEF/MF-1.4µm treated (-0.12) milk cheese.

Determination of total color difference (∆E), however, showed a very small difference of 1.14 and 1.18 for stand-alone PEF and PEF/MF-1.4µm treated milk cheese, respectively, to that of raw milk cheese.

Bermúdez-Aguirre and others (2011) also observed a significant change in a attribute of skim milk treated at a PEF intensity of 30.76 - 46.15 kV/cm. However, no trends in the values were observed and they suggested that these random changes in colour attribute were due to wearing down of the electrode due to arcing. The study was conducted using different PEF intensities at different temperatures (20-40 °C) and most of the treatments did not result in a significant change in L and b attributes even at high electric field intensity and temperature.

Sampedro and others (2005) also did not observe significant changes in color, pH and protein

100 content in skim milk when it was treated with PEF at electric field intensity of 35 kV/cm (pulse width =3 μs) for 90 μs treatment time.

Evrendilek and others (2004) did not observe significant changes in three color attributes (L, a and b) of a PEF (30 kV/cm with 1.4 µs pulse width for 32 µs) treated yoghurt-based drink stored at 4 °C for 91 days. The sample was first heated at 60 °C for 30 s and then cooled to 30 °C before PEF treatment. Pafylias and others (1996) have also recommended the use of MF-1.4 µm in the dairy industry to obtain the right balance between reduction in microbial load and milk quality.

Table 4. 2. Comparison of Hunter colour space attributes L (Lightness), a (redness) and b (yellowness) measured in Brie cheese prepared from raw milk, pulsed electric field (PEF; 40 kV/cm, 1571 µs) and combination of PEF and pilot scale cross-flow microfiltration (CFMF1.4 = 1.4 µm) treated milk, after 65 days of ripening period (Data = mean ± S.D.; n = 3).

Treatments L1 a1 b1 ∆E1 Raw milk 67.04 (±1.02)a -0.19 (±0.01) a 11.42 (±0.43) a PEF 66.94(±0.77) a -0.13(±0.02) b 11.16(±0.21) a 1.14(±0.29) a PEF/MF 67.10(±0.87) a -0.12(±0.01) b 10.92(±0.27) a 1.18(±0.27) a P2 NS ** NS NS SEM3 0.259 0.012 0.117 0.102 (L* = lightness, a* = redness, b* = yellowness) a Same superscripted letters in the same column indicate no statistical significance between the two means. 1Values in parentheses indicate the standard deviations. 2P stands for the statistical probability; NS indicates no statistical significance (P ≥ 0.05). 3SEM abbreviates the standard error of the mean.

4.4.1.3 Volatile compounds

The concentration of different volatile flavour compounds of raw milk, PEF and PEF/MF-

1.4µm treated milk cheeses were measured by SIFT-MS and results are shown in Table 4.3. The

101 volatile flavour compounds in cheese are produced by breakdown of milk components such as protein, fat, lactose and citrate by microbial and biochemical activities during the preparation and ripening process (Pillonel and others 2003). Formation of aldehydes, alcohols, acids and sulphur compounds are the result of degradation of amino acids, whereas esters, methyl ketones and secondary alcohols are formed from lipolysis of fatty acids (González-Martín and others 2014).

Out of 17 volatile flavour compounds measured, only acetic acid concentrations were significantly different (P <0.05) between cheeses made from PEF and PEF/MF-1.4µm treated milk compared to cheese made from raw milk.

Aldehydes are one of the most important volatile compounds in cheese. Aliphatic aldehydes in milk are a product of auto-oxidation of unsaturated free fatty acids (FFA) (Zabbia and others

2012). They are also produced by breakdown of amino acids (Calvo and de la Hoz 1992).

Benzaldehyde, decanal and hexanal were quantified in all cheese, and the concentration of benzaldehyde was found to be slightly higher in raw milk cheese than in cheese made from PEF or PEF-MF-1.4µm treated milk. However, the concentration of all three aldehydes did not change significantly between the cheeses (P >0.05). Benzaldehyde is responsible for the aromatic note of bitter almond in cheese and is a result of breakdown of tryptophan or phenylalanine amino acids, and it is also formed during oxidation of phenylacetic acid (Evert-Arriagada and others 2012).

Hexanal is produced by β-oxidation of unsaturated fatty acids and gives cheese its green-grass like aroma (Moio and others 1993).

Several factors including animal metabolism and feed, chemical, microbial, and enzymatic reactions, and the types of milk processing have an effect on volatile favour compounds of milk

(Calvo and de la Hoz 1992). Our results are in agreement with Zhang and others (2011), who did

102 not observe any change in hexanal and decanal in PEF treated milk. Though, hexanal concentrations were slightly decreased when low intensity PEF treatments were applied to milk, but when the treatment intensity was high (30 kV/cm) no significant change was observed.

Some ketones are passed into the milk through feed, e.g., acetone and 2-butanone (Gordon and Morgan 1972). They are also formed from FFA, released from triglycerides. Conversion of

FFA to keto-acids takes place through β-oxidation and then decarboxylated to methyl ketones.

Formation of methyl ketones in mould ripened cheese is also affected by concentration of precursors, stage of mould mycelial growth, and the cultural environment (Karahadian and others

1985). This group of compounds is responsible for fruity overtones in cheese (Bugaud and others

2001). In the category of ketones, 2-decanone and 2-nonanone were analyzed and their concentrations remain unchanged (P >0.05) in cheeses made from treated milk compared to values found for raw milk cheese. In a study of the effect of PEF (15-30 kV/cm for 800 µs) treatment on volatile flavour compounds in milk, Zhang and others (2011) observed no change in methyl ketone concentrations following treatment. These findings indicate that PEF processing does not affect the ketone groups present in skim milk, probably due to the reduced levels of fatty acids present in skim milk.

Alcoholic compounds are formed in the milk by reduction of aldehydes (Zhang and others

2011). Urbach (1990) observed a reduction of volatile carbonyls due to their conversion into their respective alcohols during refrigerated storage of raw cow milk. Secondary alcohols contribute to heavier cheesy notes in mold surface-ripened cheese (Karahadian and others 1985).

Different kinds of alcohol in cheese are related to dry or dusty odors, sweet and floral aroma

(Arora and others 1995).

103

Four types of alcohols, including 1-hexanol, 2-heptanol, 2-pentanol, 2-phenylethanol were detected in raw milk cheese as well as in cheeses made from PEF and PEF/MF-1.4µm treated milk, but the treatments did not cause a significant change (P >0.05) in alcohol concentrations in cheeses. Zhang and others (2011) showed similar effects of PEF on the alcohol concentrations in milk. They did not observe significant changes in alcohol concentrations between PEF (15-30 kV/cm for 800 µs at temperature <40 °C) and HTST (75 °C for 15 s) treated milk. Moio and others (1993) have described 2-heptanol as an important odorant in cheese.

Concentrations of short chain fatty acids are always high in milk and some of them are synthesized in the mammary gland and others are produced by degradation of triglyceride by lipase, which results in a rancid flavor (Zhang and others 2011). Long chain fatty acids are transferred from food to the mammary gland (Lin and others 1996). Four organic acids namely butanoic acid, hexanoic acid, propanoic acid, and acetic acid were quantified in cheese. Among these, the first three organic acids are produced by lipolysis, and acetic acid is synthesized from casein, lactose and citrate by microbial activity during cheese ripening (Urbach 1995). The concentration of acetic acid was found to be significantly lower (P <0.05) in cheeses produced from milk processed using both PEF and PEF in combination with MF in comparison with raw milk cheese. The concentration of acetic acid was reduced to 45.00 and 45.45 ppb, in cheese made from PEF and PEF/MF-1.4µm treated milk, respectively, compared to a value of 96.25 ppb, in raw milk cheese. Good flavour is negatively correlated with acetic acid, which is synthesized from casein (Urbach 1995), indicating that cheese made from PEF treated milk may have a more pleasant flavour compared to raw milk cheese. Since, biosynthesis of acetic acid takes place from casein (Urbach 1995), thus, a structural change in protein during PEF processing (Sharma 2014), may contribute to the change in acetic acid concentration of PEF treated milk cheese. However,

104

Zhang and others (2011) did not find significant changes in concentrations of acetic acid, butanoic acid and hexanoic acid in milk following PEF (30 kV/cm) treatment.

Two aromatic hydrocarbons (styrene and phenol) were also detected in both the raw milk and treated milk cheeses, however their concentrations were not significantly affected (P >0.05) by PEF or PEF/MF-1.4µm treatment. Phenol contributes to the pronounced flavour of Brie and

Camembert (Karahadian and others 1985), while styrene has a very strong plastic odour

(Molimard and Spinnler 1996). Dimethyl trisulfide and ethyl octanoate were the other volatile compounds detected in cheese and their levels were not significantly different (P >0.05) between all cheeses. Sepulveda-Ahumada and others (2000) prepared cheddar cheese from milk subjected to PEF (35 kV/cm at <35°C) treatment and recommended the use of PEF pasteurized milk for cheese making to improve cheese quality.

Table 4.3. Retention of volatile flavour compounds in Brie cheese prepared from raw milk, pulsed electric field (PEF; 40 Kv/cm, 1571 µs) and a combination of PEF and cross-flow microfiltration

(CFMF1.4 = 1.4 µm) treated milk, after 65 days of ripening (Data = mean ± S.D.; n =3).

