Massey Documents by Type
Permanent URI for this communityhttps://mro.massey.ac.nz/handle/10179/294
Browse
58 results
Search Results
Item Fabrication, characterisation, and application of functional protein aggregates derived from faba bean protein isolates : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Food Technology at Massey University, Auckland, New Zealand(Massey University, 2025-07-14) Hu, YinxuanThis thesis explores the preparation, characterisation, and applications of plant protein aggregates, derived from faba bean protein isolate (FPI). The formation of FPI aggregates was accomplished by various methods, including pH adjustments, salt addition, heat treatment, sonication, and thermosonication (TS). The physico-chemical properties and technofunctional characteristics of FPI aggregates formed by different treatments, such as ζ-potential, solubility, emulsification capability, and particle sizes, were also characterised in this study. Furthermore, the microstructure of the FPI aggregates in solutions was examined using various techniques, including light scattering, microscopies (TEM and SEM), and small angle neutron scattering. Additionally, this project further developed the TS method for formation of FPI fibrillar aggregates at pH 2 and amorphous aggregates at pH 7. The characteristics of FPI aggregates formed by TS and conventional heat treatment (CH) were analysed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and liquid chromatography linked to tandem mass spectrometry (LC-MS/MS). In addition, Thioflavin T (ThT) fluorescence, Fourier-transform infrared spectroscopy (FTIR) and circular dichroism (CD) were applied to investigate the differences in secondary structure between CH-treated FPI and TS-treated FPI, indicating that TS effectively converted FPI structures to be enriched in β-sheets. The gelation behaviours of different FPI aggregates at 10 wt% were studied by examining their rheological properties and observing the microstructure using scanning electron microscopy (SEM), indicating that TS-treatment of FPI at pH 7 facilitated the formation of stronger protein hydrogels. The functionality of FPI aggregates fabricated from various treatments at the oil-water and oil-air-water interfaces was also characterised. Emulsions (O/W) with various oil factions (ϕ) ranging from 0.2 (dilute emulsions) to 0.75 (high internal phase emulsions, HIPEs), were stabilised by suitable FPI aggregates selected based on their different physico-chemical properties. The findings indicate that higher FPI concentrations (~5 wt%) and pH values (~pH 9) result in better emulsification capabilities. Among all FPI aggregates studied in this project, fibrillar aggregates exhibited the best emulsification performance as they could stabilise emulsions with oil content up to 75% (v/v). However, emulsions stabilised by FPI aggregates induced from TS at pH 7 had the greatest application potential due to their long-term stability (up to 28 days) and compatibility with a neutral pH environment. Therefore, another study in this thesis was to investigate the application of FPI aggregates in stabilising vegetable oil-based whipped creams. TS-treated FPI at pH 7 exhibited superior functional properties compared to other treatments, such as CH and ultrasonication (US), in terms of visual appearance, overrun, and stability of whipped cream. Overall, this project provides fundamental insights into the physical-chemical and techno-functional properties of FPI aggregates, including their ability to stabilise and form emulsions, gels, and foams, with an emphasis on their potential applications in innovative food products such as 3D-printed emulsion gels and plant based whipped cream. The enhanced physicochemical and techno-functional properties of FPI aggregates fabricated in this study showed a great application potential as novel food ingredients for formulation of plant-based food products.Item Investigating the health benefits of Monty’s Surprise apple phytochemicals : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Nutritional Science at Massey University, Manawatū, New Zealand. EMBARGOED until 18 November 2027.(Massey University, 2024) Nezbedová, LindaCancer is a major health concern worldwide, requiring effective prevention strategies to reduce its burden. There is evidence suggesting that diet plays an important role in disease prevention. Apples are an example of a commonly available fruit that can be easily incorporated into a diet to reduce the risk of chronic disease and some cancers. Apples are rich in phytochemicals with chemo-preventative properties. However, assessing their benefits in humans is complex due to various factors affecting phytochemical composition and low bioavailability. Understanding and controlling for these factors is crucial for maximising their potential in preventing cancer and promoting human health. The PhD project presented in this thesis uses a multidisciplinary ‘from orchard to fork to function’ approach to evaluate the health benefits of phytochemicals in a New