Effect of phenolic compounds on the oxidative stability of ground walnuts and almonds

Abstract

Walnuts and almonds are mainly constituted by unsaturated fatty acids and, in consequence, are readily prone to oxidation reactions. These nuts are also rich in antioxidant substances such as polyphenols, which are mainly concentrated in the brown skins (BS) or seed coats. This study analyzes the effect of BS phenolic compounds on the oxidative stability of walnuts and almonds. For this reason, peeled and unpeeled ground nuts were comparatively evaluated after thermal treatments as well as after room temperature storage. Furthermore, volatile compound emission due to thermally induced oxidation of the ground nuts was analyzed using HS-SPME-GC-MS. Several secondary oxidation products were identified in the headspace, and volatile product formation was considerably larger in nuts with BS than in peeled ones. Consistently, the former showed higher malondialdehyde levels after heating which indicates a prooxidant effect of BS presence. In order to verify these findings, a water-in-oil microemulsion was used as a model system to assess the role of nut phenolic compounds on lipid oxidation promotion. The hydroperoxide and malondialdehyde content was determined, as a measurement of primary and secondary oxidation enhancement, respectively. A dose dependent prooxidant effect on linoleic acid thermal oxidation was also observed when BS polyphenolic extracts as well as pure phenolic compounds were added.

---

1 Introduction

Walnuts and almonds are worldwide consumed foods and frequently included as ingredients in the diet of numerous ethnical groups. Besides, it has been demonstrated that the regular consumption of such nuts exerts beneficial effects on health.¹ The main constituent of these nuts is oil, predominantly constituted by unsaturated fatty acids and hence, highly prone to oxidation reactions.²

Lipid oxidation is a multi-stage radical mediated process where the primary products are lipid hydroperoxides.³ Since these are relatively unstable, they can be further oxidized yielding diverse secondary products⁴ such as aldehydes, acids, hydrocarbons and oxoacids.⁵ Aldehydes can undergo classic auto-oxidation processes forming short-chain hydrocarbons, carboxylic acids and dialdehydes like malondialdehyde (MDA) as the main products.⁶ These volatile compounds not only impart undesirable flavors, but also decrease the nut nutritional quality and result in the formation of potentially toxic species.⁴ There are several factors which can promote lipid peroxidation. Among the endogenous agents, the enzyme lipoxygenase (LOX) is able to catalyze the oxidation of fatty acids containing a cis,cis-1,4-pentadiene system as specific substrates, like linoleic acid (LA).⁷ LOX occurrence in walnuts and almonds has been studied and the enzyme inactivation by heating has also been described.⁸ Even when LOX is not activated, lipid oxidation may take place by an auto-oxidation process but at a slower rate. Another important external factor capable of promoting the oxidation reactions of lipid-containing foods is thermal treatment in an aerobic medium, conditions which are very frequent during food preparation.

Walnuts and almonds are also rich in other bioactive molecules such as phenolic compounds. In a previous study, we have determined that hydrophilic phenolic compounds are concentrated in the nut brown skins (BS) or pellicles that cover the kernels as seed coats, whereas less polar phenolic compounds such as tocopherols are located in the oily matrix of the kernel or nut meat.⁹ BS represents a 4% w/w of the shelled seed. Numerous phenolic compounds have been found in nut BS, flavonoids and non-flavonoids being identified.¹⁰ Methanolic and aqueous extracts of nut BS are able to effectively scavenge free radicals and inhibit LOX-induced oxidation of LA.⁹ Other antioxidant mechanisms of flavonoids are also well known including metal chelation and antioxidant-enzyme activation.¹¹ On the contrary, the prooxidant effect of polyphenols has also been described in lipid systems in a dose-dependent manner in the presence of active metal ions.^12,13

Walnuts and almonds, consumed as shelled nuts, are ingredients of several processed foods frequently subjected to cooking. These nuts are used worldwide in cooked food preparation as crushed, ground, or powdered nuts. Their oxidative stability has previously been studied using whole nuts exposed to different treatments.^14,15 However, as far as we know, the effect of BS phenolic compounds on the oxidative stability of walnuts and almonds submitted to thermal treatments has not previously been reported. The main aim of this work was to compare the oxidative susceptibility of peeled and unpeeled ground nuts after heating and during room temperature storage in aerobic conditions in order to elucidate the role of the phenolic constituents in the oxidation reaction mechanism. Besides, a model system formed by a water-in-oil microemulsion was designed to mimic the nut matrix in order to isolate the main factors, and to simulate the interactions that occur when nut lipids are heated in the presence of BS phenolic compounds.

2 Experimental section

2.1 Materials

Walnuts (Junglans regia var. Franquette) and almonds (Prunus dulcis var. Amara) were purchased at a local market. Nuts were cracked and shelled. A portion was peeled by soaking in water in order to remove the BS and afterwards, the material was dried under nitrogen. Peeled and unpeeled nuts were cooled at −20 °C and ground before analysis into a fine powder (particle size approx. 0.5 mm) using a coffee grinder for 1 min.

Hexanal and acetophenone were purchased from Sigma-Aldrich (Saint Louis, MO), pentanal from Fluka (Steiheim, Germany) and 1,1,3,3-tetraethoxypropane (TEP), thiobarbituric acid (TBA), and trichloroacetic acid (TCA) from Merck (Buenos Aires, Argentina). A total amount of 37 saturated and unsaturated fatty acids were used as reference compounds (Supelco 37 FAME Mix) and were purchased from Sigma (Newport, MA). Tween 20 (polyoxyethylene-sorbitan monolaurate) and catechin hydrate were purchased from Aldrich; linoleic acid was obtained from Riedel de Häen; sodium bis(2-ethylhexyl) sulphosuccinate or docusate sodium salt (AOT) and Folin–Ciocalteu’s phenol reagent were purchased from Sigma (Sigma-Aldrich, Buenos Aires, Argentina); GA was obtained from Anedra (Buenos Aires, Argentina) and quercetin was provided by Parafarm, (Buenos Aires, Argentina).

