Food Research International 42 (2009) 165–170

Contents lists available at ScienceDirect

Food Research International

journal homepage: www.elsevier .com/ locate / foodres

Chemiluminescent assay of lipid hydroperoxides quantification in emulsions offatty acids and oils

P. Rolewski a, A. Siger b, M. Nogala-Kałucka b, K. Polewski a,*

a Department of Physics, Faculty of Food Science and Nutrition, Poznan University of Life Sciences, Polandb Department of Food Biochemistry and Food Analysis, Faculty of Food Science and Nutrition, Poznan University of Life Sciences, Poland

a r t i c l e i n f o

Article history:Received 3 July 2008Accepted 25 September 2008

Keywords:Luminol chemiluminescenceHydroperoxidePhosphatidylcholineLinoleic acidLinseed oilEmulsion

0963-9969/$ - see front matter � 2008 Elsevier Ltd. Adoi:10.1016/j.foodres.2008.09.010

* Corresponding author.E-mail address: polewski@up.poznan.pl (K. Polews

a b s t r a c t

Lipid hydroperoxides (LOOH) are relatively stable intermediates which arise during oxidation of fats andlipids. The luminol-enhanced chemiluminescent (LCL) method has been applied to detect the LOOHformed during thermal oxidation. The method was optimized for best signal to noise ratio what gave reli-able and sensitive data up to 50 pM of detected hydroperoxides (approximately 1 lmol hydroperoxides/kg lipid). The observed LCL signal was calibrated with 13-HPODE (13-(S)-hydroperoxy-9Z,11E-octadeca-dienoic acid) as an external hydroperoxide standard. This allowed quantitative determination of LOOHformed during thermal oxidation of linoleic acid and phosphatidylcholine (PC) in emulsion and duringstorage of linseed oil. The obtained LCL results were validated using HPLC, spectrophotometric assayfor conjugated dienes at 234 nm and iodometric titration peroxide value (PV) methods.

The described LCL method is direct, sensitive, fast and simple, however, it is not specific to the lipidsand may be used to determine the presence of the other types of hydroperoxides or the other oxidizingagents what finally gives overall total antioxidant estimation of the sample.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Lipid oxidation is considered a free radical chain reaction, withthe first step being hydrogen abstraction, followed usually by lipidhydroperoxide formation. In the absence of decomposing factors asredox-active metals, heat, ultraviolet light or antioxidants hydro-peroxides are relatively stable products of lipid oxidation. In suchcase the formation of hydroperoxides is a good measure to detectearly stages of lipid oxidation, however, the reaction conditionand solvent should be taken into account. There are many analyt-ical methods available to measure oxidation of lipids. The choice ofthe method depends on what parameters are being measured andwhich phase of oxidation we want to monitor. It seems that amongdifferent authors emerged the opinion that the total hydroperoxidecontent is a good measure of oxidation progress. This may be mea-sured by a variety of methods including iodometric titration(Antolovich, Prenzler, Patsalides, Mcdonald, & Robards, 2002;Simic, Jovanovic, & Niki, 1992), colorimetric detection with theFOX method (Dobarganes & Velasco, 2002; Frankel, 1991), FTIRspectroscopy (Dauben, Lober, & Fullerton, 1969; Sedman, van deVoort, & Ismail, 1997), conjugated dienes and trienes (Antolovichet al., 2002; Wanasundara, Shahidi, & Jablonski, 1995; Waters,1971). However, each of those methods has some drawbacks.Chemiluminescence (CL) offers some advantages to overcome

ll rights reserved.

ki).

shortcomings of the other methods as superior sensitivity andsimplicity.

The use of CL method for analyzing lipid peroxidation was re-ported by Vladimirov and Petrenko (1976). From this time thismethod has been modified to avoid some drawbacks and wasadopted to study the quality of food. Review of chemiluminescentmethods used in food analysis were given (Jimenez & Navas, 2002;Navas & Jimenez, 1996) and some information was included in re-views regarding lipid oxidation (Dodeigne, Thunus, & Lejeune,2000; Roginsky & Lissi, 2005; Wheatley, 2000).