Volatile Raw milk PEF PEF/MF P2 SEM3 compounds (ppb) cheese1 cheese1 cheese1 Aldehydes Benzaldehyde 26.65(±9.05)a 22.25(±7.19)a 18.48(±6.55)a NS 5.317 Decanal 8.53(±2.24)a 7.07(3.02)a 6.13(±2.95)a NS 0.780 Hexanal 10.90(±4.03)a 10.90(2.12)a 13.45(±10.65)a NS 2.132 Ketones 2-decanone 12.95 (±4.21)a 10.78(±4.08)a 10.58(±1.38)a NS 0.964 2-nonanone 15.63(±8.49)a 13.23(±3.47)a 16.18(±3.74)a NS 1.543 Alcohols 1-hexanol 1.78(±1.48)a 1.5 (±0.86)a 0.90(±0.71)a NS 0.300 2-heptanol 10.61(±5.38)a 11.48(±6.80)a 16.89(±6.96)a NS 2.087

105

2-pentanol 54.85(±8.56)a 52.65(±4.42)a 40.5(±17.84)a NS 4.071 2-phenylethanol 13.02(±6.62)a 13.83(±3.33)a 12.2(±1.41)a NS 1.833 Acids Acetic acid 96.25(±7.99)a 45.00(±9.34)b 42.45(±11.63)b * 7.571 Butanoic acid 54.43(±17.86)a 40.83(7.68)a 37.85(±1.77)a NS 4.294 Hexanoic acid 5.45(±1.66)a 8.25(±0.21)a 4.65(±2.19)a NS 0.633 Propanoic acid 10.74(±3.67)a 5.47(±4.50)a 4.54(±2.78)a NS 1.272 Other compounds Phenol 3.51(±1.58)a 1.53(±0.81)a 1.44(±0.81)a NS 0.380 Dimethyl 10.16(±1.50)a 8.90(±0.81)a 8.60(±2.41)a NS 0.489 trisulfide Ethyl octanoate 23.10(±0.14)a 18.65(±7.69)a 18.13(±4.08)a NS 1.825 Styrene 26.56(±8.20)a 27.72(±2.67)a 26.02(±3.37)a NS 2.484 Same superscripted letters in the same row indicate no statistical significance between the two means. 1Values in parentheses indicate the standard deviations. 2P stands for the statistical probability; NS indicates no statistical significance in same row (P ≥ 0.05). 3SEM abbreviates the standard error of the mean.

4.5 Presence of pathogenic bacteria

The fresh skim milk was analyzed for L. monocytogenes, Salmonella spp., and E. coli

O157:H7 before and after the treatment. E. coli O157:H7 was not detected in raw skim milk

(detection limit = 10 CFU/mL), whereas the Salmonella spp. and L. monocytogenes counts for fresh raw skim milk were 1.51 and 1.23 log10 CFU/mL, respectively. Slade and others (1988) along with Farber and others (1998) also found prevalence of Listeria spp. in bulk milk samples whereas McEwen and others (1988) have reported the presence of Salmonella spp. in Ontario in bulk milk tank. Complete inactivation (detection limit of 10 CFU/mL) of all the pathogenic bacteria were observed after PEF and PEF/MF-1.4µm treatments, since the no presumptive colonies were observed when treated milk was plated on selective media. After 65 days of ripening at 4 °C, all the cheeses were analysed for L. monocytogenes, Salmonella spp., and E.

106 coli O157:H7. None of the cheese samples showed the presence of any pathogenic bacteria

(detection limit of 10 CFU/mL).

Antimicrobial characteristics of PEF treatment can be enhanced by increasing electric field intensity and processing temperature (Qin and others 1998; Vega-Mercado and others 1996).

Several research groups have reported the inactivation of spoilage as well as pathogenic bacteria in milk by PEF treatment. Walkling-Ribeiro and others (2009) have reported a reduction of 6.4 log10 cycle in total microbial load of milk treated by PEF at 50 kV/cm for 33 µs. Bermudez-

Aguirre and others (2009, 2011) obtained 2.4 and 3.6 log10 cycle reductions in count of mesophiles in skim milk following a PEF treatment at 46.15 kV/cm (30 pulses) with processing temperatures of 50 and 60 °C, respectively. Grahl and Markl (1996) also observed a 4.0 log10 cycle reduction in count of E. coli inoculated into UHT milk by PEF treatment at 23 kV/cm and temperature <50°C. Sharma and others (2014) applied PEF with electric field intensity of 23-28 kV/cm and inlet temperature 55 °C to inoculated whole milk and achieved a 5 - 6 log10 cycle reduction of in counts of E. coli and L. innocua. In a review of microbial inactivation by PEF

Sampedro and others (2005) reported a 4.0 log10 cycle reduction in count of S. Dublin in pasteurized, homogenized inoculated milk after PEF treatment at 36.7 kV/cm. Similarly, Sensoy and others (1997) also reported the reduction of S. Dublin by 4.0 log10 cycle in skim milk by PEF treatment at 35 kV/cm for 163.9 μs treatment time.

Membrane microfiltration (1.4 µm pore size) has also been successfully used to lower microbial count in milk and reductions of up to 4.0 and 4.5 log10 CFU/mL have been reported for total microbial and spore counts, respectively (Maubois 2002). Complete removal of somatic cells and 3.5 - 4.0 log10 cycle reduction in counts of pathogenic bacteria in milk by using 1.4 µm

107 sized ceramic membrane MF has been described by Saboya and Maubois (2000). Madec and others (1992) reported the removal of Salmonella spp. and Listeria spp. by 2.5 and 1.9 log10 cycles respectively, in skim milk by Membralox® membrane of 1.4µm pore size at 35 °C. Other research groups also proposed the use of MF as a promising alternative to heat pasteurization for cheese making and claimed that cheese made with MF treated milk was superior in terms of E. coli load and possessed comparable quality to traditionally produce cheese as well as having an overall sensory acceptability (Gcsan-Guiziou 2010, Amornkul and Henning 2007). Li and others

(2013) observed a 5.2 log10 CFU/mL inactivation of E. coli in soymilk at electric field strength of

40 kV/cm for 547 μs.

Combined application of PEF/MF may be useful to achieve higher quality milk for cheese manufacture, since MF treatment can retain the dead cell debris along with potentially active enzymes such as thermoduric proteases (Olesen and Jensen 1989; Saboya and Maubois 2000).

Walkling-Ribeiro and others (2011) have also described the effectiveness of combined PEF/MF rather than single treatment to increase microbial inactivation. The increased antimicrobial activity is due to agglomeration between electroporated cells or with milk components, which increases the volume and hence filtration efficiency of MF.

4.6 Conclusions

Raw milk Brie type cheese is a specialty cheese that is popular among consumers due to its unique flavour. These types of cheese are classified under soft cheese due to its higher moisture content, which makes it susceptible to microbial growth. Therefore, soft cheese made from raw milk may pose a public health risk due to the potential for pathogens to be present in the raw milk. Brie cheese was prepared from raw milk and milk treated with PEF and a combination of

108

PEF and 1.4 µm pore-sized MF (PEF/MF-1.4µm). Different treatments did not affect the proximate composition and Hunter color value for L* and b*, whereas, the value of a* attribute increased slightly, indicating the color moved from greenish to the red region. However, very small color differences (∆E) between raw milk cheese and those produced from milk treated by

PEF and PEF/MF-1.4µm were observed. Most of the volatile flavour compounds were not affected by PEF and PEF/MF-1.4µm treatment, while the concentration of acetic acid, which is inversely related to the flavour quality of cheese, was significantly decreased in treated cheese, indicating a positive impact on flavour. Microbial analysis of PEF and PEF/MF-1.4µm treated skim milk before cheese making showed complete inactivation of Listeria monocytogenes and

Salmonella spp., whereas, Escherichia coli O157:H7 was not found in raw milk (detection limit of 10 CFU/mL). Analysis of all cheeses after 65 days of ripening at 4 °C confirmed the absence of the pathogenic bacteria. These results indicate combination of PEF and MF as a hurdle approach has potential as an alternative to heat treatment in the dairy industry to manufacture

Brie cheese of high quality.

109

CHAPTER 5. CONCLUSIONS AND FUTURE OUTLOOK

5.1 Thesis summary and general conclusions

The high water activity, water content, almost neutral pH and an ample supply of essential nutrients in milk make it a suitable growth medium for numerous spoilage and pathogenic microorganisms. Therefore, milk is a highly perishable food and easily subject to microbial contamination by the surrounding milking environment. Heat processing of milk before consumption is a well-established technology and has been practiced for generations, and now, many countries have made the thermal pasteurization of milk mandatory before consumption.

Thermal treatment of milk is the most effective method for the destruction of microorganisms to produce microbiologically safe milk, thus ensuring public health and food safety. However, it also has some limitations: some consumers object to the cooked flavour of milk treated at high temperatures and it also brings about undesirable changes to its physicochemical, nutritional and organoleptic quality. Retention of bioactive components in milk is also of concern. Non-thermal technologies could be an alternative to thermal treatment to minimize the adverse effects of heat on food quality, since this technology runs at low temperature and foods processed by non- thermal technologies are fresh-like and their color, flavour and nutrients are less affected.