2.2 Methods

2.2.1 Fatty acid characterization of nut oils. Walnut and almond oil extraction was carried out using a cold extraction method with hexane according to Miraliakbari and Shahidi’s report¹⁶ with minor modifications. All extractions were performed in triplicate. After extraction, oil samples were stored at −20 °C under nitrogen. Fatty acids were transformed into their corresponding methyl esters before GC analysis. Methyl undecanoate (C:11) was used as an internal standard.¹⁷ Results are expressed in mg of fatty acid per 100 mg of extracted oil.

2.2.2 Phenolic content in walnut and almond extracts. Phenolic extract preparation: approximately 2 g of whole nuts or 0.5 g of their BS was mixed with 50 mL of ultra-pure water and stirred for 45 min, then, filtered under vacuum and centrifuged at 10 000 rpm for 10 min. The supernatant was recovered and the total phenolic compound content was determined using spectrophotometry according to the Folin–Ciocalteu method¹⁸ and the phenolic extract concentration was expressed as μg GAE per mL. This solution was used in model-system experiments.

2.2.3 Determination of volatile secondary products by HS-SPME-GC-MS. Amounts of 5 g of ground walnuts and almonds (previously peeled or unpeeled) were placed in round-bottom flasks with hermetic lids. The samples were heated for 24 h at 100 °C under controlled conditions. Appropriate amounts of acetophenone were added to the flasks 20 min before finishing the heating time as internal standards. Volatile compounds were collected from the head-space by SPME using a 50/30 μm divinylbenzene-carboxen-polydimethylsiloxane stable flex fiber. The DVB carboxen/PDMS fiber (Supelco, Spain) was exposed to the head-space for the last 10 min of thermal treatment. After the volatile compound extraction, the fiber was inserted directly into the GC injector where it was kept during the run time. A Thermo Scientific gas chromatograph equipped with a mass spectrometer detector, split injector and a 30 m × 0.25 mm × 0.25 μm Thermo Scientific TR-5MS capillary column were used in the analysis. The mass range scanned was 50–600 m/z. The ionization mode was electron impact at 70 eV. The injector temperature was 200 °C, the injector split mode was used with a rate of 50 mL min⁻¹, while the total and column flows were 10 and 1 mL min⁻¹, respectively. The initial temperature was 37 °C and this was held for 5 min; then, it was increased at 10 °C min⁻¹ to 140 °C and held for 5 min; finally, it was increased at 10 °C min⁻¹ to 240 °C and held for 5 min. Peak identification was carried out by comparison of the mass spectra with those contained in NIST library Mass Spectral Search Program, version 2.0. In order to quantify the volatile compounds formed, pure authentic samples were used for calibration purposes. Individual response factors were calculated with the internal standard to quantify the oxidation products. Results are expressed as headspace concentration of volatile compounds in μg per g ground nut.

2.2.4 Determination of secondary products by a TBARS method. Samples of 0.6 g of ground nuts (previously peeled or unpeeled) were placed into hermetically capped flasks and heated in a water bath for different times. A series of seven samples were kept at 60 °C for 10 and 30 min, at 70 °C for 10, 30 and 60 min, and at 100 °C for 30 and 60 min. The thiobarbituric acid reactive substance (TBARS) content was determined in the nuts after treatment.¹⁹ Controls were kept at room temperature (25 °C) for 10, 30, 60 and 120 min. An experiment kept at room temperature immediately after nut grinding was considered as a blank for all the determinations. The procedure to monitor the formation of TBARS was carried out by mixing an amount of 5 mL of 0.3 M TCA solution with 0.6 g of sample. The mixture was stirred and centrifuged for 15 min at 10 000 rpm. A 2 mL supernatant aliquot was mixed with an equal volume of 34 mM TBA solution and heated at 100 °C for 1 h. Finally, the solution was cooled in a cold-water bath containing ice to stop the reaction. After reaching room temperature, the absorbance was measured at 532 nm using 700 nm values to correct the baseline. As a malondialdehyde (MDA) precursor, TEP was used for calibration purposes. Results are expressed as MDA equivalents in mg kg⁻¹ of sample. In order to verify if TBA is able to react with other carbonyl compounds different to MDA, 6 mM pentanal and hexanal pure solutions were also analyzed.

2.2.5 Thermal induced oxidation of LA in a model system. To mimic a low water content food rich in lipids, a water-in-oil model system was designed as AOT/isooctane/water reverse micelles containing LA. Stock solutions of AOT and LA were prepared both in isooctane in capped tubes to final concentrations of 25 and 8 mM, respectively. A fixed aliquot of the aqueous media was added with a microsyringe. The total water content in the system was expressed as the mole ratio between water and surfactant:

W = [H₂O]/[AOT]

The W value was fixed at 62 according to our previous report to reach the humidity level in nuts.⁹ In order to modify the phenolic concentration from 0 to 6 μg GAE per mL for walnuts, and for almonds from 0 to 0.3 μg GAE per mL, the water–aqueous extract ratio was varied in the aqueous portion of the reverse micelle system. The mixtures were incubated in sealed vials at 60 and 100 °C for time lengths of 30 min to 15 h. Subsequently, they were cooled in a cold-water bath to stop the reaction in order to analyze the oxidation enhancement. The same procedure was carried out whilst adding to the system aqueous solutions of GA, catechin and quercetin in order to compare the nut aqueous extract behavior with those of pure compounds. Phenolic concentrations varied between 0 and 6 μg mL⁻¹. No precedents have been found on using temperatures close to 100 °C in reverse micelles systems formed by AOT/isooctane/water; thus, the boiling point of the water in the oil system and its stability were previously determined. The determination was carried out by putting the glass tubes containing the micelle system in a vaseline bath. A porous plate was previously added to the tubes to prevent overheating, and the instrument used in this determination was a thermocouple. The boiling point was 103.0 ± 0.1 °C, which was higher than the working temperature.
2.2.5.1 Determination of primary oxidation products in the model system. Hydroperoxide formation was evaluated spectrophotometrically at 234 nm using a reduced optical-path cuvette, with 0.2 cm of path length. Hydroperoxide concentrations were calculated using the Lambert–Beer Law and a molar extinction coefficient of LA hydroperoxides at the same wavelength in reverse micelles of ε = 23 005 M⁻¹ cm⁻¹. To calculate this value, a given oxidized LA sample was prepared in Tween 20 aqueous micelles and measured at 234 nm. The values were converted to hydroperoxide concentrations using the Lambert–Beer Law, taking ε as 25 000 M⁻¹ cm⁻¹. This value is the molar extinction coefficient of LA hydroperoxides reported in direct micelles.²⁰ The same LA sample was used to prepare the AOT/isooctane reverse-micelle system and the molar extinction coefficient in this system was calculated by measuring the absorbance at 234 nm.
2.2.5.2 Determination of secondary products in the model system. After hydroperoxide measurements, the secondary oxidation products were determined using a TBARS method as previously described. Amounts of 2 mL of 0.3 M TCA solution and 34 mM TBA were added to the model system. The mixture was homogenized and heated at 100 °C in a water bath for 1 h. Then, the reaction was stopped by cooling in a cool-water bath to reach room temperature. Subsequently, the mixtures were centrifuged at 10 000 rpm for 10 min at 20 °C in order to separate the aqueous and organic phases. The former was used to measure the absorbance at 532 and 700 nm. Results are expressed as μg MDA per mg LA.