Zealand heritage apple cultivar known as Monty’s Surprise, with an emphasis on cancer prevention. This thesis particularly focuses on apple’s most abundant phytochemical group known as phenolic compounds. After controlling fruit harvest and maturity, Monty’s Surprise apple was found to contain high concentrations of health beneficial phenolics, especially procyanidins, with apple skin having higher total phenolic content compared to its flesh. Pureeing was identified as a suitable processing technique with minimal impact on apple’s phenolics, allowing long-term preservation and creating uniform material for use in intervention and feeding trials. Moreover, Monty’s Surprise apple extract depleted of sugars inhibited the proliferation of lung, breast, and colorectal cancer cells in a dose-dependent manner in vitro. Lastly, in a human acute feeding trial, Monty’s Surprise apple puree increased plasma antioxidant capacity and decreased postprandial glucose response compared to sugar-matched placebo. The feeding trial also showed that procyanidins were not detected in the participant’s blood, while epicatechin and chlorogenic acid, which are prevalent phenolics in Monty’s Surprise apple puree, were absorbed into the bloodstream. Therefore, these compounds were available to contribute to the improved health parameters measured in the study. This study offers practical recommendations for optimising harvesting, storage, and processing practices to preserve and more clearly elucidate the health-promoting properties of phenolics in apples. Overall, this research demonstrated the potential health benefits of Monty’s Surprise apple and its phenolics, suggesting its value as a functional food for improving human health and well-being.Item Characteristisation of acetic acid bacteria and yeast isolated from kombucha produced in New Zealand : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Food Technology at Massey University, Auckland, New Zealand. EMBARGOED until 21 August 2026.(Massey University, 2024-08-13) Wang, BoyingBackground: Kombucha is a popular functional tea beverage commonly fermented by a complex symbiotic culture of acetic acid bacteria (AAB) and yeast (SCOBY) in a base of sugared tea infusion at ambient temperature for 7-14 days. Regular consumption of Kombucha confers potential health benefits due to the presence of live cultures and high concentrations of bioactive components such as vitamins, polyphenols, and organic acids. However, industrial production of Kombucha faces challenges due to the limited information on the dynamic changes in its microbial community composition and the lack of knowledge regarding their health-promoting characteristics. The impact of fermentation conditions and added substrates on starter cultures, physicochemical characteristics and functional activities is also not well understood. Objectives: This study aimed to determine the microbiological characteristics of New Zealand Kombucha starter cultures and evaluate the probiotic potential of AAB and yeast isolated from commercial Kombucha products. The bioactive components, antioxidant activity, and antimicrobial activities of Kombucha fermented using a New Zealand starter culture under different fermentation conditions were also determined.--Shortened abstractItem Population genetics and genomics of a marsupial species : analysis of native and invasive brushtail possum populations (Trichosurus vulpecula) : a dissertation presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Conservation Biology, Massey University, Manawatū Campus, New Zealand(Masey University, 2024-04-15) Pattabiraman, NimeshikaOne of the leading causes of global biodiversity decline is the introduction of invasive pest species that destroy native flora and compete with native fauna for food and other resources. Aotearoa New Zealand is one of the foremost countries in the world that has focussed on eradicating pest species and in particular exotic mammals from the archipelago, which lacks native, terrestrial mammals. The New Zealand Government recently set in train the ambitious task of removing all mustelids, rats, and possums from the terrestrial landscape by the year 2050. Brushtail possums (Trichosurus vulpecula) were introduced to Aotearoa New Zealand from Australia in the mid-1800s, after which they were translocated across the country and have become widespread, destroying indigenous habitat, eating native birds and invertebrates, and spreading bovine TB. Control efforts have seen possum numbers decline in the last two decades from close to 75 million in 2002 to 40 million in 2020. There is, however, a gap in the scientific understanding of possum populations with respect to their genetic composition and population structure across the country, and this knowledge could help us develop effective and dynamic management strategies to eradicate possums on a nationwide scale. In this thesis, I focus on three aspects of population structure and diversity of brushtail possums. First, I investigated a small geographical study area - The