Moreover, non-thermal technologies require less energy in comparison with traditional thermal technologies, and are less expensive and more environmentally friendly than traditional thermal technologies. Pulsed electric field (PEF) and membrane microfiltration (MF) are emerging non- thermal technologies that are gaining in popularity, since they preserve food by inactivation of microorganisms using low energy inputs and reduce the time for processing as well. By combined application of PEF and MF as a hurdle technology, microbial stability and safety can

110 be achieved at low intensity of PEF treatment, thus minimizing the negative impact on the quality of foods, resulting in improved nutritional quality and sensory properties.

The aim of this project was to study the efficacy of either stand-alone or a combination of pulsed electric field (PEF) and membrane microfiltration (MF) for microbial inactivation and its effect on physico-chemical properties of skim milk compared with milk processed by HTST pasteurization. This study was also undertaken to evaluate Brie-type cheese prepared with PEF and MF treated milk on the basis of color, flavour, proximate composition and microbial screening.

Initially, a study on the effect of either stand-alone or a combination of PEF and MF on microbial inactivation and physico-chemical properties of milk was conducted, and the results compared with those achieved following HTST treatment of skim milk. Stand-alone high intensity PEF (PEF-H), MF-0.65μm and MF-1.4μm resulted in the highest level of inactivation of total microbial load in milk (4.13, 3.13 and 3.08 log10 CFU/mL respectively); whereas when the treatments were applied in combination, PEF-H/MF-0.65μm and PEF-H/1.4μm were the most efficient and reduced the total microbial load by 5.95 and 5.90 log10 cycles, respectively.

The microbial reduction achieved by this combination was higher than that obtained using HTST processing, which was 4.96 and 5.08 log10 cycle for HTST-75 and HTST-95 respectively. The reduction in microbial counts in milk obtained by PEFH/ MF-1.2μm and PEF-M/MF-1.4μm was comparable with HTST processing. Skim milk processing either by PEF alone or in combination with MF (MF-1.2μm and MF-1.4µm) did not cause significant changes in the physicochemical properties of milk (pH, electrical conductivity, °Brix, and viscosity); however, processing through either MF-0.65μm alone or in combination with PEF, resulted in a decrease in °Brix and

111 in viscosity of skim milk. These results suggest MF using a small pore-size membrane (MF-

0.65μm) is not a suitable choice for the dairy industry for commercial milk processing. Processing of skim milk inoculated with three pathogenic bacteria (E. coli, Listeria monocytogenes or

Salmonella spp.) by PEF-H and PEF-H/MF-1.2μm treatments, resulted in significant inactivation

(> 4 log10 CFU/mL) of all bacteria, indicating that hurdle technology combining PEF and MF is suitable to inactivate pathogenic bacteria.

The preparation of Brie-type cheese with PEF and MF treated milk was evaluated and color, flavour, proximate composition and presence of pathogenic bacteria were compared with raw milk cheese. Cheese prepared with PEF and MF treated milk remained unaffected with respect to proximate composition and flavour compounds; however, the concentration of acetic acid was found to be significantly lower in PEF/MF treated milk cheese compared to raw milk cheese.

Since the concentration of acetic acid is inversely related with flavour quality, this suggests that cheeses made with the PEF/MF treated milk may have a better taste. Analysis of cheese for

Hunter color value for lightness (L) and yellowness (b) were not significantly different among all the cheeses produced, whereas the value of redness (a) attribute increased slightly in cheese made from PEF/MF treated milk, indicating the color moving from greenish to the red region.

Microbial analysis of PEF and MF treated skim milk before cheese preparation did not show the presence of the two (Listeria monocytogenes or Salmonella spp.) pathogenic bacteria (detection limit of 10 CFU/mL), where as E. coli was negative in raw milk (detection limit of 10 CFU/mL), and enumeration of all cheeses after 65 days of ripening at 4 °C, confirmed the absence of the pathogenic bacteria.

112

The results obtained in this project indicate that the combination of PEF and MF as non- thermal hurdle technologies has potential in the dairy industry to produce skim milk of high quality, ensuring food safety and public health, as an alternative to traditional heat treatment.

Moreover, these non-thermal technologies can also be used to produce high quality Brie-type soft cheese.

5.2 Future outlook

The findings generated from this research are interesting and important to the combined application of PEF and MF technology. However, to make these non-thermal technologies reliable and to utilize the full potentiality of PEF and MF, these non-thermal technologies need further research. Alkaline phosphate is a pasteurization indicator for heat processing and it is unclear whether this enzyme can be used as an indicator of the efficacy of the PEF/MF treatment. Thus, research to discover a reliable indicator of process efficacy is required. Process optimization in terms of total pulsing energy, pulse frequency, pulse number or flow rate, rather than increasing electric field intensity and processing temperature may be beneficial. The main problem before commercialisation of this technology appears to be the size of the treatment chamber for PEF technology. Scale-up of the processing capacity is required for this technology to move towards commercial application. Studies on the effect of PEF on protein conformation, enzyme activity (like protease plasmin), vitamins and immunoglobulins may also help to maximize the application of this non-thermal technology. Sensory testing of Brie-type soft cheeses and other cheeses prepared from milk processed by this non-thermal technology is also necessary. Due to the nature of PEF, there is a risk of metal ion migration into the food products during processing, so more research on metal ion migration in different kinds of foods with

113 different pH and electrical conductivity at different electric intensity and processing temperature is required to prove it is safe. Some countries have mandatory laws for application of heat for any product to be sold as a pasteurized food, so thorough research on consumer acceptance and to address regulatory requirements is also essential.

114

REFERENCES

Adams MR and Moss MO. 2007. Food microbiology. Third edition. The Royal Society of Chemistry publishing, Cambridge, UK.

Agcam E, Akyıldız A, Akdemir Evrendilek G. 2014. Comparison of phenolic compounds of orange juice processed by pulsed electric fields (PEF) and conventional thermal pasteurisation. Food Chemistry 143:354-361.

Aguilar-Rosas SF, Ballinas-Casarrubias ML, Nevarez-Moorillon GV, Martin-Belloso O, Ortega- Rivas E. 2007. Thermal and pulsed electric fields pasteurization of apple juice: Effects on physicochemical properties and flavour compounds. Journal of Food Engineering 83(1):41- 46.

Alkhafaji SR and Farid M. 2007. An investigation on pulsed electric fields technology using new treatment chamber design. Innovative Food Science & Emerging Technologies 8(2):205-212.

Alvarez I, Manas P, Condon S, Raso J. 2003. Resistance variation of salmonella enterica serovars to pulsed electric fields treatments. Journal of Food Science 68(7):2316-2320.

Amornkul Y and Henning DR. 2007. Utilization of microfiltration or lactoperoxidase system or both for manufacture of cheddar cheese from raw milk. Journal of Dairy Science 90(11):4988-5000.

Arora G, Cormier F, Lee B. 1995. Analysis of odor-active volatiles in cheddar cheese headspace by multidimensional GC/MS sniffing. Journal of Agricultural and Food Chemistry 43(3):748-752.

Awad S, Ahmed N, El Soda M. 2010. Influence of microfiltration and adjunct culture on quality of domiati cheese. Journal of Dairy Science 93(5):1807-1814.

Barat M, Rodrı guez-Barona S, Andrés A, Fito P. 2003. Cod salting manufacturing analysis. Food Research International 36(5):447-453.

Barbosa-Canovas GV and Sepulveda D. 2005. Present status and the future of PEF technology. BarbosaCanovas, GV Tapia, MS Cano,MP, editor. . 44 p.

Barbosa-Canovas GV, Gongora-Nieto MM, Pothakamury UR, Swanson BG. 1999. Preservation of foods with pulsed electric fields. . xiii + 197pp.

Barbosa-Canovas GV, Pierson MD, Q Howard Zhang, Schaffner DW. 2000. Pulsed electric fields. Journal of Food Science :65-79.

Barbosa-Cánovas GV, Pothakamury UR, Palou E, Swanson BG. 1998. Nonthermal preservation of foods. Marcel Dekker.

115

Barbosa-Cánovas GV, Góngora-Nieto MM, Pothakamury UR, Swanson BG. 1999. Chapter 1 - fundamentals of high-intensity pulsed electric fields (PEF). In: Preservation of foods with pulsed electric fields. San Diego: Academic Press.

Baylis CL. 2009. Raw milk and raw milk cheeses as vehicles for infection by verocytotoxin- producing escherichia coli. International Journal of Dairy Technology 62(3):293-307.

Beckman SL and Barbano DM. 2013. Effect of microfiltration concentration factor on serum protein removal from skim milk using spiral-wound polymeric membranes. Journal of Dairy Science 96(10):6199-6212.

Bendicho S, Barbosa-Canovas GV, Martin O. 2002. Milk processing by high intensity pulsed electric fields. Trends in Food Science & Technology 13(6-7):195-204.

Berger C, Khan JA, Molimard P, Martin N, Spinnler HE. 1999. Production of sulfur flavors by ten strains of geotrichum candidum. Applied and Environmental Microbiology 65(12):5510- 5514.

Berk Z. 2013. Chapter 10 - membrane processes. Berk Z, editor. In: Food process engineering and technology (second edition). San Diego: Academic Press. 259 p.

Bermudez-Aguirre D, Mawson R, Versteeg K, Barbosa-Canovas GV. 2009. Composition properties, physicochemical characteristics and shelf life of whole milk after thermal and thermo-sonication treatments. Journal of Food Quality 32(3):283-302.