2.2.6 Statistical analysis. All analyses were performed in triplicate. Results are expressed as the mean of the values obtained for each sample (n = 3). Analysis of variance models (one-way ANOVA) was performed using Infostat computing software. A multiple comparison procedure using Duncan’s test was applied to determine which means were significantly different at the P < 0.05 confidence level.

3. Results and discussion

3.1 Fatty acid composition of nut oils

Walnut and almond oils were obtained through hexane extraction in order to determine their fatty acid profiles by gas chromatography (GC). Table 1 shows the fatty acid composition of walnut and almond oils. In the case of walnuts, LA was the major component, followed by oleic, linolenic, palmitic, and stearic acids. In contrast, oleic acid was the major component of almonds followed by LA, palmitic, stearic, and palmitoleic acids. No significant differences in the oil composition were observed between peeled and unpeeled nuts (data not shown). According to these results, walnut oil is richer in poly-unsaturated fatty acids (PUFA) than that of almonds, while the latter is richer in mono-unsaturated fatty acids, which is in agreement with previous reports.^2,14 Consistently, a lower oxidative stability has been found for walnut oil than those of other nut oils measured by a Rancimat method, with walnut and almond induction times being 4.7 and 21.8 h, respectively.²¹

Table 1 Fatty acid composition of walnut and almond oils (mg per 100 mg oil)^a

	Palmitic acid (16 : 0)	Palmitoleic acid (16 : 1, n − 7)	Stearic acid (18 : 0)	Oleic acid (18 : 1, n − 9)	Linoleic acid (18 : 2, n − 6)	Linolenic acid (18 : 3, n − 3)
a Values indicate the mean ± standard deviation of 3 replicates; n.d.: not detected.
Walnuts	6.7 ± 0.5	0.23 ± 0.01	2.3 ± 0.1	21 ± 2	57 ± 7	12 ± 2
Almonds	6.7 ± 0.3	0.49 ± 0.03	1.4 ± 0.1	68 ± 9	24 ± 3	n.d.

---

3.2 Thermally induced oxidation of nuts ground with and without BS

Concerning the oxidative stability of walnuts and almonds, different experiments were carried out by heating portions of ground peeled and unpeeled nuts in aerobic conditions to evaluate the BS effect. The oxidative status after each treatment was evaluated by analyzing the volatile compound emission by SPME-GC-MS and the TBARS formation. A control experiment was carried out by keeping the material at room temperature for the same time period.

3.2.1 Analysis of volatile oxidation products. The volatile compound emission produced as a consequence of secondary oxidation was taken as one of the parameters to evaluate the changes in the oxidation enhancement according to BS presence. Fig. 1 shows the typical chromatograms corresponding to the volatile fraction of the oxidation products formed after the thermal treatment of ground walnuts and their BS for 24 h at 100 °C. A complex mixture of volatile products was observed. Several compounds were found and identified in the headspace of the peeled and unpeeled nuts after heating. Significantly different behaviors were observed between the peeled and unpeeled nuts. Experiments performed by heating exclusively the BS taken from walnuts and almonds, indicate that this material did not produce the emission of the volatile compounds in the absence of the lipid material.

Fig. 1 Chromatographic profiles of the volatile oxidation products of ground walnuts after thermal treatment at 100 °C for 24 h as analyzed by SPME-GC-MS. (a) Walnuts with brown skin. (b) Walnuts without brown skin. (c) Walnut brown skin. Peak identification: (1) pentanal, (2) hexanal, (3) 2-heptenal, (4) 2,4-heptadienal, (5) 2-octenal, (6) acetophenone, (7) nonanal, (8) 2-nonenal, (9) 2,4-decadienal, (10) 2-undecenal, (11) 3-decen-1-ol, IS: internal standard.

Table 2 clearly indicates that BS induces oxidation product formation in walnuts as well as in almonds, with volatile compound levels in the headspace being higher in walnuts than in almonds. In the case of ground walnuts with skin, several compounds were found, the major one being 2,4-decadienal, followed by 3-decen-1-ol, 2-octenal, 2-heptenal, 2-undecenal, 2,4-heptadienal, hexanal, nonanal, 2-nonenal and pentanal. By contrast, the major product in walnuts without skin was hexanal, followed by 2,4-decadienal and 3-decen-1-ol. The total volatile content in the headspace was 183 and 5762 μg g⁻¹ for ground walnuts without and with BS, respectively after 24 h of heating.