Kenepuru Peninsula - where I sought evidence of genetic correlations with geography, time and fur colour. I used two types of genetic markers that target the nuclear and mitochondrial regions of possum DNA with large population samples. In every case, it was determined that the possums comprised one freely interbreeding population at this scale. In particular I demonstrated that colour morphs associated with distinct subspecies in Australia, freely interbreed in New Zealand. I then increased the scale of sampling to include representation of populations across New Zealand and Australia, with the same genetic markers. This threw light on the heterogenous nature of possum diversity in New Zealand, and showed that even after ~110 generations, possums retained genetic separation among spatial groups. Additionally, the data showed evidence of multiple possum lineages across New Zealand that are derived from several Australian populations. High haplotype diversity in New Zealand suggests that the rapidly expanding population has retained novel haplotypes and the data thus far indicated a non-homogenous (metapopulation) distribution of possums without geographical concordance. As the project progressed, I was able to apply high-throughput genotyping-by-sequencing to generate a large genomic dataset. This dataset provided much more detail of the genotypic distribution of possums in Australia and among invasive metapopulations in New Zealand, as well as informing us of the relationship between them. This large, robust database of possum population structure and genetic diversity throughout Aotearoa New Zealand will support future studies in providing informed management decisions to eradicate brushtail possums.Item Thin film electrochemical sensor for water quality monitoring : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Engineering, Massey University, Auckland, New Zealand(Massey University, 2023-12-11) Lal, KartikayFreshwater is the most precious natural resource, essential for supporting life. Aquatic ecosystems flourish in freshwater sources, and many regions around the world depend on aquatic food sources, such as fish. Nitrogen and phosphorous are the two nutrients, in particular, that are essential for growth of aquatic plants and algae. However, with rising population and anthropogenic activities, excessive amounts of such nutrients enter our waterways through various natural processes, thereby degrading the quality of freshwater sources. Elevated levels of nitrate-nitrogen content, in particular, lead to consequences for both aquatic life as well as human health, which has been a cause for concern for many decades. As recommended by the World Health Organization, the maximum permissible nitrate level in water is 11.3 mg/L. These levels are often exceeded in coastal areas or freshwater bodies that are close to agricultural land. Therefore, it is essential to monitor nitrate levels in freshwater sources in real-time, which can be achieved by employing detection methods commonly used to detect ionic content in water. Hence, a comprehensive review was carried out on various field-deployable electrochemical and optical detection methods that could be employed for in-situ detection of nitrate ions in water. The primary focus was on electrochemical methods that could be integrated with low-cost planar electrodes to achieve targeted detection of nitrate ions in water. Designing resilient sensors for real-time monitoring of water quality is a challenging task due to the harsh environment to which they are subjected. There is a significant need for sensors with attributes such as repeatability, sensitivity, low-cost, and selectivity. These attributes were first explored by evaluating the performance of silver and copper materials on three distinct geometric patterns of electrodes. The experiments produced promising results with interdigitated pattern of copper electrodes that were successful in detecting 0.1-0.5 mg/L of nitrate ions in deionised water. The interdigitated geometric pattern of electrodes were further analyzed in four distinct materials namely, silver, gold, copper, and tin with real-world freshwater samples that were collected from three different freshwater bodies. The water samples were used to synthesize varying concentrations of nitrate ions. The results showed tin electrodes performed better over other materials for nitrate concentrations from 0.1-1 mg/L in complex matrix of real-world sample. The nitrate sensor eventually needs to be deployed in freshwater bodies, hence a real-time water quality monitoring system was also built that incorporated sensors to monitor five basic water quality parameters with the aim to monitor and study the quality of water around the local area.Item Grape yield analysis with 3D cameras and ultrasonic phased arrays : a thesis by publications presented in fulfillment of the requirements for the degree of Doctor of Philosophy in Engineering at Massey University, Albany, New Zealand(Massey University, 2024-01-18) Parr, BadenAccurate and timely estimation