Use of Fatty Acid Metal Salts for Preventing Maillard Reaction-Driven Browning of Lecithins

Maillard four compounds. PEs heat lecithin oil sugars rearrangement ） food mixed SL, heat deterioration. Abstract: We previously reported that fluid soybean lecithin (SL) undergoes heat deterioration by the newly reported pseudo-Maillard rearrangement reaction. To inhibit heat deterioration, SLs were treated with metal silicates, such as magnesium silicate and calcium silicate. When soybean fatty acid was added to SL before treatment with calcium silicate, the color index after heating improved significantly as the acid value increased from 10 to 35 mg KOH/g. To elucidate the role of soybean fatty acid, calcium silicate treatment was carried out by adding several fatty acids to SL. Although saturated fatty acids had no effect on the heat deterioration of SL, unsaturated fatty acids were significantly more effective at inhibiting heat deterioration. SL. Based on these results, several fatty acid metal salts were added to confirm whether heat deterioration while heating SL could be inhibited. It was observed that the heat deterioration of SL could be inhibited with fatty acid metal salts, regardless of whether the fatty acids were saturated or unsaturated and whether the metal was monovalent, divalent, or trivalent. Therefore, in this study, we clarified that the heat deterioration of SL could be inhibited by adding fatty acid metal salts to SL. Among sodium stearate, calcium stearate, magnesium stearate, barium stearate, and aluminum tristearate, the divalent fatty acid metal salts had a stronger inhibitory effect on heat deterioration than the monovalent and trivalent salts.

desaccharified SL could not mitigate the heat deterioration, and a new solution is needed. Helmy et al. 1994 reported that the color of refined cottonseed oil was improved upon the addition of metal silicates in the degumming process 16 , although crude cottonseed oil is known to have high levels of lecithin. The improved color of refined cottonseed oil on addition of metal silicates suggests that the metal silicates may inhibit the above pseudo-Maillard reaction, but the mechanism of this inhibition is still unknown. In this paper, the effect of metal silicates on the inhibition of browning and the mechanism by which they inhibit the Maillard reaction of SL are reported. In addition, compounds related to the inhibition of the Maillard reaction are discussed.

Instruments and Measurements
The acid value, phospholipid composition, and color data were measured in accordance with sections 4.2.1, 4.3.3.1, and 2.2.1.1, respectively of the Standard Methods for the Analysis of Fats, Oils and Related Materials 1996 . The color index was calculated based on the color analysis using 10 B 1 Y 10 R. The calcium content was analyzed using inductively coupled plasma atomic emission spectrometry Shimadzu ICPS-7510. The UV data were obtained using a Shimadzu UV-2550 spectrophotometer. High-performance liquid chromatography HPLC was carried out with a Shimadzu LC-6AD pump equipped with a Shimadzu SPD-M10Avp diode array detector.

Materials
Fluid SL total phospholipid content: 57. 2  .05 g and soybean oil 17.95 g were mixed to make 25.0 wt. of the total phospholipid content, and heated at 60 for 30 min to easily dissolve SL in the oil. Next, metal silicate 3.00 g was added to 20.00 g of the mixture, and because of the improved fluidity of the oil-SL mixture, it was further heated at 60 for 30 min. Subsequently, the metal silicate was removed by pressure filtration, and the filtrate was vacuum-dried 50 , 0.09 MPa, 18 h . 2.3.1.2 Treatment with metal silicate after addition of soybean fatty acids SL 12.05 g , soybean oil, and soybean fatty acids were mixed to make 25.0 wt. of the total phospholipid content. The acid value of the mixture was adjusted from 15 to 35 mg KOH/g in 5 increments, and the total weight was kept at 30.00 g. Next, metal silicate 3.00 g was added to 20.00 g of the mixture, and because of the improved fluidity of the oil-SL mixture, it was heated at 60 for 30 min. Subsequently, the metal silicate was removed by pressure filtration, and the filtrate was vacuum-dried 50 , 0.09 MPa, 18 h . 2.3.1.3 Treatment with calcium silicate after addition of several fatty acids SL 12.05 g , soybean oil, and several fatty acids 0.023 mol each were mixed to make 25.0 wt. of the total phospholipid content. Next, calcium silicate 3.00 g was added to 20.00 g of these mixtures, and because of the improved fluidity of the oil-SL mixture, the mixtures were heated at 60 for 30 min. Subsequently, calcium silicate was removed by pressure filtration, and the filtrates were vacuum-dried 50 , 0.09 MPa, 18 h .