Table 2 Brown skin effect on the emission of volatile oxidation products of ground walnuts and almonds after thermal treatment^a

Peak number	Volatile oxidation products	Headspace concentration (μg per g ground nut)
		Walnuts with skin	Walnuts without skin	Almonds with skin	Almonds without skin
a Values indicate the mean ± standard deviation of 3 replicates; n.d.: not detected. Detection limit = 3 μg in the headspace.
1	Pentanal	88 ± 1	n.d.	36 ± 2	n.d.
2	Hexanal	381 ± 8	84 ± 3	261 ± 15	26 ± 2
3	2-Heptenal	452 ± 35	n.d.	143 ± 9	n.d.
4	2,4-Heptadienal	354 ± 42	n.d.	n.d.	n.d.
5	2-Octenal	576 ± 47	n.d.	176 ± 2	n.d.
6	Nonanal	142 ± 17	n.d.	142 ± 2	n.d.
7	2-Nonenal	105 ± 5	n.d.	23.2 ± 0.2	n.d.
8	2,4-Decadienal	2478 ± 379	73 ± 10	24 ± 1	22 ± 3
9	2-Undecenal	385 ± 15	n.d.	124 ± 7	16 ± 14
10	3-Decen-1-ol	799 ± 12	26 ± 2	85 ± 5	23 ± 4

---

In ground almonds with BS, the volatile compounds found in the headspace were hexanal, 2-octenal, 2-heptenal, nonanal, 2-undecenal, 3-decen-1-ol, pentanal, 2,4-decadienal and 2-undecenal. In ground peeled almonds, the compounds detected were hexanal, 3-decen-1-ol, 2,4-decadienal and 2-undecenal. The total volatile content in the headspace was 87 and 1014 μg g⁻¹ for ground almonds without and with BS, respectively after 24 h heating.

The formation of 2,4-decadienal, the major volatile aldehyde found in the oxidized walnut headspace, was markedly induced by the BS presence; being 34 times higher in walnuts ground with peel than in the ones previously peeled.

Hexanal levels in the headspace of walnuts with skin showed values 4.5 times higher than in walnuts without skin. Furthermore, almonds with skin showed emissions 10 times higher than for almonds without skin. Pentanal emission was exclusively detected in nuts with skin, being more than 2 times higher in walnuts than in almonds. The presence of 2-heptenal, 2-octenal, 2-nonenal and nonanal was detected only after heating nuts with skin, with the values being between 3 and 4.5 times higher in walnuts than in almonds. In the case of nonanal, the values were similar in both nuts. On the other hand, 3-decen-1-ol and 2,4-heptadienal were detected mainly in walnuts with skin.

The development of low-molecular weight aldehydes during storage of fat-rich food can occur as a consequence of hydroperoxide decomposition from lipid oxidation reactions.^5,22 The possible scissions of primary oxidation compounds take place to give different volatile products such as hexanal, 2,4-decadienal, 3-nonenal, and 2-heptenal.⁶ Hexanal has been reported as the main oxidation product from LA oxidation, which has the highest correlation with the sensorial perception of specific rancidity in walnuts.²³ Pentanal has been detected as one of the major volatile constituents of walnut aroma and its presence has been attributed to LA breakdown. Pentanal is then considered as an important contributor to the typical walnut aroma.²⁴ Hexanal formation in walnuts stored for a period of 12 months packaged in different conditions has been reported,²⁵ and the hexanal values from dark storage and under light exposure varied between 0.0285 and 36 mg kg⁻¹, respectively. In a similar study, the effect of packaging on the hexanal content among other parameters of almonds stored under different conditions for 12 months has been investigated, and values ranging between 0.0285 and 4.88 mg kg⁻¹ were obtained.¹⁵ As previously mentioned, hexanal and pentanal as well as 1-pentanol and 1-hexanol have been identified as the major volatile compounds in the aroma headspace of walnuts from different provenances, especially in those which contain high PUFA levels.²⁴ On the other hand, these volatile carbonylic substances formed as secondary oxidation products from unsaturated fatty acid breakdown can be toxic at high concentrations and may be associated with the development of off-flavors, making the products undesirable to consumers.^26,27

3.2.2 Secondary oxidation products by the TBARS method. TBARS levels were analyzed in ground peeled and unpeeled nuts treated at different temperatures and time lengths. Nut samples kept at room temperature for the same time periods were taken as controls. The results corresponding to the analysis of the TBARS content in these nut samples are shown in Table 3. Significant differences were observed in the oxidation degrees of the treated samples depending on BS presence. It is important to emphasize that this is the first study where the BS effect on the oxidative stability of ground nuts is taken into account.

Table 3 Brown skin effect on the oxidation levels of ground walnuts and almonds after thermal treatment and room temperature storage^a

Treatment		TBARS (mg MDA per kg nut)
Temperature (°C)	Time (min)	Walnuts		Almonds
		Without skin	With skin	Without skin	With skin
a Mean ± SD of triplicate. Values in a row for walnuts or almonds followed by different letter indicate significant differences (P < 0.05) between the peeled and unpeeled nuts.
25	10	3.5 ± 0.1^a	2.2 ± 0.1^b	2.5 ± 0.1^a	1.7 ± 0.3^b	Antioxidant
	30	3.8 ± 0.1^a	2.8 ± 0.2^b	2.9 ± 0.3^a	1.4 ± 0.1^b
	60	3.9 ± 0.2^a	2.9 ± 0.1^b	2.5 ± 0.4^a	1.6 ± 0.1^b
	120	5.1 ± 0.2^a	3.2 ± 0.4^b	2.93 ± 0.03^a	1.72 ± 0.03^b
60	10	3.5 ± 0.1^b	6.0 ± 0.4^a	3.0 ± 0.2^a	1.6 ± 0.3^b
	30	4.4 ± 0.2^b	7.5 ± 0.1^a	4.2 ± 0.1^a	2.9 ± 0.2^b
70	10	3.7 ± 0.1^b	6.5 ± 0.1^a	3.2 ± 0.2^a	1.8 ± 0.1^b
	30	7.6 ± 0.1^b	10.5 ± 0.1^a	4.4 ± 0.1^b	6.9 ± 0.1^a	Prooxidant
	60	8.5 ± 0.3^b	11.5 ± 0.1^a	5.6 ± 0.3^b	8.4 ± 0.3^a
100	30	8.1 ± 0.1^b	10.9 ± 0.1^a	5.1 ± 0.2^b	7.2 ± 0.2^a
	60	9.7 ± 0.1^b	12.1 ± 0.5^a	6.5 ± 0.3^b	8.7 ± 0.4^a