of vineyard yield is crucial for the profitability of vineyards. It enables better management of vineyard logistics, precise application of inputs, and optimization of grape quality at harvest for higher returns. However, the traditional manual process of yield estimation is prone to errors and subjectivity. Additionally, the financial burden of this manual process often leads to inadequate sampling, potentially resulting in sub-optimal insights for vineyard management. As such, there is a growing interest in automating yield estimation using computer vision techniques and novel applications of technologies such as ultrasound. Computer vision has seen significant use in viticulture. Current state-of-the-art 2D approaches, powered by advanced object detection models, can accurately identify grape bunches and individual grapes. However, these methods are limited by the physical constraints of the vineyard environment. Challenges such as occlusions caused by foliage, estimating the hidden parts of grape bunches, and determining berry sizes and distributions still lack clear solutions. Capturing 3D information about the spatial size and position of grape berries has been presented as the next step towards addressing these issues. By using 3D information, the size of individual grapes can be estimated, the surface curvature of berries can be used as identifying features, and the position of grape bunches with respect to occlusions can be used to compute alternative perspectives or estimate occlusion ratios. Researchers have demonstrated some of this value with 3D information captured through traditional means, such as photogrammetry and lab-based laser scanners. However, these face challenges in real-world environments due to processing time and cost. Efficiently capturing 3D information is a rapidly evolving field, with recent advancements in real-time 3D camera technologies being a significant driver. This thesis presents a comprehensive analysis of the performance of available 3D camera technologies for grape yield estimation. Of the technologies tested, we determined that individual berries and concave details between neighbouring grapes were better represented by time-of-flight based technologies. Furthermore, they worked well regardless of ambient lighting conditions, including direct sunlight. However, distortions of individual grapes were observed in both ToF and LiDAR 3D scans. This is due to subsurface scattering of the emitted light entering the grapes before returning, changing the propagation time and by extension the measured distance. We exploit these distortions as unique features and present a novel solution, working in synergy with state-of-the-art 2D object detection, to find and reconstruct in 3D, grape bunches scanned in the field by a modern smartphone. An R2 value of 0.946 and an average precision of 0.970 was achieved when comparing our result to manual counts. Furthermore, our novel size estimation algorithm was able accurately to estimate berry sizes when manually compared to matching colour images. This work represents a novel and objective yield estimation tool that can be used on modern smartphones equipped with 3D cameras. Occlusion of grape bunches due to foliage remains a challenge for automating grape yield estimation using computer vision. It is not always practical or possible to move or trim foliage prior to image capture. To this end, research has started investigating alternative techniques to see through foliage-based occlusions. This thesis introduces a novel ultrasonic-based approach that is able to volumetrically visualise grape bunches directly occluded by foliage. It is achieved through the use of a highly directional ultrasonic phased array and novel signal processing techniques to produce 3D convex hulls of foliage and grape bunches. We utilise a novel approach of agitating the foliage to enable spatial variance filtering to remove leaves and highlight specific volumes that may belong to grape bunches. This technique has wide-reaching potential, in viticulture and beyond.Item The gross composition, fatty acid composition, and processability of milk from cows milked once and twice a day under New Zealand grazing conditions : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Animal Science, Massey University, Palmerston North, New Zealand(Massey University, 2023) Sanjayaranj, InthujaaIn New Zealand, milking once a day (OAD) is practised by some farmers, with the majority of farmers milking twice a day (TAD). Once-a-day milking provides several benefits such as reduced labour cost, improved labour efficiency, improved health and reproductive performance of cows, and additional employment opportunities for farmers. Once-a-day milking causes a reduction in milk yield and modifies milk composition by increasing the fat and protein percentages. The main aim of this thesis was to study the effect of OAD milking on the gross composition, fatty acids composition, and processability of milk. To achieve this, the milk from 232 cows from a OAD herd and 182 