CL originates from excited state product of two ground statemolecules. The chemical energy gained is translated into electron-ically excited state which finally, trough photon emission, relaxesto ground state. One mechanism suggests the coupling of two per-oxy radicals (ROO�) to form tetroxodie (ROOOOR), which furtherdecomposes it results in photon emission, so called Russel mecha-nism. Another mechanism indicates that degradation of hydroper-oxides (ROOH) will produce alkoxy radicals (RO�) or similarreactive species that in next step will lead to b-scission mechanismfollowed by photon emission. This ultraweak CL is related to onlypart of the oxidation process where decomposition reactions gen-erate excited carbonyls, which eventually relax and emit photons.It has been shown that ultraweak CL is accompanied during oxida-tion of many compounds including hydrocarbons and lipids (Navas& Jimenez, 1996). Due to the very low quantum yield (less than10�4) the detection of CL intensity requires sophisticated methodslike single photon counting method. At early stages of oxidation

166 P. Rolewski et al. / Food Research International 42 (2009) 165–170

the reaction of addition to double bonds dominates since reactiveconjugated dienes are present at this stage. However, these addi-tion reactions do not cause any chemiluminescence. In order to in-crease the CL intensity and avoid specialized apparatus the luminoland lucygenin as light amplifiers have been introduced. The lumi-nol-enhanced chemiluminescence (LCL) involves oxidation ofluminol in basic solution generating a free radical intermediatewhich reacts with flux of oxidizing agents present in the system,e.g. H2O2, lipid hydroperoxides. This leads to formation luminol-derived product in excited state which eventually returns toground state emitting strong blue light at 430 nm. In general, theluminol chemiluminescent reaction in aqueous solution occursaccording mechanism given by Miyazawa, Fujimoto, Suzuki, andYasuda (1994), however, the exact mechanism is not known. Basedon this mechanism we may assume that LCL observed from lipidhydroperoxides/luminol system will follow the mechanism givenbellow. In basic solution the formation of mono- and dianion ofluminol occurs, Lum2�. Next, hemin catalyzed decomposition ofhydroperoxide occur. LOOH + hemin ? LO� + LO� + Lum2�. Luminolanions react with LO� radicals. This leads to formation of luminol-derived free radical (diazamiquinone radical) Lum� and LOH. Inter-action of oxygen O2 with luminol radical leads to formation ofsuperoxide radical, O�2�, which interacting with Lum� leads to for-mation of transient luminol endoperoxide, Lum–O–O�. Finally,endoperoxide decomposes to give light emission (CL) and prod-ucts, aminophthalate and N2. Hemin in this reaction plays a roleof catalyst and cooxidant.

LCL method has frequently been used to monitor on-line detec-tion for lipid hydroperoxides in connection with HPLC what hasallowed the quantification of various hydroperoxides at picomolelevel (Chan, Cheng, Tsao, Niu, & Hong, 1996; Henderson, Slickman,& Henderson, 1999; Kondo, Kawai, Miyazawa, & Mizutani,1993; Miyazawa, 2000; Ohba, Kuroda, & Nakashima, 2002;Sanchez-Moreno & Larrauri, 1998). In all reported cases the appli-cation of LCL shifted the detection limits of hydroperoxide to thelevel of single picomoles, the sensitivity unreachable by the otheranalytical methods (Navas & Jimenez, 1996).

Several LCL methods for lipid hydroperoxide detection are re-ported in the literature like stationary (Matthaeus, Wiezorek, &Eichner, 1995) or flow injection method (Bunting & Gray, 2003),or providing only qualitative data (Burkow, Moen, & Overbo,1992; Pettersen, 1994; Saito & Nakamura, 1989), or post-columnquantitative data (Miyazawa, 2000; Miyazawa, Kunika, Fujimoto,Endo, & Kaneda, 1995; Yasaei, Yang, Warner, Daniels, & Yuoh,1996). However, no report is available which directly measuresthe content of hydroperoxides during oxidation process. The objec-tive of this study was to report a new procedure for quantitativedetermination of lipid hydroperoxide formation using luminol-en-hanced CL in the phospholipids and oil emulsions. The use of adetergent increases the emulsion stability and efficiency of the sys-tem due to preventing the side reactions. In order to validate ourstudies we have calibrated the chemiluminescence scale with lin-oleic acid hydroperoxide 13-HPODE (13-(S)-hydroperoxy-9Z,11E-octadecadienoic acid) as external standard. The LCL intensity isdose-dependent and increases with increasing hydroperoxidesconcentration. In this work we present a rapid, reliable LCL methodto directly measure the amount of formed hydroperoxides in lipidemulsions during oxidation.