Bermúdez-Aguirre D, Fernández S, Esquivel H, Dunne PC, Barbosa-Cánovas GV. 2011. Milk processed by pulsed electric fields: Evaluation of microbial quality, physicochemical characteristics, and selected nutrients at different storage conditions. Journal of Food Science 76(5):S289.

Bi X, Liu F, Rao L, Li J, Liu B, Liao X, Wu J. 2013. Effects of electric field strength and pulse rise time on physicochemical and sensory properties of apple juice by pulsed electric field. Innovative Food Science & Emerging Technologies 17(0):85-92.

Boluda-Aguilar M, Taboada-Rodríguez A, López-Gómez A, Marín-Iniesta F, Barbosa-Cánovas GV. 2013. Quick cooking rice by high hydrostatic pressure processing. LWT - Food Science and Technology 51(1):196-204.

Brans G, Schroen CGPH, van der Sman RGM, Boom RM. 2004. Membrane fractionation of milk: State of the art and challenges. Journal of Membrane Science 243(1-2):263-272.

Brooks JC, Martinez B, Stratton J, Bianchini A, Krokstrom R, Hutkins R. 2012. Survey of raw milk cheeses for microbiological quality and prevalence of foodborne pathogens. Food Microbiology 31(2):154-158.

116

Buchin S, Delague V, Duboz G, Berdague JL, Beuvier E, Pochet S, Grappin R. 1998. Influence of pasteurization and fat composition of milk on the volatile compounds and flavor characteristics of a semi-hard cheese. Journal of Dairy Science 81(12):3097-3108.

Buckow R, Chandry PS, Ng SY, McAuley CM, Swanson BG. 2014. Opportunities and challenges in pulsed electric field processing of dairy products. International Dairy Journal 34(2):199-212.

Bugaud C, Buchin S, Hauwuy A, Coulon JB. 2001. Relationships between flavour and chemical composition of abondance cheese derived from different types of pastures. Lait 81(6):757- 773.

Bushnell AH, Clark RW, Dunn JE, Lloyd SW, inventors; May 7, 1996. Process for reducing levels of microorganisms in pumpable food products using a high pulsed voltage system. 1.

Buzrul S, Alpas H, Largeteau A, Demazeau G. 2008. Modeling high pressure inactivation of escherichia coli and listeria innocua in whole milk. European Food Research and Technology = Zeitschrift Für Lebensmittel-Untersuchung Und -Forschung.A 227(2):443-448.

Calderón-Miranda ML, Barbosa-Cánovas GV, Swanson BG. 1999. Inactivation of listeria innocua in skim milk by pulsed electric fields and nisin. International Journal of Food Microbiology 51(1):19-30.

Calvo MM and de la Hoz L. 1992. Flavour of heated milks. A review. International Dairy Journal 2(2):69-81.

Caplan Z and Barbano DM. 2013. Shelf life of pasteurized microfiltered milk containing 2% fat. Journal of Dairy Science 96(12):8035-8046.

Castro AJ, Barbosacanovas GV, Swanson BG. 1993. Microbial inactivation of foods by pulsed electric-fields. Journal of Food Processing and Preservation 17(1):47-73.

Chambers DH, Esteve E, Retiveau A. 2010. Effect of milk pasteurization on flavor properties of seven commercially available french cheese types. Journal of Sensory Studies 25(4):494-511.

Chen G, Campanella OH, Barbosa-Canovas GV. 2012. Estimating microbial survival parameters under high hydrostatic pressure. Food Research International 46(1):314-320.

Cheryan M. 1998. Ultrafiltration and microfiltration handbook. Technomic Publishing Company, Inc. xvii + 527pp. p.

Chye FY, Abdullah A, Ayob MK. 2004. Bacteriological quality and safety of raw milk in malaysia. Food Microbiology 21(5):535-541.

117

Claeys WL, Cardoen S, Daube G, De Block J, Dewettinck K, Dierick K, De Zutter L, Huyghebaert A, Imberechts H, Thiange P, and others. 2013. Raw or heated cow milk consumption: Review of risks and benefits. Food Control 31(1):251-262.

Cortés C, Esteve MJ, Frígola A. 2008. Effect of refrigerated storage on ascorbic acid content of orange juice treated by pulsed electric fields and thermal pasteurization. European Food Research and Technology = Zeitschrift Für Lebensmittel-Untersuchung Und -Forschung.A 227(2):629-635.

Coster HGL and Zimmermann U. 1975. Mechanism of electrical breakdown in membranes of valonia-utricularis. Journal of Membrane Biology 22(1):73-90.

Cserhalmi Z, Sass-Kiss A, Toth-Markus M, Lechner N. 2006. Study of pulsed electric field treated citrus juices. Innovative Food Science & Emerging Technologies 7(1-2):49-54.

Daufin G, Escudier JP, Carrere H, Berot S, Fillaudeau L, Decloux M. 2001. Recent and emerging applications of membrane processes in the food and dairy industry. Food and Bioproducts Processing 79(C2):89-102.

Debon J, Prudêncio ES, Petrus JCC, Fritzen-Freire CB, Müller CMO, Amboni RDdMC, Vieira CRW. 2012. Storage stability of prebiotic fermented milk obtained from permeate resulting of the microfiltration process. LWT - Food Science and Technology 47(1):96-102.

Devlieghere F, Vermeiren L, Debevere J. 2004. New preservation technologies: Possibilities and limitations. International Dairy Journal 14(4):273-285.

Dunn J. 1996. Pulsed light and pulsed electric field for foods and eggs. Poultry Science. 75(9):1133-1136.

Dunn J and Pearlman JS, inventors; Methods and apparatus for extending the shelf life of fluid food products. US Patent 4,695,472.

Dutreux N, Notermans S, Wijtzes T, Góngora-Nieto MM, Barbosa-Cánovas GV, Swanson BG. 2000. Pulsed electric fields inactivation of attached and free-living Escherichia coli and Listeria innocua under several conditions. International Journal of Food Microbiology 54(1– 2):91-98.

Eduardo Garcia-Amezquita L, Ricardo Primo-Mora A, Angel Guerrero-Beltran J, Barbosa- Canovas G, Roberto Sepulveda D. 2013. Rennetability of cheese-making milk processed by nonthermal technologies. Journal of Food Process Engineering 36(2):247-253.

El Darra N, Grimi N, Maroun RG, Louka N, Vorobiev E. 2013. Pulsed electric field, ultrasound, and thermal pretreatments for better phenolic extraction during red fermentation. European Food Research and Technology 236(1):47-56.

118

Elwell MW and Barbano DM. 2006. Use of microfiltration to improve fluid milk quality. Journal of Dairy Science 89 Suppl 1:E20-E30.

Evert-Arriagada K, Hernandez-Herrero MM, Juan B, Guamis B, Trujillo AJ. 2012. Effect of high pressure on fresh cheese shelf-life. Journal of Food Engineering 110(2):248-253.

Evrendilek GA and Zhang QH. 2005. Effects of pulse polarity and pulse delaying time on pulsed electric fields-induced pasteurization of E. coli O157 : H7. Journal of Food Engineering 68(2):271-276.

Evrendilek GA, Li S, Dantzer WR, Zhang QH. 2004a. Pulsed electric field processing of beer: Microbial, sensory, and quality analyses. Journal of Food Science 69(8):M228-M232.

Evrendilek GA, Yeom HW, Jin ZT, Zhang QH. 2004b. Safety and quality evaluation of a yogurt- based drink processed by a pilot plant pef system. Journal of Food Process Engineering 27(3):197-212.

Farber JM, Sanders GW, Malcolm SA. 1988. The presence of Listeria spp. in raw milk in Ontario. Canadian Journal of Microbiology 34(2):95-100.

Fernandes R. 2009. Microbiology handbook. LFRA microbiology handbook. 2nd ed. Leatherhead Pub.

Fernandez Garcia L, Alvarez Blanco S, Riera Rodriguez FA. 2013. Microfiltration applied to dairy streams: Removal of bacteria. Journal of the Science of Food and Agriculture 93(2):187-196.

Fernandez-Molina JJ, Fernandez-Gutierrez SA, Altunakar B, Bermudez-Aguirre D, Swanson BG, Barbosa-Canovas GV. 2005. The combined effect of pulsed electric fields and conventional heating on the microbial quality and shelf life of skim milk. Journal of Food Processing and Preservation. 29(5-6):390-406.

Fleischman GJ, Ravishankar S, Balasubramaniam VM. 2004. The inactivation of listeria monocytogenes by pulsed electric field (PEF) treatment in a static chamber. Food Microbiology 21(1):91-95.

Floury J, Grosset N, Lesne E, Jeantet R. 2006a. Continuous processing of skim milk by a combination of pulsed electric fields and conventional heat treatments: Does a synergetic effect on microbial inactivation exist? Lait 86(3):203-211.

Floury J, Grosset N, Leconte N, Pasco M, Madec MN, Jeantet R. 2006b. Continuous raw skim milk processing by pulsed electric field at non-lethal temperature: Effect on microbial inactivation and functional properties. Lait 86(1):43-57.