cows from a TAD herd were collected in a full season in early, mid-, and late lactation. The samples were analysed for milk composition and fatty acid composition using Fourier-transform mid-infrared (FT-MIR) spectroscopy. Calibration equations for individual and grouped fatty acids were developed using FT-MIR calibrator software. Milk coagulation properties (rennet coagulation time, curd firming rate, and curd firmness) were analysed using a Formagraph. In addition, a SNP (rs109421300) in the diacylglycerol acyltransferase 1 (DGAT1) region on bovine chromosome 14, which is associated with a higher fat percentage in milk, was selected to study its association with fatty acid composition. Solid fat content at 10°C (SFC₁₀), a parameter that predicts butter hardness, was calculated using an equation developed from the proportions of fatty acids. The proportions of de novo synthesised fatty acids (C8:0-C14:0) were significantly higher, and the proportions of preformed, long-chain fatty acids (mainly C16:0 and C18:0) were lower in milk from cows milked OAD compared with cows milked TAD. Milk coagulation properties were superior for Jersey cows compared with Holstein-Friesian and Holstein-Friesian × Jersey cows milked OAD and TAD. The CC genotype of SNP rs109421300 was associated with higher fat yield, fat percentage, protein percentage, and SFC₁₀. Milk from OAD cows had lower proportions of C16:0 suggesting it would produce more easily spreadable butter. Selecting cows with the CC genotype would be suitable for the New Zealand dairy industry with the current payment system, however, the CC genotype was less suitable for making more easily spreadable butter. Overall, there were some significant differences noted between milk from cows milked OAD and TAD in terms of fatty acid composition and processability.Item Assessment of process variables on the structural, material, and physical properties of acid milk gels : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Food Technology at Massey University, Palmerston North, New Zealand. EMBARGOED until 9 June 2026.(Massey University, 2023) Kim, Ji UkThe aim of this study was to investigate the effect of process and formulation variables on cream cheese microstructure, properties, and their interactions. A systematic approach toward assessing these effects was conducted with a model cream cheese that was developed as part of the study design. Preliminary analysis compared the structural and material properties of commercial cream cheeses against the model composition to determine the extent of formulation variations on the performance of the cream cheese. In terms of compositional variables, both the ratios of fat to moisture and protein to moisture were found to positively correlate with large deformation material properties cream cheeses. These ratios were accordingly fixed in the model cream cheese formulation used within the study. The impact of coagulation kinetics on the structural and material properties of the model cream cheese was then investigated. Milk coagulation was conducted using either glucono delta-lactone (GDL) or lactic acid starter. Acid milk gels formed using GDL showed faster pH decrease during the early stage of acidification, causing earlier gelation at higher gelation pH when compared to gels using culture as acidulant. Amongst other observations, it was seen that the application of GDL as acidulant produced greater acid gel firmness (i.e., storage modulus (G′)) and higher rigidity in the corresponding cream cheeses when compared to acid gels and cream cheeses prepared using starter cultures as the acidulant. The role of additional processing and formulation variables (homogenisation pressure, coagulation pH, coagulation temperature, and stabiliser level) on cream cheese physical, material, and microstructural properties was also determined. Protein-protein and protein-fat interactions were hypothesised as the primary structural contributors to the physical properties of cream cheese, and thus variables that impacted on these interactions were considered as most actively influencing cream cheese properties. Findings were able to show that cream cheese microstructure and its associated properties demonstrated well-defined correlations to specific and controllable processing elements within the manufacturing process, showing significance in interactions between parameters in multivariable linear regression analysis (P < 0.05). Finally, the role of mineral balance of the properties of creams cheese was investigated. Here, different concentrations of the calcium chelating agent, ethylenediaminetetraacetic acid (EDTA), were added to the milk prior to coagulation. Progressive addition of EDTA was seen to reduce levels of insoluble calcium, which in turn led to weakened structural and material properties of both acid milk gels and corresponding cream cheeses. Findings here indicated the particular contribution of ionic content and equilibrium on cream cheese properties during manufacture.Item Characterising honeys in situ by spectral methods : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Food Technology at Massey University, Manawatu, New Zealand. EMBARGOED to 20 September 2024.