2. Materials and methods

2.1. Chemicals

Luminol (5-amino-2,3-dihydro-1,4-phthalazinedione), TRITONX-100 (t-octylphenoxypoly-ethoxyethanol), buffer CAPS (3-[cyclo-

hexylamino]-1-propansulfonic acid, methanol, hemin, linoleic acidand phosphaditylcholine (PC) were purchased from SIGMA.Linseed oil (Linolia) was purchased from Natural Fiber Institute(Poznan, Poland). Linseed oil containing essential fatty acids, phy-tosterols, phospholipids, vitamins and content of a-linolenic acidwas estimated to 60%. Peroxide value (PV) was determinedaccording to the standard ISO 3960:1977 (E). The method is basedon the iodometric titration, which measures the iodine producedfrom potassium iodide by the peroxides present in the oil. Theresult is reported in milli-equivalents of oxygen per kilogram offat. For PV measurement the volume of 50 ml of linseed oil washeated and every 30 min appropriate amount of oil was taken foranalysis.

Doubly distilled water and the other reagents used were ofHPLC grade. Pure 13-HPODE needed for chemiluminescencecalibration was obtained by enzymatic oxygenation of linoleicacid (Nogala-Kalucka, Kupczyk, Polewski, Siger, & Dwiecki, 2007).

2.2. Methods

The synthesis of free fatty acid hydroperoxides (FAOOHs) stan-dards was carried out as described by Spiteller, Kern, Reiner, andSpiteller (2001) and Spiteller and Spiteller (1997) but their proce-dures were modified for the synthesis of 13-hydroperoxy-octade-cadienoic acid (13-HPODE) standards and details are givenelsewhere (Nogala-Kalucka et al., 2007).

2.3. Oxidation of linoleic acid and lipids

Weighted amount of linoleic acid, linseed oil or 2 mg/ml of PCsuspended in water – sonicated until clear solution was obtained– were incubated at 60 ± 1 �C for maximum 24 h without light inthe presence of air. The hydroperoxide content was determinedevery 30 min. Oxidized solution (100 ll) was added to 0.1 M CAPSbuffer with 5% methanol and 0.025% TRITON X-100. The sampleswere sonicated (IS-K1, InterSonic, Poland) for 2–3 min in ice-coldwater to form emulsion. After sonication the solution cooled downto room temperature and was used for LCL analysis by injectingluminol and hemin. For temperature dependent measurementsthe heating temperature was varied.

UV absorption of conjugated diene hydroperoxides during lipidoxidation was measured using Shimadzu UV 1200 spectrophotom-eter (Shimadzu Ltd.)

2.4. Chemiluminescence measurements

The chemiluminescent solution consisted of: luminol 0.1 mM,5% methanol, 0.025% TRITON X-100, buffer CAPS 0.1 M pH 10,hemin 5 � 10�6 M and source of hydroperoxides (standard, linoleicacid, linseed oil or PC). Chemiluminescent reaction started by add-ing hemin to the mixture containing thermally oxidized lipidemulsion. A 3 ml round, flat cuvette was placed directly before aphotocathode of an RCA EMI 9558 QB photomultiplier connectedto data collecting board in PC computer. As a measure of CL theintegrated area under registered signal was used, I0. In order to ob-tain blank level of CL emission the registration started before anycomponent was added to the cuvette.

2.5. Statistical analysis

Experiments were repeated several times and when appropriatethe SD, errors and correlation coefficients were calculated. Statisti-cal calculation and graphs were done in program ORIGIN (Origin-Lab Corp., MA, USA).

P. Rolewski et al. / Food Research International 42 (2009) 165–170 167

3. Results and discussion

3.1. Relation between 13-HPODE concentration and LCL

In order to estimate the sensitivity and feasibility of the LCL-method, the LCL intensity scale was calibrated using 13-HPODEas external standard with concentrations ranging from 10 pM to100 nM. The calibration curve, Fig. 1, shows linear relationship be-tween 13-HPODE concentration and LCL with high correlationcoefficient (r = 0.986). Background CL due to hemin radicals oxida-tion of fatty acids and the other reactions related to formed tran-sient free radicals were observed in the absence of luminol andwere measured separately to be during calculations subtractedfrom the LCL signal. The background level determined the minimaldetectable amount of lipid hydroperoxides and the sensitivity wasestimated about 50 pM. These data demonstrated that detection ofhydroperoxides using the assay is possible and that LCL response islinear in the range studied.