---

After storage at 25 °C, the oxidative levels of ground peeled nuts were higher than those nuts ground with their peels even for up to 120 min. This indicates a protective effect of the BS in lipid auto-oxidation or enzymatically induced lipoxidation. In contrast to this antioxidant effect, walnuts with BS showed higher TBARS levels than the peeled ones after heating at 60, 70 and 100 °C, whereas almonds showed behavior similar to walnuts only at 70 and 100 °C. In the case of almonds, the protective effect remained after heating at 60 °C for 10 and 30 min, although at 70 °C this remained only for 10 min. However, BS produced a pro-oxidant effect under harsher conditions, indicating that the nut skins exerted a pro-oxidant effect when thermally induced lipoxidation was the main mechanism.

After thermal treatment, the MDA equivalents determined in walnuts were larger, in all cases, than those of almonds. In both cases, the nuts ground with peels presented larger amounts than the peeled nuts after thermal treatment. The increases in TBARS levels were between 25 and 38% in walnuts, whilst they were between 34 and 57% in almonds, for the different temperatures and time lengths assayed, and were the highest for 70 °C and 30 min.

On the other hand, the walnut control samples also presented higher MDA values than the almond ones. However, differences were found between the peeled and unpeeled nuts in the control samples. The MDA values of ground peeled nuts were higher than those of nuts ground with their peels; the BS protective effect being of 37 and 41% for walnuts and almonds, respectively.

The BS effect on TBARS levels is consistent with the observed variations in volatile compound formation ascribed to the BS presence. Control experiments were also performed with pure pentanal and hexanal using the TBARS method conditions. It was verified that these carbonylic compounds do not interfere with MDA in this analysis. This indicates that both parameters have independently a good correlation with lipid oxidation.

The TBARS values in raw unpeeled almonds varied in a range between 0.65 and 1.8 mg MDA per kg.⁴ In the case of raw unpeeled walnuts, the TBARS values measured varied between 0.2 and 11 mg MDA per kg of walnut stored for a period of 12 months packaged in different conditions, when they were stored in the dark and exposed to light, respectively.²⁵

Nut heating increases the rate of all single reactions involved in the oxidation process. Furthermore, the concentration of lipid oxidation markers in nuts increases dramatically from room temperature to storage temperatures above 40 °C.²⁸ High temperatures and light exposure during storage dramatically increase lipid oxidation in almonds; besides, the presence of oxygen has been recognized as one of the most significant extrinsic factors affecting lipid oxidation of nuts.¹⁵ Moreover, roasting of walnut kernels was found to increase the peroxide value and conjugated dienoic acid value of the oil, parameters which were used as indicators of their oxidative stability.²⁹ Walnuts and almonds are foods rich in polar and nonpolar phenolic compounds with recognized antioxidant ability as free radical scavengers, but also as LOX inhibitors. In both nuts, the major fraction corresponds to polar compounds concentrated in the BS, and non-polar ones located in the oil are minor constituents, including tocopherols and tocotrienols.

We have previously reported the phenolic content of whole nuts and of their BS.⁹ For both nuts, walnuts and almonds, the BS and whole walnut extracts presented approximately 10 times higher phenolic content than those of the almond ones. Besides, the BS extracts presented noticeably higher content values than the whole nut extracts, indicating that phenolic compounds are concentrated in the nut BS.

Numerous phenolic compounds have been found in nut BS. Flavonoids and non-flavonoid phenolic compounds have been identified in almond skins.¹⁰ Among the non-flavonoid compounds, protocatechuic, vainillic and p-hydroxybenzoic acids have been found, while among the flavonoids, several flavonols, flavanols, dihydroflavonols and flavanons have been reported. Besides, the presence of different kinds of proanthocyanidins has also been found.³⁰ Several flavonoids and phenolic acids have been found exclusively in almond BS;³¹ among them, quercetin, quercetin-3-O-galactoside, quercetin-3-O-rutinoside, quercetin-3-O-glucoside, dihydroxykaempferol, eriodictyol, kaempferol and p-hydroxy-benzoic acid can be mentioned. Many other compounds such as catechin and epicatechin have been found in the kernel, although they are mainly located in the almond BS. The potential use of almond skins as a source of phytochemicals has driven several studies of their composition.³² In walnut BS, compounds such as ellagic acid, gallic acid and methyl gallate have been identified.^33,34 These compounds can be present as polymers and bound to sugars, and constitute a major fraction of nut phenolic compounds. Moreover, sixteen phenolic constituents have been found in walnut extracts, three of them were hydrolysable ellagitannins named glansrins and thirteen are well known polyphenolic compounds.³⁵ Besides, walnuts presented a noticeably higher antioxidant activity and larger phenolic content than almonds.⁹ Walnut oil has been demonstrated to present higher levels of both non-polar phenolic compounds and polyunsaturated fatty acids than the almond oil. However, the BS protective effect towards thermally induced lipid oxidation was not observed, on the contrary to our initial expectations. Moreover, the thermal treatment of nuts ground with their peels indeed induced the volatile carbonyl compound formation. By contrast, the differences found in the control samples between the peeled and unpeeled nuts can be attributed to a protective effect of phenolic compounds present in the BS towards enzymatic oxidations. Taking into account that phenolic compounds are good LOX inhibitors, these results are new evidence that BS phenolic compounds prevent LOX-induced lipoperoxidation, which is the major reaction pathway taking place at room temperature.

3.3 Effect of BS phenolic extract on thermally induced lipid oxidation in a model system

In order to isolate the main factors controlling the processes and simulating the interactions that occur when peeled and unpeeled nuts are heated, a model system formed of a water-in-oil microemulsion was proposed. Nuts as dried foods have a 2–4% moisture content but are rich in lipids. Since the lipids are located in a hydrophobic region called lipid droplets or oily bodies and water is located in pools or “pockets”, a liquid model was chosen to mimic the nut microstructure. Reverse micelles of AOT/isooctane/water at W = 62 allow mimicking of such a low water content dispersed in a lipophilic medium enriched in LA. In the aqueous core, phenolic extracts of nut BS at different concentration levels were assayed. Samples were heated for 15 h to monitor primary and secondary oxidation products.