(Massey University, 2023) Truong, Hien Thi DieuNew Zealand mānuka honey is derived mainly from Leptospermum scoparium nectar and is valuable through accumulation of antibacterial methyl glyoxal (MGO). Mānuka honey also has a strong polyphenolic profile. Some phenolics act as chemical markers aiding verification of botanical origin as Leptospermum scoparium. A total of eight key chemical markers (DHA, MGO, 3-PLA, 4-HPLA, 2’-MAP, 2-MBA, Leptosperin and LepteridineTM) are found at higher levels in mono-mānuka honey than in multi-mānuka honey, with little or none in other floral honeys. These key markers signify mānuka honey quality and purity (i.e., monoflorality of L. scoparium). The quality and purity of mānuka honeys depend on multiple factors, largely determined by botanical source, which define the value of the final honey product. Available nectar is, in turn, influenced by geographic district and season. Wild harvest honey is naturally a mixture from different nectars. Honey quality varies among apiaries, between beehives and even in a honey frame. Current industry practice lumps all frames of the same apiary together for extraction. Potentially, “good” quality frames of mānuka honey could be mixed with “bad” quality frames. This bulked process can limit the monetary value of mānuka honey. Quality assessment of honey while still in the frame before bulk extraction is of great of interest to the honey industry to preserve the value of mānuka honey at source and to ensure authenticity. The current study used rapid and non-destructive methods such as NIR and fluorescence combined with chemometrics, machine learning and deep learning to evaluate mānuka honey in the frame. The study focuses on assessment of mānuka honey quality in two ways: 1) direct measurement of levels of eight key chemical makers; 2) indirect measurement of potency (based on UMFTM score) and purity (verification of botanical origin as L. scoparium) that are built from key chemical markers. Honey samples (n ~ 1656) representing 200 L drums, each extracted from multiple frames, spanning eight geographic districts across New Zealand, were scanned with NIR non-imaging (350 - 2500 nm) and imaging (547 – 1701 nm) sensors. A sub-dataset of 100 honey samples was scanned in excitation-emission matrix mode (250 - 400/300 - 600 nm) under in-line geometry by a fluorescence sensor. Once techniques were verified, freshly uncapped seven honey frames were scanned and modelled to evaluate the current optical methods used. Overall, the research showed the capability of NIR methods for measurement of honey potency and purity in the frame, achieving 70 - 80 % accuracy. However, NIR methods showed limited ability to measure levels of individual key chemical markers, giving 60 - 70 % accuracy, due to the complexity of the honey matrix. This study has calculated the economic benefit of using NIR methods for sorting honey frames into different quality buckets (UMFTM buckets and MPI honey buckets) before lumped extraction. The greatest revenue increase is found for apiaries with large variation between frames and in seasons with high curvature in the price-quality curve. Later, this study employed fluorescence-based methods that further improved prediction of almost all key chemical markers, in particular two polyphenolic fluorescence markers, Leptosperin and LepteridineTM, to above 80 % accuracy. Moreover, the fusion of NIR and fluorescence data further enhanced predictability of chemical markers, potency and purity of mānuka honey to 90 - 100 % accuracy. In conclusion, this study confirms the fusion of NIR and fluorescence methods has great potential for at-line/on-line assessment of New Zealand mānuka honey while still in the frame. This research provides basic scientific guidance for future application of NIR and fluorescence methods for quality assessment of honey in general and has implications for other wild harvest foods.Item Browning and blistering of Mozzarella during high temperature pizza baking : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Chemical and Bioprocess Engineering at Massey University, Manawatū, New Zealand(Massey University, 2022) Thornton, DanielMozzarella and pizza baking are large and growing industries, and manufacturers and commercial operators need to know how their cheese will function during baking. The aim of this research was to develop a kinetic model describing the Maillard browning reaction occurring at the surface of Mozzarella during high temperature cooking, and to develop a conceptual model for how blister formation and growth is influenced by baking conditions. Image analysis using lightness, L*, from the CIELAB colour space (L*a*b*) colour scale was used to measure the development of brown colour during baking. An experimental method was developed using a DSLR camera to record colour data and an Infrared (IR) camera to study the Maillard browning