Towards a wicking rapid test for rejection assessment of reused fried oils: Results and analysis for extra virgin olive oil John S. Lioumbas, Angelos Zamanis, Thodoris D. Karapantsios ⇑ Division of Chemical Technology, Department of Chemistry, Aristotle University of Thessaloniki, University Box 116, 541 24 Thessaloniki, Greece article info Article history: Received 28 March 2013 Received in revised form 13 May 2013 Accepted 24 May 2013 Available online 4 June 2013 Keywords: Fried oil Rejection criteria Wicking Rapid test abstract This work investigates the potential of developing a rapid test based on wicking of oil into paper for determining whether a reused frying oil is to be rejected. To achieve this goal, wicking patterns (oil pen- etration rate and oil front shape versus time) of both fresh and prolonged fried extra virgin olive oil are optically registered at six different paper stripes. Four of them are double-ply towel papers whereas the other two are single-ply chromatographic papers. Wicking tests are performed at 20 °C and 30 °C. It is shown that the type of paper affects seriously the wicking patterns. Double-ply papers present high oil penetration rates but very irregular oil front shapes whereas single-ply papers yield lower oil penetration rates but pretty flat oil fronts. Furthermore, it is found that only under certain conditions the penetration rates obey the well known Lucas–Washburn equation. A discussion is made on the phenomena that take place during wicking of oil into paper which may cause deviations from the Lucas–Washburn equation. A semi-empirical model is proposed to describe the above deviations by incorporating the effect of time evolving pore sizes. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction The intense and complex heat and mass transfer processes dur- ing deep fat frying result in significant oil degradation (Bouchon et al., 2003) which imposes oil replenishment in sequential frying batches. The determination of the exact instant that frying oil must be replenished is a major concern for avoiding possible health risks (Bansal et al., 2010a) but also for estimating the cost of fried foods in food industry and catering applications (Kress-Rogers et al., 1990). The official analytical procedures used in laboratories to as- sess the quality of prolonged (or repeated) fried oils are very accu- rate but are also time consuming, expensive and cannot be used on the spot in industrial or catering applications. Hence, use of rapid tests is gaining attention in actual applications where there is need for quick evaluation of whether the total polar compound content of the oil has surpassed 25% by mass and so has to be discarded according to international laws (Bansal et al., 2010b; Paul and Mit- tal, 1997). A rapid test should exhibit the following characteristics: (i) correlate well with official analytical methods, (ii) provide an objective index, (iii) quantify the degree of oil degradation, (iv) be easy and inexpensive to use, (v) be independent from the nature of frying oil, (vi) have no influence from the fried food, and (vii) be safe to use in food production area (no toxic chemical, no glassware, etc.) (Gertz, 2000). Gupta (2005) and Bansal et al. (2010a) in a comprehensive re- view categorized the existing rapid tests on those based on mea- suring physical parameters (i.e. color, dielectric constant and viscosity) and those based on measuring chemical parameters (i.e. total polar compounds, oxidized/free fatty acids and carbonyl compounds). Using an analytical laboratory method (size exclusion chromatography, HPSEC) Gertz (2000) determined polar materials, polymerized triglycerides and acid value of 150 different samples of deteriorated oil. He employed these results as reference to com- pare with and evaluate different commercial rapid tests based on the determination of chemical parameters (alkali colour number: Fritest Ò , and oxidation products: Oxifrit-Test Ò ) and physical param- eters (dielectric constant: Foodoil Sensor™- FOS™, Northern Instru- ments Corp., and relative viscosity and density: Fri-Check Ò ). He found that the tests based on chemical parameters depend highly on the particular frying conditions (e.g. type of oil and food) and give poor correlation with analytical methods (r 2  0.45). On the contrary, physical parameters show good correlation with analyti- cal methods (r 2  0.894). Marmesat et al. (2007) evaluated also the performance of several commercially available rapid tests in a sig- nificant number of fried oil samples (i.e. 105 samples). These rapid tests were based on changes of either oil physical properties (i.e. viscosity – Viscofrit Ò , dielectric constant – FOS™) or oil chemical properties (i.e. carbonyl compounds – Fritest Ò , total amount of oxi- dized compounds – Oxifrit-Test Ò ) during frying. Similarly to Gertz (2000) these authors also observed that rapid tests based on 0260-8774/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jfoodeng.2013.05.037 ⇑ Corresponding author. Tel./fax: +30 2310997772. E-mail address: karapant@chem.auth.gr (T.D. Karapantsios). Journal of Food Engineering 119 (2013) 260–270 Contents lists available at SciVerse ScienceDirect Journal of Food Engineering journal homepage: www.elsevier.com/locate/jfoodeng