3.2. Reagent composition studies

The data reported by Matthaeus et al. (1995) and Bunting andGray (2003) indicated on the problem connected with the inter-faces existing in multiphasic system. In order to solve this problemBunting and Gray (2003) proposed monophasic system whereluminol and cyt c were replaced by lucygenin. The final systemshowed the linear relationship between LOOH and CL intensity,however it remained monophasic only when oil concentration be-low 1% (wt/vol) was used.

Taking into account the reported results and known obstacleswe decided to optimize the emulsion system by carrying out someexperiments with changing concentration and ratios between thecomponents of the system. This should allow to obtain the highestsensitivity and wide concentration range to detect hydroperoxideswith high reproducibility and low background.

The most pronounced changes of LCL are connected with pH ofthe solution. However, higher pH increases the rate of hydroperox-ide decomposition and the other reaction elevating the back-ground. Fig. 2a shows the relation between LCL, background CLand pH of the solution. The chosen pH value is a tradeoff betweenthose important factors affecting sensitivity and reproducibility of

Fig. 1. Plot of integrated LCL signal versus 13-HPODE concentration. The reactionmixture is optimized with final concentrations of luminol 0.1 mM, 5% methanol,0.025% TRITON X-100, buffer CAPS 0.1 M pH 10 and hemin 5 � 10�6 M. Thebackground CL, the signal recorded without luminol in the sample, was subtractedbefore plotting calibration data.

the method. Hemin concentration is another factor influencing theshape of CL. Its increasing concentration is narrowing the shapeand at higher concentration it inhibits the CL, what is shown inFig. 2b. The plot of LCL signal with increasing luminol concentra-tion exhibits maximum at 40 lM and then remain constant, asshown in Fig. 2c. During the measurements even higher concentra-tion of luminol was used in order to increase oxidative action of thesubstrate towards luminol. Detergent, below its cmc, was used inorder to assure better dispersion, to improve solubility and main-tain stability of the emulsion, not to increase CL intensity. Deter-gent, when used with concentration above its cmc, produced avery broad peak with slowly decaying LCL, probably due to micelleformation and diluting the substrate, data not shown. The finalconcentrations of the substrates were chosen on the basis of thesensitivity to reproducibility ratio.

The kinetics of optimized chemiluminescent reaction is pre-sented in Fig. 3. It shows maximum a few seconds after additionof hemin, which decays exponentially until CL intensity reachesabout 10% of the maximal value, what takes about 2 min. In the ab-sence of luminol a measurable ultraweak CL from the other oxida-tion reactions was recorded and this background emission is alsoshown in Fig. 3.

3.3. Quantification of LCL with LOOH by conjugated dienes method

In order to determine whether LCL method would providequantitative data with agreement with another method the otherexperiment has been carried out. The dependence between theLCL signal and concentration of 13-HPODE sample diluted in meth-anol determined by absorption at kmax = 234 nm by so called con-jugated dienes method was measured. The results are presentedin Fig. 4. The plot is linear up to 500 nM of 13-HPODE and it showshigh correlation coefficient (r = 0.993) between observed LCL andconjugated dienes formation. The data confirmed that the LCL as-say provides quantitative data on lipid hydroperoxide formation.

3.4. LOOH determination during thermal oxidationof linoleic acid and PC

In order to check feasibility of LCL method we determined theamount of LOOH formed during thermal oxidation of linoleic acidat 50, 60 and 70 �C. The results obtained are presented in Fig. 5.As expected the data show that increasing temperature increasesLCL intensity as well as at given temperature the increase ofLCL with elapsed oxidation time is observed. In both cases increas-ing LCL intensity indicates higher quantities of the formedhydroperoxides.

Temperature study at 50 �C was carried for the phosphatidyl-choline emulsion with results presented in Fig. 5. Also in this casethe increasing time of oxidation led to stronger LCL proportional tohigher LOOH concentration.

The above presented results show that calibration with LOOHstandard allowed directly obtain the concentrations of hydroper-oxides formed in the investigated system.