3.3.1 Primary oxidation products. The oxidation enhancement after 15 h of thermal treatment at 60 and 100 °C was analyzed in terms of hydroperoxide formation. In systems containing walnut and almond BS extracts, heating at 60 °C did not form hydroperoxides in the absence of BS phenolic compounds, or in their presence. LA thermally induced oxidation at 100 °C yielded a peroxide level of 6 ± 0.5 μg per mg LA after 15 h in the absence of phenolic compounds. The presence of nut phenolic extract markedly increased the formation of oxidation products giving as the highest peroxide levels 64.5 ± 4.1 and 21.3 ± 1.2 μg per mg LA, at 4.4 and 0.05 μg GAE per mL for walnut and almond extracts, respectively. After reaching the maximum level of hydroperoxides, a decrease was observed for the samples of both kinds of extracts, as a function of phenolic concentration. The highest values of hydroperoxides in the systems containing walnut extracts are consistent with their higher LA content. The fact that non-detectable levels of hydroperoxides were found at higher phenolic concentrations, suggests that the oxidation-promotion effect of the BS extract shifted the reaction to more advanced stages leading to secondary oxidation products. Room temperature experiments were also carried out to simulate those conditions where enzymatically induced oxidation is the major pathway taking place. An aliquot of LOX aqueous solution was added to the reverse micelle system containing LA to induce oxidation. The hydroperoxide formation was evidenced by the absorbance increase at 234 nm due to the conjugated diene formation. When the same experiment was performed with BS extract addition, an abundant precipitate was formed which impeded spectrophotometric measurement. This fact, far from being negative, was very positive since it demonstrated that BS phenolic compounds have the ability to interact with LOX. The association between tannins and proteins has been widely studied and reported. However, these studies were carried out in solution whereas the results presented in this work provide evidence that this interaction can take place in a food matrix and this explains the protective effect of BS phenolic compounds towards enzyme-induced oxidation observed in ground nuts stored at room temperature.

3.3.2 Secondary oxidation products. After hydroperoxide determination in the model system, TBARS levels were also analyzed as a function of phenolic concentration. This investigation was carried out in order to verify if the decrease in hydroperoxide values previously observed was related to the secondary oxidation product formation under the experimental conditions. Samples containing BS extracts showed higher MDA levels than samples without BS extracts after thermal treatment at 100 °C for 15 h. Significant differences were observed in the oxidation levels of the treated samples, depending on their phenolic compound content. The maximum yield of MDA was found in the samples with walnut extract, this value being 31% higher than those of the samples with almond extract. The increment in MDA level was 53% in the samples with walnut extract, and 33% in the samples with almond extract. This increase in MDA level was coincident with the decrease in the hydroperoxide values previously observed.

Fig. 2A and B and 3A and B show the effect of phenolic concentration on LA oxidation as the variation of the primary and secondary oxidation products after thermal treatment in the model system. The results clearly indicate a prooxidant action of the phenolic compounds during 100 °C treatment of the reverse micelle system. This observation is in good agreement with a previous report where the prooxidant effect of polyphenols over an oxidisable substrate has been reported to be promoted by metal ions.³⁶ Besides, Zhou and Elias¹³ have recently reported the change from antioxidant to prooxidant activity of phenolics in direct micelles in the presence of active metal ions.

Fig. 2 Effect of walnut brown skin phenolic extracts on LA oxidation in the water-in-oil micelle systems after 15 h of thermal treatment. (A) Primary oxidation products; (B) secondary oxidation products.

Fig. 3 Effect of almond brown skin phenolic extracts on LA oxidation in the water-in-oil micelle systems after 15 h of thermal treatment. (A) Primary oxidation products; (B) secondary oxidation products.

To the best of our knowledge, this is the first report on the promotion of lipid oxidation by phenolic compounds in a model system as well as its correlation with the natural food behavior in the absence of active metal ions.

3.4 Effect of pure phenolic compounds on thermally induced lipid oxidation in a model system

Three different pure phenolic compounds were used to study their antioxidant/prooxidant behavior, and compare it with those of the nut aqueous extracts. They were chosen since as was previously cited, GA is an important phenolic acid present in walnut BS,^33,34 and both flavonoids catechin and quercetin are main constituents of almond BS.³¹

3.4.1 Primary oxidation products. LA thermally induced oxidation yielded an initial peroxide level of approximately 10 μg per mg LA in the absence of phenolic compounds. The three pure substances employed exerted a prooxidant action under the working conditions. After 15 h at 100 °C, an increase in the peroxide levels was observed, followed by a decrease in all cases. The highest peroxide levels were 60.8 μg mg⁻¹ in the presence of 4.4 μg mL⁻¹ of GA, 51.2 μg per mg LA with 4.4 μg mL⁻¹ of catechin and 80.4 μg per mg LA with 2.2 μg mL⁻¹ of quercetin. The behavior observed for the three compounds studied was consistent with that observed in the systems containing aqueous extracts. Fig. 4 shows the variation of LA primary oxidation products in a dose dependent manner, after thermal treatment in the model system. The results clearly indicate a prooxidant action of the phenolic compounds during 100 °C treatment of the reverse micelle system.

Fig. 4 Effect of pure phenolic compounds on the primary oxidation products of LA oxidation in the water-in-oil micelle systems after 15 h of thermal treatment.

