Short communication: Electrospinning of casein/pullulan blends for food-grade applications.

Electrospinning is a complex process that produces fibers with diameters on the micrometer or nano-scale from an electrified jet of a polymer solution. The objective of this study was to create electrospun fibers for food use from aqueous solutions of calcium (CaCAS) or sodium caseinate (NaCAS). Fibers were not formed from electrospinning of solutions of either caseinate (CAS) at 50 °C, but were formed from blends of either CAS solution with aqueous solutions of the food-grade polysaccharide, pullulan (PUL), when using mass ratios from 2:1 to 1:4 of PUL/CAS. The CAS in the spinning solutions ranged from 3 to 15% (wt/wt) and the PUL ranged from 5 to 15% (wt/wt). The PUL/CaCAS 1:2 fibers showed the lowest fiber diameter sizes (FDS) of 172 ± 43 nm, as determined by scanning electron microscopy, and were smaller in size than fibers electrospun from 15% (wt/wt) PUL solution. The PUL/NaCAS solutions were more viscous and formed fibers with occasional branching and less uniform FDS at higher NaCAS contents. Reductions in NaCAS in these solutions reduced viscosity and improved jet stabilities with consequent improvement in fiber morphology leading to more uniform FDS. Fibers with less defects and more homogeneous FDS were formed from PUL/CaCAS blends with more CaCAS, showing that each CAS interacted differently with PUL and formed the best fibers at different solution conditions. Calcium bridging may also underlie the anomalous behavior of the PUL/CaCAS blends by forming crosslinks with the phosphoserine residues, further enabling chain entanglements for fiber formation. The PUL/NaCAS fibers tended to be larger than the PUL/CaCAS fibers, which may also be due to other factors such as solution surface tension and conductivity, which also affect fiber quality and size. The shear viscosities at 100 s(-1) of the solutions producing fibers were within the range of 0.07 to 0.16 Pa/s, with the smallest standard deviations in FDS noted for solutions with viscosities within about 25% that of PUL. This is the first example of caseinate fibers prepared using a food-grade carrier rendering a product with potential use in food and packaging applications.

Computer simulation of energy use, greenhouse gas emissions, and costs for alternative methods of processing fluid milk.

Computer simulation is a useful tool for benchmarking electrical and fuel energy consumption and water use in a fluid milk plant. In this study, a computer simulation model of the fluid milk process based on high temperature, short time (HTST) pasteurization was extended to include models for processes for shelf-stable milk and extended shelf-life milk that may help prevent the loss or waste of milk that leads to increases in the greenhouse gas (GHG) emissions for fluid milk. The models were for UHT processing, crossflow microfiltration (MF) without HTST pasteurization, crossflow MF followed by HTST pasteurization (MF/HTST), crossflow MF/HTST with partial homogenization, and pulsed electric field (PEF) processing, and were incorporated into the existing model for the fluid milk process. Simulation trials were conducted assuming a production rate for the plants of 113.6 million liters of milk per year to produce only whole milk (3.25%) and 40% cream. Results showed that GHG emissions in the form of process-related CO2 emissions, defined as CO2 equivalents (e)/kg of raw milk processed (RMP), and specific energy consumptions (SEC) for electricity and natural gas use for the HTST process alone were 37.6g of CO2e/kg of RMP, 0.14 MJ/kg of RMP, and 0.13 MJ/kg of RMP, respectively. Emissions of CO2 and SEC for electricity and natural gas use were highest for the PEF process, with values of 99.1g of CO2e/kg of RMP, 0.44 MJ/kg of RMP, and 0.10 MJ/kg of RMP, respectively, and lowest for the UHT process at 31.4 g of CO2e/kg of RMP, 0.10 MJ/kg of RMP, and 0.17 MJ/kg of RMP. Estimated unit production costs associated with the various processes were lowest for the HTST process and MF/HTST with partial homogenization at $0.507/L and highest for the UHT process at $0.60/L. The increase in shelf life associated with the UHT and MF processes may eliminate some of the supply chain product and consumer losses and waste of milk and compensate for the small increases in GHG emissions or total SEC noted for these processes compared with HTST pasteurization alone. The water use calculated for the HTST and PEF processes were both 0.245 kg of water/kg of RMP. The highest water use was associated with the MF/HTST process, which required 0.333 kg of water/kg of RMP, with the additional water required for membrane cleaning. The simulation model is a benchmarking framework for current plant operations and a tool for evaluating the costs of process upgrades and new technologies that improve energy efficiency and water savings.

Computer simulation of energy use, greenhouse gas emissions, and process economics of the fluid milk process.

Energy-savings measures have been implemented in fluid milk plants to lower energy costs and the energy-related carbon dioxide (CO2) emissions. Although these measures have resulted in reductions in steam, electricity, compressed air, and refrigeration use of up to 30%, a benchmarking framework is necessary to examine the implementation of process-specific measures that would lower energy use, costs, and CO2 emissions even further. In this study, using information provided by the dairy industry and equipment vendors, a customizable model of the fluid milk process was developed for use in process design software to benchmark the electrical and fuel energy consumption and CO2 emissions of current processes. It may also be used to test the feasibility of new processing concepts to lower energy and CO2 emissions with calculation of new capital and operating costs. The accuracy of the model in predicting total energy usage of the entire fluid milk process and the pasteurization step was validated using available literature and industry energy data. Computer simulation of small (40.0 million L/yr), medium (113.6 million L/yr), and large (227.1 million L/yr) processing plants predicted the carbon footprint of milk, defined as grams of CO2 equivalents (CO2e) per kilogram of packaged milk, to within 5% of the value of 96 g of CO 2e/kg of packaged milk obtained in an industry-conducted life cycle assessment and also showed, in agreement with the same study, that plant size had no effect on the carbon footprint of milk but that larger plants were more cost effective in producing milk. Analysis of the pasteurization step showed that increasing the percentage regeneration of the pasteurizer from 90 to 96% would lower its thermal energy use by almost 60% and that implementation of partial homogenization would lower electrical energy use and CO2e emissions of homogenization by 82 and 5.4%, respectively. It was also demonstrated that implementation of steps to lower non-process-related electrical energy in the plant would be more effective in lowering energy use and CO2e emissions than fuel-related energy reductions. The model also predicts process-related water usage, but this portion of the model was not validated due to a lack of data. The simulator model can serve as a benchmarking framework for current plant operations and a tool to test cost-effective process upgrades or evaluate new technologies that improve the energy efficiency and lower the carbon footprint of milk processing plants.