3.5. LOOH formation in linseed oil during oxidation. LCL and PV study

To explore further applications of the LCL method it was appliedto determine the amount of hydroperoxides formed during oxida-tion of linseed oil at temperature 60 �C. The data presented in Fig. 6shows that during 2 h of oxidation the amount of hydroperoxideincreases almost linearly. At longer times LCL first reaches plateauand then slow decrease is observed. To explain observed results wehave to consider that after 2 h of such intensive thermal treatment,and according to lipid oxidation scheme (Porter, Caldwell, & Mills,

Fig. 2. The optimization studies of LCL intensity, expressed in LOOH concentration, in dependence on: (a) pH; (b) luminol concentration and (c) hemin concentration; insetshows the shapes of LCL kinetics with 2 � 10�6 M (sharp peak) and 5 � 10�8 M hemin in the sample. This last one was multiplied nine times to visualize the difference inobtained shapes. The concentrations of reaction mixture components were as given in legend of Fig. 1, except for varying components.

Fig. 3. Typical LCL signal obtained during measurements. The reaction mixturecontained in final volume of 3 ml: 100 pM of LOOH, luminol 0.1 mM, 5% methanol,0.025% TRITON X-100, buffer CAPS 0.1 M pH 10. The reaction started by addition ofhemin 5 � 10�6 M 30 s after the beginning of data acquisition. The lower curveshows emission in the absence of luminol, which, above 1 min, practicallyrepresents background emission before adding hemin to the sample.

Fig. 4. LCL of 13-HPODE versus absorbance measured at 234 nm, (conjugateddienes formation).

168 P. Rolewski et al. / Food Research International 42 (2009) 165–170

1995), the oxidation process progressed and decrease of doublewith conjugated dienes is expected leading to the decreased LCL.

The formed quasi-stationary state is observed at longer oxidationtime. We may also notice that the amount of the formed LOOHin case of linseed oil is one order of magnitude higher than as ob-served for linoleic acid. This is connected with the fact that linseed

Fig. 5. LOOH concentration measured after 0, 30, 60 and 90 min of oxidation oflinoleic acid emulsion at temperatures 50, 60 and 70 �C, colorized rectangles, and PCoxidized at 50 �C, black rectangle. All LOOH concentrations are the averages of fivereplicates and the error bar represents SD.

Fig. 6. LOOH concentration measured after 0, 6, 12, 18, 24 hours of oxidation of2 mg/ml linseed oil emulsion at 40 �C and after 336 hours of storage at closed, darkbottle at room temperature. All LOOH concentrations are the averages of fivereplicates and the error bar represents SD. The PV values, measured by iodometrictitration, are given for the samples.

P. Rolewski et al. / Food Research International 42 (2009) 165–170 169

oil contains mostly unsaturated fatty acids with 67% linolenic acid(18:3) and 11% linoleic acid (18:2). In such case the thermally in-duced hydroperoxide formation occur fast and come up in bigamount compared to linoleic acid itself. These above reported re-sults for linseed oil obtained with LCL method were compared withthe data obtained by PV method. The PV value calculated for thefresh sample was 1.1 and this value was practically constant dur-ing 2 h of the oxidation. The sample of linseed oil stored in closed,dark flask at room temperature after 2 weeks gave the PV value of1.6. Similar results describing small changes of PV value of linseedoil measured at early stages of the storage were reported by Rud-nik, Szczucinska, Gwardiak, Szulc,and Winiarska (2001). The differ-ence observed between the LCL method and PV method during firsthours of oxidation reflects the methodological differences con-nected first of all with the amount of the sample and number ofsteps necessary to obtain the final result. Interestingly, at longertimes, both methods gave similar information. This comparisonclearly shows that LCL assay gives information faster, with highersensitivity and in much simpler way compared to the six-stepselaborative iodometric titration PV method.

The LCL method like the other experimental methods poses itslimitations. The presence of the other food components like caro-tenes, flavones, polyphenols may obscure applicability of LCLmethod. However, due to its high sensitivity the concentration ofsubstrate is low and usually the sample is diluted so the presenceof the other food components would not influence the strong lumi-nol chemiluminescence. The presence of certain natural fluoro-phores, like riboflavin, with an absorption band located in theemission peak position of luminol may lead to energy transferand appearance of sensitized chemiluminescence may occur.