3.4.2 Secondary oxidation products. The systems containing phenolic substances showed that TBARS levels increased as a function of their concentration. In the same way as for the aqueous extracts, the decrease in the hydroperoxide values was related to the secondary oxidation product increase. The maximum oxidation levels were 0.32, 0.3 and 0.41 μg MDA per mg AL in the presence of 6 μg mL⁻¹ of GA, catechin and quercetin, respectively. Fig. 5 shows the effect of phenolic concentration on LA oxidation as the variation of the secondary oxidation products after thermal treatment in the model system. The results clearly indicate a prooxidant action of the phenolic compounds during 100 °C treatment of the reverse micelle system. Summarizing, the model system designed allowed corroboration of the prooxidant effect on LA of the phenolic compounds after thermal treatment, in the same way as that observed in the food matrix.

Fig. 5 Effect of pure phenolic compounds on the secondary oxidation products of LA oxidation in the water-in-oil micelle systems after 15 h of thermal treatment.

The main factors influencing the pro-oxidant activity of polyphenols in oil-in-water emulsions have been well described for a system containing iron ions.¹³

In our present study, polyphenol prooxidant action takes place in reverse micelles, as the micro-heterogeneous systems, which describes well what takes place in the food matrix of nuts. As far as we know, this is the first report evidencing the prooxidant action of phenolic compounds in systems free of active metal ion addition.

The interaction of phenolic compounds with free radicals (like ROS) produces phenoxyl radicals as intermediates stabilized by resonance; further oxidation of these species leads to final degradation products of the phenolics. In the case of low free radical levels (few initiation events for a radical reaction), phenolics result in the breaking of the reaction chain, and due to the stability of the phenoxyl radical intermediates formed, this result leads to antioxidant behavior. However, prooxidant behavior can take place in a lipid rich system containing peroxide and hydroperoxide thermally unstable precursors of free radicals. When the production of these radicals is excessive due to heat induction, this leads to a redox imbalance (oxidative stress) and these phenoxyl radical species are able to promote the propagation of the radical chain reaction and thus, to induce the enhancement of the oxidation reaction in a dose dependent manner.

4 Conclusion

Walnuts are more susceptible to oxidation than almonds as a consequence of their lipid composition. Thermal treatment of nuts ground with their BS produced higher TBARS levels and a higher volatile emission than those of ground peeled nuts; hence, the presence of BS promotes lipid oxidation in ground walnuts and almonds subjected to short-term heating.

On the contrary, during room temperature storage of ground nuts, antioxidant behavior was found to be exerted by the BS. This action is ascribed to the inhibition of LOX induced oxidation by BS phenolic constituents. This protective ability was no longer effective when the nuts were heated, with the BS prooxidant action being more important than the antioxidant effect. The proposed water-in-oil model system validates these results; a dose-dependent prooxidant effect on LA thermal oxidation was observed after heating for the phenolic extract additions; this behavior was reverted in the LOX induced process at room temperature. Nut skin induces specific lipid oxidation reactions yielding volatile compounds.

This work shows the importance of peeling the nuts before grinding for culinary preparations that require cooking, to preserve not only their sensorial and nutritional characteristics but also their safety. On the other hand, for those recipes without cooking that include ground nuts, keeping their skins is important to prevent oxidation.

Abbreviations

BS	Brown skin
GA	Gallic acid
GAE	Gallic acid equivalents
GC	Gas chromatography
HS	Head space
LA	Linoleic acid
LOX	Lipoxygenase
MDA	Malondialdehyde
MS	Mass spectrometry
PUFA	Polyunsaturated fatty acids
ROS	Reactive oxygen species
SPME	Solid phase micro-extraction
TBARS	Thiobarbituric acid reactive substance
TCA	Trichloroacetic acid
TEP	1,1,3,3-Tetraethoxypropane

Acknowledgements

C. Salcedo acknowledges her postdoctoral fellowship from CONICET. This work was supported by CONICET, and CICYT-UNSE. We are sincerely grateful to Prof. Maria Pilar Almajano for her support with the SPME system.

