|LONGTIN, MADYSON - NORTH CAROLINA STATE UNIVERSITY|
|CONLEY PAYTON, SUMMER - FORMER ARS EMPLOYEE|
|OSBORNE, JASON - NORTH CAROLINA STATE UNIVERSITY|
|BITZER, DONALD - NORTH CAROLINA STATE UNIVERSITY|
Submitted to: Journal of Food Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/31/2020
Publication Date: 4/1/2020
Citation: Price, R.E., Longtin, M., Conley Payton, S., Osborne, J.A., Johanningsmeier, S.D., Bitzer, D., Breidt, F. 2020. Modeling buffer capacity and pH in acid and acidified foods. Journal of Food Science. 85(4):918-925. https://doi.org/10.1111/1750-3841.15091.
Interpretive Summary: Buffer capacity is a critical factor for determining how pH may change with the addition of ingredients to acid or acidified foods, which are made safe by low pH. While models have been published in the scientific literature for predicting pH in water-based solutions, a method for identifying buffers in undefined food ingredient solutions and using the data to predict pH is needed. The models will be of value to FDA and industry for determining the pH buffering of typical ingredients in acidic food products, and calculating how these ingredients may change pH, a vital safety factor for foods primarily protected from disease causing bacteria by pH. We have developed mathematical models and methods for the analysis of food ingredients to allow pH predictions in a final product with mixed ingredients and acids typical of salad dressing products and other acid food products. The models have been shown to accurately predict pH acid-base solutions, and can predict the buffers present in food ingredients. The models may have broad applicability with many acid or acidified foods, and aid producers and regulatory agencies in determining the safety of acidified foods products.
Technical Abstract: Standard ionic equilibria equations may be used for calculating pH of weak acid and base solutions. These calculations are difficult or impossible to solve analytically for foods that include many unknown buffering components, making pH prediction in these systems impractical. We combined buffer capacity (BC) models with a pH prediction algorithm to allow pH prediction in complex food matrices from BC data. Numerical models were developed using Matlab software to estimate the pH and buffering components for mixtures of weak acid and base solutions. The pH model was validated with laboratory solutions of acetic or citric acids with ammonia, in combinations with varying salts using Latin hypercube designs. Linear regressions of observed versus predicted pH values based on the concentration and pK values of the solution components resulted in estimated slopes between 0.96 and 1.01 with and without added salts. BC models were generated from titration curves for 0.6 M acetic acid or 12.4 mM citric acid resulting in acid concentration and pK estimates. Predicted pH values from these estimates were within 0.11 pH units of the measured pH. Acetic acid concentration measurements based on the model were within 6% accuracy compared to high-performance liquid chromatography measurements for concentrations less than 400 mM, although they were underestimated above that. The models may have application for use in determining the BC of food ingredients with unknown buffering components. Predicting pH changes for food ingredients using these models may be useful for regulatory purposes with acid or acidified foods and for product development.