The presence of transition metal ions or enzymes, which willcompete with lipid hydroperoxides as oxidizing agents, may inter-fere with proper determination of hydroperoxides. The availablereferences show that the LCL method is very useful in the simplesystems like fats, oils and solutions of extracts from solids. In morecomplicated systems with the recognized problems of interferingsources the LCL method still poses an advantage as a fast and sim-ple method for preliminary estimation of oxidative status of thesample.

4. Conclusions

According to calibration curve the LCL method may detecthydroperoxides from 50 pM to 4 lM. However, care should be ta-ken regarding the range of LOOH concentration. The presented re-sults indicate that calibration curve should be prepared for eachtype of the sample being studied, however, for a given samplethe values obtained from the calibration curve are very accuratewith excellent sensitivity and wide linear dynamic range usingsimple instruments.

The presence in the sample the other components like carotenesor phenols, which interfere with absorption measurements, in caseof LCL do not influence the LOOH determination but calibration ineach case is required.

The described chemiluminescent method is not specific for fattyacids and lipid hydroperoxides but may be used to follow the otherenzymatic or non-enzymatic oxidation processes.

The main advantage of this direct method is that it provide sen-sitive, fast and simple assay for quantitative determination of thehydroperoxides formed in the sample during oxidation.

Despite existing limitations the LCL method gives adequate testbecause it reacts with most relevant species, i.e. hydroperoxides orthe other oxidizing agents what finally gives overall total antioxi-dant estimation of the sample.

Acknowledgments

This work was partially supported by research Grant fromPoznan University of Life Sciences and Grant N312 1410 33 fromthe Polish Ministry of Science and Higher Education.

References

Antolovich, M., Prenzler, P. D., Patsalides, E., Mcdonald, S., & Robards, K. (2002).Methods for testing antioxidant activity. Analyst, 127, 183–198.

Bunting, J. P., & Gray, D. A. (2003). Development of a flow injectionchemiluminescent assay for the quantification of lipid hydroperoxides. Journalof the American Oil Chemists’ Society, 80, 951–955.

Burkow, I. C., Moen, P., & Overbo, K. (1992). Chemiluminescence as a method foroxidative rancidity assessment in autoxidized marine oils. Journal of theAmerican Oil Chemists’ Society, 69, 1108–1111.

Chan, P., Cheng, J. T., Tsao, C. W., Niu, C. S., & Hong, C. Y. (1996). The in vitroantioxidant activity of trilinolein and other lipid-related natural substances asmeasured by enhanced chemiluminescence. Life Sciences, 59, 2067–2073.

Dauben, W. G., Lober, M., & Fullerton, D. S. (1969). Allylic oxidation of olefins withchromium trioxide pyridine complex. Journal of Organic Chemistry, 34,3587–3592.

Dobarganes, M. C., & Velasco, J. (2002). Analysis of lipid hydroperoxides. EuropeanJournal of Lipid Science and Technology, 104, 420–428.

170 P. Rolewski et al. / Food Research International 42 (2009) 165–170

Dodeigne, C., Thunus, L., & Lejeune, R. (2000). Chemiluminescence as a diagnostictool. A review. Talanta, 51, 415–439.

Frankel, E. N. (1991). Recent advances in lipid oxidation. Journal of the Science ofFood and Agriculture, 54, 495–511.

Henderson, D. E., Slickman, A. M., & Henderson, S. K. (1999). Quantitative HPLCdetermination of the antioxidant activity of capsaicin on the formation of lipidhydroperoxides of linoleic acid: A comparative study against BHT andmelatonin. Journal of Agricultural and Food Chemistry, 47, 2563–2570.

Jimenez, A. M., & Navas, M. J. (2002). Chemiluminescence methods (present andfuture). Grasas y Aceites, 53, 64–75.

Kondo, Y., Kawai, Y., Miyazawa, T., & Mizutani, J. (1993). Chemiluminescence–HPLCanalysis of fatty acid hydroperoxide isomers. Bioscience, Biotechnology, andBiochemistry, 57, 1575–1576.

Matthaeus, B., Wiezorek, C., & Eichner, K. (1995). Fast chemiluminescence methodfor detection of oxidized lipids. Fett Wissenschaft Technologie, 97, 327–329.