reaction. A first order kinetic model was fitted to the lightness data using non-linear regression, with the temperature dependence found to follow the Arrhenius Law. The activation energy (Ea) and the kinetic rate constant at 120°C (k120), were fitted to two data sets. One based on an average region of interest (Average ROI kinetics, Eₐ = 124 kJ.mol⁻¹ & k₁₂₀ = 5.0x10⁻⁴ s⁻¹); and the other based on a discrete localised region where blistering occurred (P1 kinetics, Eₐ = 101.3 kJ.mol⁻¹ and k₁₂₀ = 3 .65x10⁻⁴ s⁻¹). Predicted L* was used to create a visual representation of predicted browning for the whole cheese surface which matched well with the shape and the outlines of the browned regions of the experimental data. The goodness of fit was better based on the Average ROI kinetics with an R2 of 0.88, compared to 0.66 for the P1 kinetics. A modified commercial pizza oven (Lincoln Impinger II 1100 Series) was characterised to provide information and data for the development of experimental techniques and analysis of baked pizzas. The modifications to the oven allowed for accurate temperature measurement, bake time, mass logging, and variable air flow. Air flow measurements showed that most of the air comes from below, based on the impinger layout, and there was minimal change to air flow within the control limits. The heat transfer coefficient (HTC) within the oven was measured using two methods; heating an aluminium block (29.1 to 37.1 Wm⁻²K⁻¹), and water evaporation (43.4 to 45.7 Wm⁻²K⁻¹). To analyse pizzas baked in the modified oven, an image analysis method was developed to process post baking images. MATLAB was used to process images of pizzas baked under a range of conditions, which involved colour conversion and correction, thresholding and determining characteristics. A dynamic thresholding method was developed to account for the variation in overall surface colour and provided consistent and accurate identification of blistering and browning of baked pizzas. The blister analysis process was applied to the experimental and predicted data from the kinetics experiment to determine blister characteristics. A general trend was identified, where blister numbers increased to a maximum and then reduced over time. Blister size was found to increase with time, which can be attributed to blister growth and also the joining of adjacent blisters as they continue to grow. Overall, the P1 kinetics gave a better fit across all of the experimental trials with R2 values ranging from 0.94 to 0.99, compared to the fit for the average kinetics at 0.90 to 0.98. The overall process of cheese baking can be categorised into several stages; melt and flow, steam formation, bubbling, drying and blistering and browning. The conceptual model developed, proposed that low heat flux conditions trend towards fewer but larger blisters and high heat flux conditions trend towards a greater number of small blisters. Blister nucleation was found to be likely more affected by bake time and temperature, than from initial conditions. It was found that the initial form factor (shreds vs slices) of Mozzarella did not affect the nucleation of blisters. The effect of baking conditions on blister characteristics was studied by baking pizza at a range of time and temperature combinations, with temperatures from 220 to 288°C and bake times from 6 to 15 minutes. Baking repeatability was found to be reliable and reproducible, with the % mass loss consistent with a standard error of 0.2% and the level of browning with a standard error of less than 3% for each set of operating conditions assessed. Blister characteristics were assessed from binary images based on the dynamic thresholding process. Observations from the images showed that the size of blisters decreased as oven temperature increased and baking time decreased. It was found that a higher oven temperature, and hence higher heat flux, resulted in greater blister numbers where blister coverage was less than 50%. Above 50%, it was found that further blister growth and characteristics were not sensitive to oven temperature. Analysis of the time for the onset of moisture loss and blistering relative to oven temperature found that as oven temperature increased, the time difference between the initiation of moisture loss and blister formation reduced from approximately 5 minutes at 220°C down to approximately 2 minutes at 280°C. This indicated that heat flux has a greater effect on blister initiation than overall moisture loss which suggested that heat flux has a greater effect on surface moisture loss in comparison to bulk moisture loss. This connection suggested that the pizza surface dried out faster relative to the overall moisture loss as oven temperature increased, and as moisture is lost, the viscosity of the cheese increases. At high heat flux, there is limited time for bubbles to form before a stable skin forms and the surface dries out. In addition, an increased viscosity means that there would be more resistance to growth. The resulting blisters would be smaller than for low