119

Fox MB. 2007. Microbial inactivation kinetics of pulsed electric field treatment. Lelieveld HLM, Notermans S,Haan SWHd, editors. In: Food preservation by pulsed electric fields. Woodhead Publishing. 127 p.

Fox PF, Paul L. H. McSweeney PLH, Cogan TM, Guinee TP. 2000. Fundamentals of cheese science. Aspen Pub.

Fritsch J and Moraru CI. 2008. Development and optimization of a carbon dioxide-aided cold microfiltration process for the physical removal of microorganisms and somatic cells from skim milk. Journal of Dairy Science 91(10):3744-3760.

Gaya P, Medina M, Rodriguez-Marin MA, Nuñez M. 1990. Accelerated ripening of ewes’ milk manchego cheese: The effect of elevated ripening temperatures. Journal of Dairy Science 73(1):26-32.

GCsan-Guiziou G. 2010. Removal of bacteria, spores and somatic cells from milk by centrifugation and microfiltration techniques. Griffiths MW, editor. Improving the Safety and Quality of Milk, Volume 1: Milk Production and Processing: 349-372.

Giner J, Gimeno V, Espachs A, Elez P, Barbosa-C novas GV, Martı n O. 2000. Inhibition of tomato (licopersicon esculentum mill.) pectin methylesterase by pulsed electric fields. Innovative Food Science & Emerging Technologies 1(1):57-67.

Goff HD and Griffiths MW. 2006. Major advances in fresh milk and milk products: Fluid milk products and frozen desserts. Journal of Dairy Science 89(4):1163-1173.

González-Martín I, Hernández-Hierro JM, González-Pérez C, Revilla I, Vivar-Quintana A, Lobos Ortega I. 2014. Potential of near infrared spectroscopy for the analysis of volatile components in cheeses. LWT - Food Science and Technology 55(2):666-673.

Gordon DT and Morgan ME. 1972. Principal volatile compounds in feed flavored milk. Journal of Dairy Science 55(7):905-912.

Government of Canada. 2014. Food and Drug Regulations (C.R.C., c 870) Department of Justice Minister of Justice, http://laws-lois.justice.gc.ca 2014 March 31

Gosch T, Apprich S, Kneifel W, Novalin S. 2014. A combination of microfiltration and high pressure treatment for the elimination of bacteria in bovine colostrum. International Dairy Journal 34(1):41-46.

Grahl T and Markl H. 1996. Killing of microorganisms by pulsed electric fields. Applied Microbiology and Biotechnology 45(1-2):148-157.

Grappin R and Beuvier E. 1997. Possible implications of milk pasteurization on the manufacture and sensory quality of ripened cheese. International Dairy Journal 7(12):751-761.

120

Griffiths MW. 2010. Improving pasteurised and extended shelf-life milk. Improving the Safety and Quality of Milk, Volume 1: Milk Production and Processing :277-301.

Guerra A, Jonsson G, Rasmussen A, Nielsen EW, Edelsten D. 1997. Low cross-flow velocity microfiltration of skim milk for removal of bacterial spores. International Dairy Journal 7(12):849-861.

Guerrero-Beltrán JÁ, Sepulveda DR, Góngora-Nieto MM, Swanson B, Barbosa-Cánovas GV. 2010. Milk thermization by pulsed electric fields (PEF) and electrically induced heat. Journal of Food Engineering. 100(1):56-60.

Halpin RM, Cregenzán-Alberti O, Whyte P, Lyng JG, Noci F. 2013. Combined treatment with mild heat, manothermosonication and pulsed electric fields reduces microbial growth in milk. Food Control 34(2):364-371.

Hansen R. 1988. Better market milk, better cheese milk and better low-heat milk powder with bactocatch treated milk. North European Food and Dairy Journal 54(1):39-41.

Harrison SL, BarbosaCanovas GV, Swanson BG. 1997. Saccharomyces cerevisiae structural changes induced by pulsed electric field treatment. Food Science and Technology- Lebensmittel-Wissenschaft & Technologie 30(3):236-240.

Heinz V and Knorr D. 2000. Effect of pH, ethanol addition and high hydrostatic pressure on the inactivation of bacillus subtilis by pulsed electric fields. Innovative Food Science & Emerging Technologies 1(2):151-159.

Heinz V, Toepfl S, Knorr D. 2003. Impact of temperature on lethality and energy efficiency of apple juice pasteurization by pulsed electric fields treatment. Innovative Food Science and Emerging Technologies 4(2):167-175.

Hemar Y, Augustin MA, Cheng LJ, Sanguansri P, Swiergon P, Wan J. 2011. The effect of pulsed electric field processing on particle size and viscosity of milk and milk concentrates. Milchwissenschaft-Milk Science International 66(2):126-128.

Hitzel A, Pöhlmann M, Schwägele F, Speer K, Jira W. 2013. Polycyclic aromatic hydrocarbons (PAH) and phenolic substances in meat products smoked with different types of wood and smoking spices. Food Chemistry 139(1–4):955-962.

Ho S and Mittal GS. 2000. High voltage pulsed electrical field for liquid food pasteurization. Food Reviews International 16(4):395-434.

Ho S, Mittal GS, Cross JD, Griffiths MW. 1995. Inactivation of pseudomonas fluorescens by high voltage electric pulses. Journal of Food Science 60(6):1337-1340.

121

Hoffmann W, Kiesner C, Clawin-Rädecker I, Martin D, Einhoff K, Lorenzen PC, Meisel H, Hammer P, Suhren G, Teufel P. 2006. Processing of extended shelf life milk using microfiltration. International Journal of Dairy Technology. 59(4):229-235.

Holsinger VH, Rajkowski KT, Stabel JR. 1997. Milk pasteurisation and safety: A brief history and update. Revue Scientifique Et Technique De L Office International Des Epizooties 16(2):441-451.

Huang E, Mittal GS, Griffiths MW. 2006. Inactivation of salmonella enteritidis in liquid whole egg using combination treatments of pulsed electric field, high pressure and ultrasound. Biosystems Engineering 94(3):403-413.

Huang K and Wang J. 2009. Designs of pulsed electric fields treatment chambers for liquid foods pasteurization process: A review. Journal of Food Engineering 95(2):227-239.

Huang K, Yu L, Wang W, Gai L, Wang J. 2014. Comparing the pulsed electric field resistance of the microorganisms in grape juice: Application of the weibull model. Food Control 35(1):241-251.

Huisman IH. 2000. Membrane Separations | microfiltration. Wilson ID, editor. In: Encyclopedia of separation science. Oxford: Academic Press. 1764 p.

Hulsheger H and Niemann EG. 1980. Lethal effects of high-voltage pulses on escherichia-coli- K12. Radiation and Environmental Biophysics 18(4):281-288.

Hulsheger H, Potel J, Niemann EG. 1983. Electric-field effects on bacteria and yeast-cells. Radiation and Environmental Biophysics 22(2):149-162.

Hulsheger H, Potel J, Niemann EG. 1981. Killing of bacteria with electric pulses of high-field strength. Radiation and Environmental Biophysics 20(1):53-65.

Jayarao BM, Donaldson SC, Straley BA, Sawant AA, Hegde NV, Brown JL. 2006. A survey of foodborne pathogens in bulk tank milk and raw milk consumption among farm families in pennsylvania. Journal of Dairy Science 89(7):2451-2458.

Jensen RG. 1995. Chapter 1 - introduction. Jensen RG, editor. In: Handbook of milk composition. San Diego: Academic Press.

Jeyamkondan S, Jayas DS, Holley RA. 1999. Pulsed electric field processing of foods: A review. Journal of Food Protection 62(9):1088-1096.

Johnson EA, Nelson JH, Johnson M. 1990. Microbiological safety of cheese made from heat- treated milk .3. technology, discussion, recommendations, bibliography. Journal of Food Protection 53(7):610-623.

122

Johansson T. 1998. Enhanced detection and enumeration of Listeria monocytogenes from foodstuffs and food-processing environments. International Journal of Food Microbiology 40(1–2):77-85.

Karahadian C, Josephson DB, Lindsay RC. 1985. Contribution of Penicillium spp. to the flavors of brie and camembert cheese. Journal of Dairy Science 68(8):1865-1877.

Kazmierska M, Rudzinska M, Jarosz B, Dobrzanski Z, Trziszka T. 2012. Changes of the fatty acid composition, cholesterol and cholesterol oxide contents in whole egg after pulsed electric field treatment. Archiv Fur Geflugelkunde 76(4):246-253.

Khanal SN, Anand S, Muthukumarappan K, Huegli M. 2014. Inactivation of thermoduric aerobic sporeformers in milk by ultrasonication. Food Control 37(0):232-239.

Knorr D, Geulen M, Grahl T, Sitzmann W. 1994. Food application of high electric field pulses. Trends in Food Science & Technology 5(3):71-75.

Kosikowski FV and Mistry VV. 1990. Microfiltration, ultrafiltration, and centrifugation separation and sterilization processes for improving milk and cheese quality. Journal of Dairy Science 73(6):1411-1419.

Kumar P, Sharma N, Ranjan R, Kumar S, Bhat ZF, Jeong DK. 2013. Perspective of membrane technology in dairy industry: A review. Asian-Australasian Journal of Animal Sciences 26(9):1347-1358.

Lado BH and Yousef AE. 2002. Alternative food-preservation technologies: Efficacy and mechanisms. Microbes and Infection / Institut Pasteur 4(4):433-440.