References

1. A. P. Simopoulos, Health Effects of Eating Walnuts, Food Rev. Int., 2004, 20, 91–98.
2. M. Venkatachalam and S. K. Sathe, Chemical composition of selected edible nut seeds, J. Agric. Food Chem., 2006, 54, 4705–4714.
3. B. Halliwell and J. M. C. Gutteridge, Free Radicals in Biology and Medicine, Oxford University Press, London, 2nd edn, 1989.
4. R. Yaacoub, R. Saliba, B. Nsouli, G. Khalaf and I. Birlouez-Aragon, Formation of lipid oxidation and isomerization products during processing of nuts and sesame seeds, J. Agric. Food Chem., 2008, 56, 7082–7090.
5. P. Spiteller, W. Kern, J. Reiner and G. Spiteller, Aldehydic lipid peroxidation products derived from linoleic acid, Biochim. Biophys. Acta, 2001, 1531, 188–208.
6. E. Frankel, Volatile lipid oxidation products, Prog. Lipid Res., 1982, 22, 1–33.
7. J. M. López-Nicolás, M. Pérez-Gilabert and F. García Carmona, Eggplant lipoxygenase (Solanum melongena): product characterization and effect of physicochemical properties of linoleic acid on the enzymatic activity, J. Agric. Food Chem., 2001, 49, 433–438.
8. A. Buranasompob, J. Tang, J. R. Powers, J. Reyes, S. Clark and B. G. Swanson, Lipoxygenase activity in walnuts and almonds, LWT--Food Sci. Technol., 2007, 40, 893–899.
9. C. L. Salcedo, B. A. López de Mishima and M. A. Nazareno, Walnuts and almonds as model systems of foods constituted by oxidisable, pro-oxidant and antioxidant factors, Food Res. Int., 2010, 43, 1187–1197.
10. M. Monogas, I. Garrido, R. Lebrón-Aguilar, B. Bartolomé and C. Gómez-Cordovés, Almond (Prunus dulcis (Mill.) D. A. Webb) skins as a potential source of bioactive polyphenols, J. Agric. Food Chem., 2007, 55, 8498–8507.
11. K. E. Heim, A. R. Tagliaferro and D. J. Bobilya, Flavonoid antioxidants: Chemistry, metabolism and structure–activity relationships, J. Biochem., 2002, 13, 572–584.
12. L. R. Fukumoto and G. Mazza, Assessing antioxidant and prooxidant activities of phenolic compounds, J. Agric. Food Chem., 2000, 48, 3597–3604.
13. L. Zhou and R. J. Elias, Factors influencing the antioxidant and pro-oxidant activity of polyphenols in oil-in-water emulsions, J. Agric. Food Chem., 2012, 60, 2906–2915.
14. J. S. Amaral, S. Casal, J. A. Pereira, R. M. Seabra and B. P. P. Oliveira, Determination of sterol and fatty acid compositions, oxidative stability, and nutritional value of six walnut (Juglans regia L.) cultivars grown in Portugal, J. Agric. Food Chem., 2003, 51, 7698–7702.
15. S. F. Mexis, A. V. Badeka and M. G. Kontominas, Quality evaluation of raw ground almond kernels (Prunus dulcis): Effect of active and modified atmosphere packaging, container oxygen barrier and storage conditions, Innovative Food Sci. Emerging Technol., 2009, 10, 580–589.
16. H. Miraliakbari and F. Shahidi, Antioxidant activity of minor components of tree nut oils, Food Chem., 2008, 111, 421–427.
17. R. A. S. Sancho, F. A. Lima, G. G. Costa, L. R. B. Mariutti and N. Bragagnolo, Effect of annatto seed and coriander leaves as natural antioxidants in fish meatballs during frozen storage, J. Food Sci., 2011, 76, 838–845.
18. V. Singleton and J. Rossi, Colorimetry of total phenolics with phosphomolybdic-phosphotungstic acid reagents, Am. J. Enol. Vitic., 1965, 16, 144–158.
19. G. Zacheo, A. R. Capello, L. M. Perrone and G. V. Gnoni, Analysis of factors influencing lipid oxidation of almond seeds during accelerated ageing, LWT--Food Sci. Technol., 1998, 31, 6–9.
20. J. M. López-Nicolás, R. Bru and F. García-Carmona, Kinetic characteristics of the enzymatic conversion in presence of cyclodextrins: study of the oxidation of polyunsaturated fatty acids by lipoxygenase, Biochim. Biophys. Acta, 1997, 1347, 140–150.
21. S. Arranz, R. Cert, J. Pérez-Jiménez, A. Cert and F. Saura-Calixto, Comparison between free radical scavenging capacity and oxidative stability of nut oils, Food Chem., 2008, 110, 985–990.
22. S. Pastorelli, S. Valzacchi, A. Rodriguez and C. Simoneau, Solid-phase microextraction method for the determination of hexanal in hazelnuts as an indicator of the interaction of active packaging materials with food aroma compounds, Food Addit. Contam., Part A, 2006, 23, 1236–1241.
23. T. D. Crowe, T. W. Crowe, L. A. Johnson and P. J. White, Impact of extraction method on yield of lipid oxidation products from oxidized and unoxidized walnuts, J. Am. Oil Chem. Soc., 2002, 79, 453–456.
24. J. S. Elmore, I. Nisyrios and D. S. Mottram, Analysis of the headspace aroma compounds of walnuts (Juglans regia L.), Flavour Fragrance J., 2005, 20, 501–506.
25. S. F. Mexis, A. V. Badeka, K. A. Riganakos, K. X. Karakostas and M. G. Kontominas, Effect of packaging and storage conditions on quality of shelled walnuts, Food Control, 2009, 20, 743–751.
26. M. D. Guillén, N. Cabo, M. L. Ibargoitia and A. Ruiz, Study of both sunflower oil and its headspace throughout the oxidation process. Occurrence in the headspace of toxic oxygenated aldehydes, J. Agric. Food Chem., 2005, 53, 1093–1101.
27. E. Goicoechea, E. F. A. Brandon, M. H. Blokland and M. D. Guillén, Fate in digestion in vitro of several food components, including some toxic compounds coming from omega-3 and omega-6 lipids, Food Chem. Toxicol., 2011, 49, 115–124.
28. S. Tazi, F. Plantevin, C. Di Falco, A. Puigserver and E. H. Ajandouz, Effects of light, temperature and water activity on the kinetics of lipoxidation in almond-based products, Food Chem., 2009, 115, 958–964.
29. B. Vaidya and J. B. Eun, Effect of roasting on oxidative and tocopherol stability of walnut oil during storage in the dark, Eur. J. Lipid Sci. Technol., 2013, 115, 348–355.
30. C. H. Brieskon and R. Betz, Procyanidin polymers crucial to the structure of the almond seed coat, Z. Lebensm.-Unters. Forsch., 1998, 187, 347–353.
31. P. E. Milbury, C. Y. Chen, G. G. Dolnikowski and J. B. Blumberg, Determination of Flavonoids and Phenolics and Their Distribution in Almonds, J. Agric. Food Chem., 2006, 54, 5027–5033.
32. B. W. Bolling, G. Dolnikowski, J. B. Blumberg and C. Y. O. Chen, Quantification of Almond Skin Polyphenols by Liquid Chromatography-Mass Spectrometry, J. Food Sci., 2009, 74, 326–332.
33. K. J. Anderson, S. S. Teuber, A. Gobeille, P. Cremin, A. L. Waterhouse and F. M. Steinberg, Walnut polyphenolics inhibit in vitro human plasma and LDL oxidation, Biochem. Mol. Action Nutr., 2001, 131, 2837–2842.
34. Z. Papoutsi, E. Kassi, I. Chinou, M. Halabalaki, L. A. Skaltsounis and P. Moutsatsou, Walnut extract (Juglans regia L.) and its component ellagic acid exhibit anti-inflammatory activity in human aorta endothelial cells and osteoblastic activity in the cell line KS483, Br. J. Nutr., 2007, 99, 715–722.
35. T. Fukuda, H. Ito and T. Yoshida, Antioxidative polyphenols from walnuts (Juglans regia L.), Phytochem., 2003, 63, 795–801.
36. Y. Sakihama, M. F. Cohen, S. C. Grace and H. Yamasaki, Plant phenolic antioxidant and prooxidant activities: phenolics-induced oxidative damage mediated by metals in plants, Toxicology, 2002, 177, 67–80.

---