heat flux conditions. If the cheese surface dries slower, there is more time for bubble growth before viscosity increases and a stable skin and hence larger blisters are formed. This suggested that there is a critical moisture content where the cheese surface forms a stable skin where temperature can rise above 100°C. To further study the effects of baking conditions on blister formation and growth, a benchtop oven was modified to measure and record surface colour, surface temperature and mass loss in real time during baking. A series of trials were carried out to bake cheese within the modified oven at temperatures of 168°C, 175°C, and 185°C for a minimum of 25 minutes each. The 10th percentile, 90th percentile, minimum and maximum values showed a pattern where initially the overall range is wide but reduces as the average temperature approaches 100°C, and then variation increases beyond 110°C. This pattern was similar for the L* data. Across all oven temperatures the 90th percentile and maximum L* values were similar, which agreed with the observations of lower surface temperatures in the 10th percentile and minimums across all oven temperatures. In contrast, there was a significant decrease in 10th percentile and minimum L* values as oven temperature increases, due to the increased surface temperatures relative to oven temperature. It was found that the original dynamic threshold was not suitable for analysis of the benchtop oven data as a result of different lighting conditions, so a new dynamic threshold equation was developed based on sensory evaluation. The analysis of binary images created using the new equation showed that blisters formed in less time at higher temperatures and blister coverage increased linearly relative to bake time, which agreed with the earlier analysis. Heat flux was found to be significantly lower for the benchtop oven, with the time to achieve 10% blister coverage at the highest oven temperature taking approximately 6 minutes longer compared to the lowest oven temperature from the commercial oven trials. The heat flux to the cheese layer for the commercial oven experiments was estimated to compare the effect of heat flux on the average blister size between the two experiments. It was found that there was a linear correlation between heat flux and average blisters size for the combined results, which suggested that this correlation may hold true irrespective of differences in oven setup and operating conditions. For blister initiation, a similar trend was observed from the benchtop data analysis as for the commercial oven data. Combining the blister initiation data suggested that the onset of browning is highly correlated to oven temperature, which is consistent with the temperature dependence of the Maillard reaction. The developed kinetic parameters were applied to the integrated kinetic equation to assess the performance of the model in a different application. It was found that the predicted images and subsequent blister characteristics matched well with the experimental data, especially for blister coverage, with and R2 of 0.99. The findings from these experiments and analysis indicate that surface drying dynamics are fundamental to browning and blister development. Blister growth and size is governed by the moisture content and therefore viscosity, at the cheese surface. Oven temperature (and therefore heat flux) dictates how fast the surface will dry and therefore the general blister characteristics. Bake time determines the desired level of baking based on blister/browning coverage. Hence heat flux and cheese surface dynamics, in regard to the rate of moisture loss and viscosity, are the critical factors in determining browning and blistering characteristics. The experiments and methods developed, and the understanding of cheese behaviour and influential factors during baking have potential benefits for the high temperature food processing industry, especially cheese manufacture and applications. Benefits include improved accuracy and consistency of product assessment, control over baking characteristics and reduced production costs. The image analysis process developed provides an objective measure of blister characteristics, which are otherwise subject to individual interpretation. This process could be applied in a commercial setting give a robust, repeatable and consistent means of product quality and performance assessment. Given the developed understanding of the fundamentals that drive blister formation and browning it is possible to manipulate parameters that affect surface drying dynamics. This could be through external baking conditions, such as a multi-stage baking process, or adjusting the cheese composition itself to achieve a desired result. The findings from this study resulted in a kinetic model describing Maillard browning of Mozzarella during baking as well as a good insight into the fundamentals that affect blister formation and growth to better understand the Maillard reaction and its application in high temperature food processing.