Langer AJ, Ayers T, Grass J, Lynch M, Angulo FJ, Mahon BE. 2012. Nonpasteurized dairy products, disease outbreaks, and state laws-united states, 1993-2006. Emerging Infectious Diseases 18(3):385-391.

Lau KY, Barbano DM, Rasmussen RR. 1991. Influence of pasteurization of milk on protein breakdown in cheddar cheese during aging. Journal of Dairy Science 74(3):727-740.

Ledenbach LH and Marshall RT. 2009. Microbiological Spoilage of Dairy Products. Sperber W.H. Doyle MP, editors, Compendium of the Microbiological Spoilage of Foods and Beverages, Food Microbiology and Food Safety, Springer-Verlag New York, 41-67.

Leistner L and Gorris LGM. 1995. Food preservation by hurdle technology. Trends in Food Science & Technology 6(2):41-46.

Leistner L. 2000. Basic aspects of food preservation by hurdle technology. International Journal of Food Microbiology 55(1–3):181-186.

123

Li Y, Tian W, Mo H, Zhang Y, Zhao X. 2013. Effects of pulsed electric field processing on quality characteristics and microbial inactivation of soymilk. Food and Bioprocess Technology 6(8):1907-1916.

Li-Chan E, Kummer A, Losso JN, Kitts DD, Nakai S. 1995. Stability of bovine immunoglobulins to thermal treatment and processing. Food Research International 28(1):9- 16.

Lin MP, Staples CR, Sims CA, OKeefe SF. 1996. Modification of fatty acids in milk by feeding calcium-protected high oleic sunflower oil. Journal of Food Science 61(1):24-27.

Little CL and Knochel S. 1994. Growth and survival of yersinia-enterocolitica, salmonella and bacillus-cereus in brie stored at 4 °C, 8 °C and 20 °C. International Journal of Food Microbiology 24(1-2):137-145.

Madec MN, Mejean S, Maubois JL. 1992. Retention of listeria and salmonella cells contaminating skim milk by tangential membrane microfiltration (bactocatch process). Lait 72(3):327-332.

Marco-Moles R, Rojas-Graue MA, Hernando I, Perez-Munuera I, Soliva-Fortuny R, Martin- Belloso O. 2011. Physical and structural changes in liquid whole egg treated with high- intensity pulsed electric fields. Journal of Food Science 76(2):C257-C264.

Marcos A, Esteban MA, Alcala M. 1990. Determination of water activity in brie and camembert cheese varieties by 4 different methods. Food Chemistry 38(3):189-199.

Marquez VO, Mittal GS, Griffiths MW. 1997. Destruction and inhibition of bacterial spores by high voltage pulsed electric field. Journal of Food Science 62(2):399-&.

Marshall RT. 1992. Standard methods for the examination of dairy products. 16th ed. American Public Health Association.

Martin O, Qin BL, Chang FJ, BarbosaCanovas GV, Swanson BG. 1997. Inactivation of escherichia coli in skim milk by high intensity pulsed electric fields. Journal of Food Process Engineering 20(4):317-336.

Matak KE, Sumner SS, Duncan SE, Hovingh E, Worobo RW, Hackney CR, Pierson MD. 2007. Effects of ultraviolet irradiation on chemical and sensory properties of goat milk. Journal of Dairy Science 90(7):3178-3186.

Mathys A, Toepfl S, Siemer C, Favre L, Benyacoub J, Hansen CE, inventors; 2013. Pulsed electric field treatment process and dairy product comprising bioactive molecules obtainable by the process. .

Maubois JL. 2002. Membrane microfiltration: A tool for a new approach in dairy technology. Australian Journal of Dairy Technology 57(2):92-96.

124

McEwen SA, Martin SW, Clarke RC, Tamblyn SE, McDermott JJ. 1988. The Prevalence, Incidence, Geographic Distribution, Antimicrobial Sensitivity Patterns and Plasmid Profiles of milk Filter Salmonella Isolates from Ontario Farms. Can. J. Vet. Res. 52:18-22.

McGuffey RK and Shirley JE. 2011. Introduction | history of dairy farming. Fuquay JW, editor. In: Encyclopedia of dairy sciences (second edition). San Diego: Academic Press. 2 p.

McSweeney PLH, Fox PF, Lucey JA, Jordan KN, Cogan TM. 1993. Contribution of the indigenous microflora to the maturation of cheddar cheese. International Dairy Journal 3(7):613-634.

Mcsweeney PLH, Lucey JA, Jordan K, Cogan TM, Fox PF. 1991. Cheddar cheese made from microfiltered milk. Journal of Dairy Science 74(SUPPL. 1):91-91.

Meersohn M. 1989. Nitrate free cheese making with bactocatch. North European Food and Dairy Journal 55(4/5):108-113.

Michalac S, Alvarez V, Ji T, Zhang QH. 2003. Inactivation of selected microorganisms and properties of pulsed electric field processed milk. Journal of Food Processing and Preservation 27(2):137-151.

Min S, Evrendilek GA, Zhang HQ. 2007. Pulsed electric fields: Processing system, microbial and enzyme inhibition, and shelf life extension of foods. IEEE Transactions on Plasma Science 35(1):59-73.

Modi R, Hirvi Y, Hill A, Griffiths MW. 2001. Effect of phage on survival of Salmonella Enteritidis during manufacture and storage of cheddar cheese made from raw and pasteurized milk. Journal of Food Protection 64(7):927-933.

Moio L, Langlois D, Etievant P, Addeo F. 1993. Powerful odorants in bovine, ovine, caprine and water-buffalo milk determined by means of gas-chromatography olfactometry. Journal of Dairy Research 60(2):215-222.

Molimard P and Spinnler HE. 1996. Review: Compounds involved in the flavor of surface mold- ripened cheeses: Origins and properties. Journal of Dairy Science 79(2):169-184.

Monfort S, Mañas P, Condón S, Raso J, Álvarez I. 2012. Physicochemical and functional properties of liquid whole egg treated by the application of pulsed electric fields followed by heat in the presence of triethyl citrate. Food Research International 48(2):484-490.

Moody A, Marx G, Swanson BG, Bermúdez-Aguirre D. 2014. A comprehensive study on the inactivation of escherichia coli under nonthermal technologies: High hydrostatic pressure, pulsed electric fields and ultrasound. Food Control 37(0):305-314.

Morgan F, Bodin JP, Gaborit P. 2001. Link between goat milk lipolysis and sensorial quality of lactic goat cheeses made from raw or pasteurised milk. Lait 81(6):743-756.

125

Motarjemi Y, Moy GG, Jooste PJ, Anelich LE. 2014. Chapter 5 - milk and dairy products. Motarjemi Y and Lelieveld H, editors. In: Food safety management. San Diego: Academic Press. 83 p.

Muir DD, Hunter EA, Banks JM, Horne DS. 1996. Sensory properties of cheddar cheese: Changes during maturation. Food Research International 28(6):561-568.

Nada S, Ilija D, Igor T, Jelena M, Ruzica G. 2012. Implication of food safety measures on microbiological quality of raw and pasteurized milk. Food Control 25(2):728-731.

Neville MC and Jensen RG. 1995. E - the physical properties of human and bovine milks. Jensen RG, editor. In: Handbook of milk composition. San Diego: Academic Press. 81 p.

Noci F, Walkling-Ribeiro M, Cronin D, Morgan DJ, Lyng JG. 2009. Effect of thermosonication, pulsed electric field and their combination on inactivation of listeria innocua in milk. International Dairy Journal 19(1):30-35.

Noziere P, Graulet B, Lucas A, Martin B, Grolier P, Doreau M. 2006. Carotenoids for ruminants: From forages to dairy products. Animal Feed Science and Technology 131(3-4):418-450.

Nuumann E. 1992. Membrane electroporation and direct gene transfer. Bioelectrochemistryc and Bioenergetics, 28 (1-2), 247-267.

Odriozola-Serrano I, Bendicho-Porta S, Martín-Belloso O. 2006. Comparative study on shelf life of whole milk processed by high-intensity pulsed electric field or heat treatment. Journal of Dairy Science 89(3):905-911.

Olesen N and Jensen F. 1989. Microfiltration - the influence of operation parameters on the process. Milchwissenschaft-Milk Science International 44(8):476-479.

Ortigosa M, Arizcun C, Torre P, Izco JM. 2005. Use of wild lactobacillus strains in an adjunct culture for a roncal-type cheese. Journal of Dairy Research 72(2):168-178.

Park HS, Marth EH, Goepfert JM, Olson NF. 1970. The fate of salmonella typhimurium in the manufacture and ripening of low-acid cheddar cheese. Journal of Milk and Food Technology (7):280-284.

Pafylias I, Cheryan M, Mehaia MA, Saglam N. 1996. Microfiltration of milk with ceramic membranes. Food Research International 29(2):141-146.

Pagan R, Condon S, Raso J. 2005. Microbial inactivation by pulsed electric fields. Novel Food Processing Technologies 141:45-68.

Pagan R, Esplugas S, Gongora-Nieto MM, Barbosa-Canovas GV, Swanson BG. 1998. Inactivation of Bacillus subtilis spores using high intensity pulsed electric fields in

126

combination with other food conservation technologies. Food Science and Technology International 4(1):33-44.

