Recipient Organization
VIRGINIA POLYTECHNIC INSTITUTE
(N/A)
BLACKSBURG,VA 24061
Performing Department
(N/A)
Non Technical Summary
Dry foods such as spices, herbs, edible seeds, and nuts (known as low water activity foods (LWAF)) may harbor harmful bacteria. Each year, products are recalled due to the presence of Salmonella enterica, Escherichia coli, and Listeria monocytogenes that sometimes results in outbreaks. Reducing foodborne illnesses and food waste are important goals. Vacuum-assisted steam pasteurization is a safe and rapid method to reduce human pathogens on a variety of ready-to-eat LWAF. LWAF vary in size, composition and packaging which may impact the effectiveness. Currently, costly testing is required to establish the process parameters associated with reduction of pathogens for each type of LWAF and each piece of processing equipment. Our approach will use new state-of-the-art thermal sensing and analysis techniques with the ability to predict reduction reliably for a variety of different LWAF that vary in composition and size, allowing processors to identify "worst-case scenario" products for microbial challenge trials. Moreover, this will solve the usual scale-up problem transitioning from lab testing to commercial equipment. We will also identify non-pathogenic surrogates that can be used in the pasteurization facilities to verify the process effectiveness. The goal of the project is to aid food processors in providing safe LWAF with good sensory quality (visual, aroma, taste) to consumers.
Animal Health Component
100%
Research Effort Categories
Basic
(N/A)
Applied
100%
Developmental
(N/A)
Goals / Objectives
Objective 1: Develop validated processes to reduce Salmonella, E. coli O157:H7 and L. monocytogenes and the potential surrogate bacteria E. faecium and P. acidilactici using vacuum-assisted steam pasteurization on LWAF (macadamia nuts, mint leaves, raisins, pumpkin seeds, mustard seeds) that differ in their inherent chemical, physical and thermal properties.Hypothesis: Reduction of foodborne pathogens in LWAF will be dependent on the product density, and the relationship between the particle size and shape that, together, will determine the porosity and heat transfer within the package. We hypothesize that LWAF with increased porosity and food chemical compositions that promote increased heat transfer will require decreased heat or contact times in standardized packaging compared to LWAF with decreased porosity. Less dense products such as mint leaves will require shorter contact times or lower temperatures to achieve comparable reductions. There may be an interaction between low water activity and product density.Hypothesis: Heat resistance of E. faecium and P. acidilactici will be greater than Salmonella, E. coli O157, L. monocytogenes on each of the tested LWA commodities. We will also determine the stability of the surrogates over time to simulate conditions where the commodity would be inoculated, shipped, processed and returned for enumeration in a period that may be 7-14 d.Objective 2: Develop mathematical models to describe the dynamic heat transfer rate (heat flux) of the product as a function of steam penetration, density and porosity of packaged product. These models can be used to create a 3D visualization that will allow the prediction of the temperature of the product at any point in the package. Compare the heat transfer rate for several LWAF that have been considered for grouping based on similarities in water activity and density, and porosity.Hypothesis: Intrinsic properties of the LWAF including density and porosity will alter the heat transfer. Commodities with similar compositions may have different heat flux if the products differ in density and porosity. The heat flux can be directly measured with a sensor in both laboratory and commercial scale systems. The predicted heat transfer from the mathematical model will be comparable to the measured heat flux using the miniatured sensorsObjective 3: Determine the suitability of dynamic models for heat inactivation kinetics to predict the inactivation of the three pathogens and potential surrogates. Compare the predicted vs. observed log reductions of Salmonella, E. coli O157, L. monocytogenes and the identified surrogates on different spices, seeds and a mixed LWAF (e.g. trail mix) to determine the effectiveness of the model to predict treatment times.Hypothesis: We anticipate that inactivation times predicted for LWAF of similar properties will be effective for decontamination on other LWAF provided the heat flux and spatial temperature profiles are comparable. This will include successfully scaling to commercial facilities from small laboratory test facilities.Objective 4: Determine if vacuum-assisted saturated steam-pasteurized LWAF are different in overall sensory quality compared to control (no treatment) LWAF based on similarity testing (triangle tests). For significantly different products, characterize quality changes: color, and water activity as a function of process.Hypothesis: We anticipate that vacuum steam pasteurization of the macadamia nuts, pumpkin seeds, mustard seeds, and raisins will not result in visible or aroma differences detected by human subjects using a similarity sensory test.