Miyazawa, T. (2000). Analysis of lipid hydroperoxide formation in food andbiological systems. Journal of Oleo Science, 49, 1377–1382.

Miyazawa, T., Fujimoto, K., Suzuki, T., & Yasuda, K. (1994). Determination ofphospholipid hydroperoxides using luminol chemiluminescence–high-performance liquid chromatography. Methods in Enzymology, 233, 324–332.

Miyazawa, T., Kunika, H., Fujimoto, K., Endo, Y., & Kaneda, T. (1995).Chemiluminescence detection of mono-, bis-, and tris-hydroperoxytriacylglycerols present in vegetable oils. Lipids, 30, 1001–1006.

Navas, M. J., & Jimenez, A. M. (1996). Review of chemiluminescent methods in foodanalysis. Food Chemistry, 55, 7–15.

Nogala-Kalucka, M., Kupczyk, B., Polewski, K., Siger, A., & Dwiecki, K. (2007).Influence of native antioxidants on the formation of fatty acid hydroperoxidesin model systems. European Journal of Lipid Science and Technology, 109,1028–1037.

Ohba, Y., Kuroda, N., & Nakashima, K. (2002). Liquid chromatography of fatty acidswith chemiluminescence detection. Analytica Chimica Acta, 465, 101–109.

Pettersen, J. (1994). Chemiluminescence of fish oils and its flavour quality. Journal ofthe Science of Food and Agriculture, 65, 307–313.

Porter, N. A., Caldwell, S. E., & Mills, K. A. (1995). Mechanisms of free radicaloxidation in unsaturated lipids. Lipids, 30, 277–290.

Roginsky, V., & Lissi, E. A. (2005). Review of methods to determine chain-breakingantioxidant activity in food. Food Chemistry, 92, 235–254.

Rudnik, E., Szczucinska, A., Gwardiak, H., Szulc, A., & Winiarska, A. (2001).Comparative studies of oxidative stability of linseed oil. Thermochimica Acta,370, 135–140.

Saito, H., & Nakamura, K. (1989). Effect of phospholipids in fish oils onchemiluminescence. Agricultural and Biological Chemistry, 53, 897–899.

Sanchez-Moreno, C., & Larrauri, J.-A. (1998). Main methods used in lipid oxidationdetermination. Food Science and Technology International, 4, 391–399.

Sedman, J., van de Voort, F. R., & Ismail, A. A. (1997). In R. E. McDonald & M. M.Mossoba (Eds.), New techniques and applications in lipid analysis (pp. 283–324).Champaign, IL: AOCS Press.

Simic, M. G, Jovanovic, S. V, & Niki, E. (1992). Lipid oxidation in food. Washington, DC:American Chemical Society (pp. 14–32).

Spiteller, P., Kern, W., Reiner, J., & Spiteller, G. (2001). Aldehydic lipid peroxidationproducts derived from linoleic acid. Biochimica et Biophysica Acta, 1531,188–208.

Spiteller, P., & Spiteller, G. (1997). 9-Hydroxy-10, 12-octadecadienoic acid (9-HODE)and 13-hydroksy-9-11-octadecadienoic acid (13-HODE): Excellent markers forlipid peroxidation. Chemistry and Physics of Lipids, 89, 131–139.

Vladimirov, Y. A., & Petrenko, Y. M. (1976). Determination of the mechanism ofaction of antioxidants in lipid systems from chemiluminescence in the presenceof ferrous oxide. Biofizika, 21, 424–427.

Wanasundara, U. N., Shahidi, F., & Jablonski, C. R. (1995). Comparison of standardand NMR methodologies for assessment of oxidative stability of canola andsoybean oils. Food Chemistry, 52, 249–253.

Waters, W. A. (1971). The kinetics and mechanism of metal-catalyzed autoxidation.Journal of the American Oil Chemists’ Society, 48, 427–433.

Wheatley, R. A. (2000). Some recent trends in the analytical chemistry in lipidoxidation. Trends in Analytical Chemistry, 19, 617–627.

Yasaei, P.-M., Yang, G.-C., Warner, C.-R., Daniels, D.-H., & Yuoh, K. (1996). Singletoxygen oxidation of lipids resulting from photochemical sensitizers in thepresence of antioxidants. Journal of the American Oil Chemists’ Society, 73,1177–1181.