Palgan I, Munoz A, Noci F, Whyte P, Morgan DJ, Cronin DA, Lyng JG. 2012. Effectiveness of combined pulsed electric field (PEF) and manothermosonication (MTS) for the control of listeria innocua in a smoothie type beverage. Food Control 25(2):621-625.

Papadatos A, Neocleous M, Berger AM, Barbano DM. 2003. Economic feasibility evaluation of microfiltration of milk prior to cheesemaking. Journal of Dairy Science 86(5):1564-1577.

Park HS, Marth EH, Goepfert JM, Olson NF. 1970. The fate of salmonella typhimurium in the manufacture and ripening of low-acid cheddar cheese. Journal of Milk and Food Technology (7):280-284.

Peleg M. 1995. A model of microbial survival after exposure to pulsed electric-fields. Journal of the Science of Food and Agriculture 67(1):93-99.

Pereda J, Ferragut V, Quevedo JM, Guamis B, Trujillo AJ. 2009. Heat damage evaluation in ultra-high pressure homogenized milk. Food Hydrocolloids 23(7):1974-1979.

Picart L, Dumay E, Cheftel JC. 2002. Inactivation of Listeria innocua in dairy fluids by pulsed electric fields: Influence of electric parameters and food composition. Innovative Food Science & Emerging Technologies. 3(4):357-369.

Pillonel L, Ampuero S, Tabacchi R, Bosset JO. 2003. Analytical methods for the determination of the geographic origin of emmental cheese: Volatile compounds by GC/MS-FID and electronic nose. European Food Research and Technology 216(2):179-183.

Piyasena P, Mohareb E, McKellar RC. 2003. Inactivation of microbes using ultrasound: A review. International Journal of Food Microbiology 87(3):207-216.

Pothakamury UR, BarbosaCanovas GV, Swanson BG, Spence KD. 1997. Ultrastructural changes in staphylococcus aureus treated with pulsed electric fields. Food Science and Technology International 3(2):113-121.

Pothakamury UR, Vega H, Zhang QH, BarbosaCanovas GV, Swanson BG. 1996. Effect of growth stage and processing temperature on the inactivation of E-coli by pulsed electric fields. Journal of Food Protection 59(11):1167-1171.

Puértolas E, López N, Condón S, Raso J, Álvarez I. 2009. Pulsed electric fields inactivation of wine spoilage yeast and bacteria. International Journal of Food Microbiology. 130(1):49-55.

Qin BL, Barbosa-Canovas GV, Swanson BG, Pedrow PD, Olsen RG. 1998. Inactivating microorganisms using a pulsed electric field continuous treatment system. IEEE Transactions on Industry Applications 34(1):43-50.

127

Qin BL, Zhang QH, Barbosacanovas GV, Swanson BG, Pedrow PD. 1994. Inactivation of microorganisms by pulsed electric-fields of different voltage wave-forms. IEEE Transactions on Dielectrics and Electrical Insulation 1(6):1047-1057.

Qin BL, Pothakamury UR, Vega H, Martin O, BarbosaCanovas GV, Swanson BG. 1995. Food pasteurization using high-intensity pulsed electric fields. Food Technology 49(12):55-60.

Ramsaran H, Chen J, Brunke B, Hill A, M. W. Griffiths MW.1998. Survival of Bioluminescent Listeria monocytogenes and Escherichia coli 0157:H7 in soft cheeses. J Dairy Sci 81:1810– 1817.

Raso J and Barbosa-Canovas GV. 2003. Nonthermal preservation of foods using combined processing techniques. Critical Reviews in Food Science and Nutrition 43(3):265-85.

Raso J, Calderon ML, Gongora M, Barbosa-Canovas GV, Swanson BG. 1998. Inactivation of Zygosaccharomyces bailii in fruit juices by heat, high hydrostatic pressure and pulsed electric fields. Journal of Food Science 63(6):1042-1044.

Reina LD, Jin ZT, Zhang QH, Yousef A. 1998. Inactivation of Listeria monocytogenes in milk by pulsed electric field. Journal of Food Protection 61(9):1203-1206.

Reitsma CJ and Henning DR. 1996. Survival of enterohemorrhagic Escherichia coli O157:H7 during the manufacture and curing of cheddar cheese. Journal of Food Protection 59(5):460- 464.

Ripperger S and Altmann J. 2002. Crossflow microfiltration – state of the art. Separation and Purification Technology 26(1):19-31.

Rivas A, Sampedro F, Rodrigo D, Martínez A, Rodrigo M. 2006. Nature of the inactivation of escherichia coli suspended in an orange juice and milk beverage. European Food Research and Technology = Zeitschrift Für Lebensmittel-Untersuchung Und -Forschung.A 223(4):541- 545.

Robert Marselles-Fontanet A, Puig-Pujol A, Olmos P, Minguez-Sanz S, Martin-Belloso O. 2013. A comparison of the effects of pulsed electric field and thermal treatments on grape juice. Food and Bioprocess Technology 6(4):978-987.

Rodiles-López JO, Arroyo-Maya IJ, Jaramillo-Flores ME, Gutiérrez-López GF, Hernández- Arana A, Barbosa-Cánovas GV, Niranjan K, Hernández-Sánchez H. 2010. Effects of high hydrostatic pressure on the structure of bovine α-lactalbumin. Journal of Dairy Science 93(4):1420-1428.

Rodriguez-Gonzalez O, Walkling-Ribeiro M, Jayaram S, Griffiths MW. 2011. Cross-protective effects of temperature, pH, and osmotic and starvation stresses in Escherichia coli O157:H7 subjected to pulsed electric fields in milk. International Dairy Journal 21(12):953-962.

128

Rodríguez-González O, Walkling-Ribeiro M, Jayaram S, Griffiths MW. 2011. Factors affecting the inactivation of the natural microbiota of milk processed by pulsed electric fields and cross-flow microfiltration. The Journal of Dairy Research 78(3):270-8.

Rosenberg M. 1995. Current and future applications for membrane processes in the dairy- industry. Trends in Food Science & Technology 6(1):12-19.

Rowe MT and Donaghy J. 2008. Mycobacterium bovis: The importance of milk and dairy products as a cause of human tuberculosis in the UK. A review of taxonomy and culture methods, with particular reference to artisanal cheeses. International Journal of Dairy Technology 61(4):317-326.

Ryser ET and Marth EH. 1987. Behavior of listeria-monocytogenes during the manufacture and ripening of cheddar cheese. Journal of Food Protection 50(1):7-13.

Sable S and Cottenceau G. 1999. Current knowledge of soft cheeses flavor and related compounds. Journal of Agricultural and Food Chemistry 47(12):4825-4836.

Saboya LV and Maubois JL. 2000. Current developments of microfiltration technology in the dairy industry. Lait 80(6):541-553.

Saldana G, Monfort S, Condon S, Raso J, Alvarez I. 2012. Effect of temperature, pH and presence of nisin on inactivation of salmonella typhimurium and escherichia coli O157:H7 by pulsed electric fields. Food Research International 45(2):1080-1086.

Sale AJH and Hamilton WA. 1967. Effects of high electric fields on microorganisms: I. killing of bacteria and yeasts. Biochimica Et Biophysica Acta (BBA) - General Subjects 148(3):781- 788.

Sampedro F, Rodrigo M, Martinez A, Rodrigo D, Barbosa-Canovas GV. 2005. Quality and safety aspects of PEF application in milk and milk products. Critical Reviews in Food Science and Nutrition 45(1):25-47.

Sanaa M, Coroller L, Cerf O. 2004. Risk assessment of listeriosis linked to the consumption of two soft cheeses made from raw milk: Camembert of normandy and brie of meaux. Risk Analysis 24(2):389-399.

Sensoy I, Zhang QH, Sastry SK. 1997. Inactivation kinetics of salmonella dublin by pulsed electric field. Journal of Food Process Engineering 20(5):367-381.

Sepulveda DR, Góngora-Nieto MM, Guerrero JA, Barbosa-Cánovas GV. 2009. Shelf life of whole milk processed by pulsed electric fields in combination with PEF-generated heat. LWT - Food Science and Technology 42(3):735-739.

129

Sepulveda-Ahumada DR, Ortega-Rivas E, Barbosa-Canovas GV. 2000. Quality aspects of cheddar cheese obtained with milk pasteurized by pulsed electric fields. Food and Bioproducts Processing 78(C2):65-71.

Shamsi K, Versteeg C, Sherkat F, Wan J. 2008. Alkaline phosphatase and microbial inactivation by pulsed electric field in bovine milk. Innovative Food Science & Emerging Technologies 9(2):217-223.

Sharma P, Oey I, Everett DW. 2014. Effect of pulsed electric field processing on the functional properties of bovine milk. Trends in Food Science & Technology 35(2):87-101.

Sharma P, Bremer P, Oey I, Everett DW. 2014. Bacterial inactivation in whole milk using pulsed electric field processing. International Dairy Journal 35(1):49-56.

Shin J, Jung K, Pyun Y, Chung M. 2007. Application of pulsed electric fields with square wave pulse to milk inoculated with E. coli, P.fluorescens, and B. stearothermophilus. Food Science and Biotechnology 16(6):1082-1084.