Project Methods
We will choose a variety or LWAF (mint leaves, mustard seeds, cumin seeds, raisins, pumpkin seeds, macadamia nuts that vary in inherent properties of chemical composition, water activity, form (whole/seeds, pieces/leaves, ground), density, size and shapes including nuts, dried fruits, edible seeds, herbs and spices. These LWAF will be inoculated individually with a cocktail of strains of Salmonella, L. monocytogenes, E. coli O157:H7 that were isolated from LWAF foods. Stationary-phase agar grown cells, previously shown to result in increased Salmonella thermal resistance[27] will be mist-inoculated directly onto the food particles in a thin film [34] to assure the necessary high inoculum is delivered while minimizing the resultant change to water activity before subsequently dried [21]. This method compromises between wet and dry inoculation by misting a liquid inoculum onto the product, achieving only a 0.1 increase in the water activity of the inoculated product based on testing in macadamia nuts and raisins. Drying time is still necessary, but far less than if products are dipped or soaked with an inoculum.In this study, we will determine the reduction of target pathogens and potential surrogate bacteria associated with standard vacuum steam pasteurization parameters designed to achieve product temperatures of 65°C and 80°C that are commonly used by industry. We currently have a small benchtop vacuum steam pasteurization device that uses in-house steam and a vacuum pump. To determine reductions of the target pathogens for the 6 different LWAF (mint leaves, mustard seeds, cumin seeds, raisins, pumpkin seeds, macadamia nuts dried, inoculated samples will be individually packaged within muslin bags and subjected to different steam pressure, temperature and exposure times. The local temperature of the food particles and heat flux will be monitored using thermal couples inserted within the product and a novel heat flux sensor. The target pathogens and the surrogate bacteria will be enumerated from LWAF at two time periods: after inoculation and stabilization (storage at typical RH and temperature for 1 week), and after vacuum steam pasteurization. Knowing the complete temperature distribution in the bulk material allows confident prediction of the level of inactivation at every location inside. This can be accomplished by a combination of surface heat transfer and temperature measurements and a thermal model of the system.The limitations of previous inactivation models for processing LWAF are due to the lack of knowledge of the actual temperature throughout the materials as a function time. The current inactivation models use the temperature outside of the material to predict the temperature inside the material without knowing the heat transfer that occurs. This requires empirical modeling based on assumptions about how the temperature rises in the material. The actual inactivation occurs in seconds of reaching the prescribed temperature. In partnership with CO-PDs Diller and Huang we will develop a model that considers the dynamic heat transfer in the system in the prediction of the time necessary for reduction of our three target pathogens within the system. While difficult to predict the exact model fit we anticipate adapting the Weibull model, which has been widely applied to characterize inactivation of low water activity stressed pathogens by thermal treatments.Challenge and validation are vital for assessing the performance and applicability of models that are meant to be applied for predictions. Predictions of 3-log reductions will be made from these models generated in objective 3 and compared against 3-log reductions of the same microorganisms on LWAF that have similar properties to those tested in Objective 1. Examples of these may include cardamom, fennel leaves, fennel seeds, sunflower kernels, Brazil nuts and cranberries. A mixture emulating a trail mix will be used to determine the suitability of the model to predict the time needed for inactivation based on the temperature and heat flux measured in the package. For products that have models with a good fit according to the calculated Akaike's Information Criteria (AIC) and root mean squared error (RMSE) will be validated at a commercial steam processing facility. In the case of these products, only LWAF inoculated with the surrogate can be used to prevent cross-contamination of other commercial products. This can test the inactivation kinetics as well as the suitability of the heat flux sensors for commercial applications.Sensory testing and quality analysis - Sensory Similarity Testing: It is important to determine whether the most effective vacuum steam pasteurization parameters affect the sensory quality of LWAF. A control (no treatment) LWAF will be used for comparison. Similarity testing will be conducted using tetrad-test methodology, in which panelists must successfully distinguish between treated and untreated LWAF by aroma, taste/texture, and/or appearance. Only non-contaminated products will be used for this study. A minimum of 54 subjects (over 18 years old) who are frequent consumers of all six target LWAFs (mustard seeds, mint leaves, macadamia nuts, raisins, cumin seeds, and pumpkin seeds) will be recruited from the Virginia Tech and Blacksburg community. The study design will be reviewed by Virginia Tech's IRB, and panelists will provide consent prior to participation. Each subject will participate in six tetrad tests (one for each LWAF) in randomized order; samples will be uncontaminated LWAFs, treated or untreated from the same batches (e.g., one test will comprise presentation of treated and untreated raisins--two treated and two untreated). Subjects will be asked to identify which 2 samples are the same, making 2 groups of 2 samples, and to identify what quality of that sample is different. Between tetrad tests, panelists will be forced to wait for one minute to avoid sensory fatigue. Expected guessing rate is 33%. Bayesian approaches will be used to estimate a 95% compatibility interval for the actual true proportion of discriminators, which will let us determine with high certainty the in vivo effects of applying these treatments to the product. Sensory tests will be conducted using standard good practices (individual booths, white light, sample anonymization and randomization, positive air pressure and climate control, etc). Results will be collected through touchscreen computers running Compusense20 software.Color: Color will be measured with a Minolta CR-300 chromameter on the CIE L*a*b* color scale in triplicate for each sample. Color readings will be taken through clear plexiglass to ensure a flat surface for color readings and prevent dust from contaminating the lens of the chromameter.