Simonne A H, H Simonne EH, Eitenmiller RR, Mills HA and Cresman III CP (1997). Could the Dumas Method Replace the Kjeldahl Digestion for Nitrogen and Crude Protein Determinations in Foods? J Sci Food Agric 1997, 73, 39-45

Simpson RK, Whittington R, Earnshaw RG, Russell NJ. 1999. Pulsed high electric field causes 'all or nothing' membrane damage in listeria monocytogenes and salmonella typhimurium, but membrane H+-ATPase is not a primary target. International Journal of Food Microbiology 48(1):1-10.

Singh RP and Heldman DR. 2014. Introduction to food engineering, 5th edition, Academic Press publications.

Slade PJ, Collins-Thompson DL, Fletcher F. 1988. Incidence of Listeria species in Ontario raw milk. Canadian Institute of Food Science and Technology Journal 21(4):425-429.

Sobrino-Lopez A, Raybaudi-Massilia R, Martin-Belloso O. 2006. High-intensity pulsed electric field variables affecting staphylococcus aureus inoculated in milk. Journal of Dairy Science 89(10):3739-3748.

Stack A and Sillen G. 1998. Bactofugation of liquid milks. Nutrition and Food Science (No. 4/5):280-282.

Stanley DW. 1991. Biological membrane deterioration and associated quality losses in food tissues. Critical Reviews in Food Science and Nutrition 30(5):487-553.

Steele JH. 2000. History, trends, and extent of pasteurization. Journal of the American Veterinary Medical Association. 217(2):175-178.

130

Toepfl S. 2011. Pulsed electric field food treatment - scale up from lab to industrial scale. Procedia Food Science 1(0):776-779.

Toepfl S, Heinz V, Knorr D. 2007. High intensity pulsed electric fields applied for food preservation. Chemical Engineering and Processing: Process Intensification 46(6):537-546.

Toepfl S, Heinz V, Knorr D. 2005. 4 - overview of pulsed electric field processing for food. Sun D, editor. In: Emerging technologies for food processing. London: Academic Press. 69 p.

Tran MTT and Farid M. 2004. Ultraviolet treatment of orange juice. Innovative Food Science & Emerging Technologies 5(4):495-502.

Trouve E, Maubois JL, Piot M, Madec MN, Fauquant J, Rouault A, Tabard J, Brinkman G. 1991. Retention of various microbial species during milk epuration by cross-flow microfiltration. Lait 71(1):1-13.

Urbach G. 1995. Contribution of lactic acid bacteria to flavour compound formation in dairy products. International Dairy Journal 5(8):877-903.

Urbach G. 1990. Headspace volatiles from cold-stored raw-milk. Australian Journal of Dairy Technology 45(2):80-85.

Vanderhorst HC and Hanemaaijer JH. 1990. Cross-flow microfiltration in the food-industry - state-of-the-art. Desalination 77(1-3):235-258.

Varnam AH and Sutherland JP. 1994. Milk and milk products. technology, chemistry and microbiology. . viii + 451pp. p.

VegaMercado H, Pothakamury UR, Chang FJ, Barbosa-Canovas GV, Swanson BG. 1996a. Inactivation of escherichia coli by combining pH, ionic strength and pulsed electric fields hurdles. Food Research International 29(2):117-121.

VegaMercado H, MartinBelloso O, Chang FJ, BarbosaCcanovas GV, Swanson BG. 1996b. Inactivation of Escherichia coli and Bacillus subtilis suspended in pea soup using pulsed electric fields. Journal of Food Processing and Preservation 20(6):501-510.

Walkling-Ribeiro M, Rodríguez-González O, Jayaram S, Griffiths MW. 2011a. Microbial inactivation and shelf life comparison of ‘cold’ hurdle processing with pulsed electric fields and microfiltration, and conventional thermal pasteurisation in skim milk. International Journal of Food Microbiology 144(3):379-386.

Walkling-Ribeiro M, Rodriguez-Gonzalez O, Jayaram S, Griffiths MW. 2011b. Processing temperature, alcohol and carbonation levels and their impact on pulsed electric fields (PEF) mitigation of selected characteristic microorganisms in beer. Food Research International 44(8):2524-2533.

131

Walkling-Ribeiro M, Noci F, Cronin DA, Lyng JG, Morgan DJ. 2009a. Shelf life and sensory evaluation of orange juice after exposure to thermosonication and pulsed electric fields. Food and Bioproducts Processing 87(C2):102-107.

Walkling-Ribeiro M, Noci F, Cronin DA, Lyng JG, Morgan DJ. 2009b. Antimicrobial effect and shelf-life extension by combined thermal and pulsed electric field treatment of milk. Journal of Applied Microbiology 106(1):241-248.

Walkling-Ribeiro M, Noci F, Cronin DA, Lyng JG, Morgan DJ. 2008. Inactivation of Escherichia coli in a tropical fruit smoothie by a combination of heat and pulsed electric fields. Journal of Food Science 73(8):M395-M399.

Walkling-Ribeiro M, Noci F, Riener J, Cronin DA, Lyng JG, Morgan DJ. 2009. The impact of thermosonication and pulsed electric fields on Staphylococcus aureus inactivation and selected quality parameters in orange juice. Food and Bioprocess Technology. 2(4):422-430.

Walkling-Ribeiro M, Noci F, Cronin DA, Riener J, Lyng JG, Morgan DJ. 2008. Reduction of staphylococcus aureus and quality changes in apple juice processed by ultraviolet irradiation, pre-heating and pulsed electric fields. Journal of Food Engineering 89(3):267-273.

Walkling-Ribeiro M, Noci F, Cronin DA, Lyng JG, Morgan DJ. 2010. Shelf life and sensory attributes of a fruit smoothie-type beverage processed with moderate heat and pulsed electric fields. Lwt-Food Science and Technology 43(7):1067-1073.

Wall K, Beveridge J, Macgregor S, Anderson J, Rowan NJ. 2003. Use of high intensity pulsed electric fields to inactivate spoilage microorganisms in wine. Abstracts of the General Meeting of the American Society for Microbiology 103:P-013.

Westhoff DC. 1978. Heating milk for microbial destruction - historical outline and update. Journal of Food Protection 41(2):122-130.

Wiedmann M, Boor KJ, Eisgruber H, Zaadhof K. 1999. Clostridium | Clostridium tyrobutyricum. Robinson RK, editor. In: Encyclopedia of food microbiology. Oxford: Elsevier. 451 p.

Wirjantoro T and Lewis M. 1996. Effect of nisin and high temperature pasteurization on the shelf-life of whole milk. Journal of the Society of Dairy Technology 49(4):99-102.

Wouters PC, Dutreux N, Smelt JPPM, Lelieveld HLM. 1999. Effects of pulsed electric fields on inactivation kinetics of Listeria innocua. Applied and Environmental Microbiology 65(12):5364-5371.

Wouters PC, Alvarez I, Raso J. 2001. Critical factors determining inactivation kinetics by pulsed electric field food processing. Trends in Food Science & Technology 12(3–4):112-121.

Young ND, Drake M, Lopetcharat K, McDaniel MR. 2004. Preference mapping of cheddar cheese with varying maturity levels. Journal of Dairy Science 87(1):11-19.

132

Yu LJ, Ngadi M, Raghavan GSV. 2009. Effect of temperature and pulsed electric field treatment on rennet coagulation properties of milk. Journal of Food Engineering 95(1):115-118.

Yu LJ, Ngadi M, Raghavan V. 2012. Proteolysis of cheese slurry made from pulsed electric field-treated milk. Food and Bioprocess Technology 5(1):47-54.

Zabbia A, Buys EM, De Kock HL. 2012. Undesirable sulphur and carbonyl flavor compounds in UHT milk: A review. Critical Reviews in Food Science and Nutrition 52(1-3):21-30.

Zenker M, Heinz V, Knorr D. 2003. Application of ultrasound-assisted thermal processing for preservation and quality retention of liquid foods. Journal of Food Protection 66(9):1642- 1649.

Zhang Q, Monsalvegonzalez A, Barbosacanovas GV, Swanson BG. 1994. Inactivation of Escherichia coli and Saccharomyces cerevisiae by pulsed electric-fields under controlled temperature conditions. Transactions of the ASAE 37(2):581-587.

Zhang QH, Qin BL, Barbosacanovas GV, Swanson BG. 1995. Inactivation of Escherichia coli for food pasteurization by high-strength pulsed electric-fields. Journal of Food Processing and Preservation 19(2):103-118.

Zhang Q, Barbosa-Cánovas GV, Swanson BG. 1995. Engineering aspects of pulsed electric field pasteurization. Journal of Food Engineering 25(2):261-281.

Zhang S, Yang R, Zhao W, Hua X, Zhang W, Zhang Z. 2011. Influence of pulsed electric field treatments on the volatile compounds of milk in comparison with pasteurized processing. Journal of Food Science 76(1):C127-C132.

Zhao W, Yang R, Shen X, Zhang S, Chen X. 2013. Lethal and sublethal injury and kinetics of Escherichia coli, Listeria monocytogenes and Staphylococcus aureus in milk by pulsed electric fields. Food Control 32(1):6-12.

Zimmermann U, Pilwat G, Riemann F. 1974. Dielectric breakdown of cell membranes. Biophysical Journal, 14,881–899.

Zimmermann U. 1986. Electrical breakdown, electropermeabilization and electrofusion. Reviews of Physiology, Biochemistry and Pharmacology 105:176-256.

133