Heat deterioration tests
SLs prepared using the various processing methods described above were diluted with soybean oil to prepare mixtures with total phospholipid content of 1 wt. , and 6 g of these mixtures were heated in a test tube at 200 for 15 min.

Calcium concentration measurement
To measure the calcium concentration in the mixtures treated with calcium silicate, the samples 0.2 g were heated with sulfuric acid 2 mL and nitric acid 20 mL at 200 for 18 h in a perfluoroalkoxy alkane-lined reaction vessel 100 mL . After cooling, perchloric acid 2.5 mL was added to the solution and concentrated at 200 for 3 h. The volume was adjusted to 50 mL with 1 vol. hydrochloric acid aq. . Finally, the calcium content was analyzed using inductively coupled plasma atomic emission spectrometry ICP-AES, ICPS-7510 .

Inhibition of heat deterioration of SL with addition of fatty acid metal salts 2.4.1 Mixing SL with several fatty acid metal salts
Soybean oil and several fatty acid metal salts 8.5 10 5 mol were mixed to obtain a total weight of 9.83 g, and the mixture was heated at 130 for 10 min to easily dissolve the fatty acid metal salts in oil. After the mixture was cooled to 60 or below, SL 0.17 g was added to prepare mixtures with total phospholipid content of 1 wt. , and the mixture was heated at 60 for 10 min to easily dissolve SL in oil. Subsequently, 6 g of these mixtures were heated in a test tube at 200 for 15 min.
were mixed to obtain a total weight of 9.83 g, and the mixture was heated at 130 for 10 min to easily dissolve calcium stearate in oil. After cooling to 60 or below, SL 0.17 g was added to prepare mixtures with total phospholipid content of 1 wt. , and the mixture was heated at 60 for 10 min to easily dissolve SL in oil. Subsequently, 6 g of these mixtures were heated in a test tube at 200 for 15 min.

Heating SL in octane
Based on the methods described in previous papers 9, 10 , SL 5 g was dissolved in octane and refluxed for 9 h. Mixtures of SL with stearic acid 0.71 g, 2.5 mmol or calcium stearate 1.52 g, 2.5 mmol were also dissolved in octane and refluxed for 9 h. After heating for 9 h, the reactants were dried in a rotary evaporator under vacuum conditions and passed through a Sep-Pak silica cartridge. Fractions with increasing concentrations of methanol were eluted in sequence, resulting in chloroform, 25 methanol in chloroform, and 50 methanol in chloroform fractions. The fraction with 50 methanol in chloroform was analyzed using HPLC Column: Senshu Pak. AQUASIL SS-4251 250 mm 10 mm, i.d., Senshu, Japan; flow rate: 3.0 mL/min; solvent: a mixture of chloroform, methanol, and water 80:20:1, v/v/v , detector: UV at 350 nm .

Results and Discussion
3.1 Inhibition of heat deterioration of SL by treatment with metal silicate Helmy et al. 1994 16 reported that the color of refined cottonseed oil was improved when crude cottonseed oil was added to the metal silicate during the degumming process; however, the mechanism of oil color improvement remains unknown. Still, we investigated whether metal silicate treatment could inhibit the heat deterioration of SL.
Color changes caused by the heating of metal silicatetreated SL are shown in Fig. 2A. Intact SL, i.e., untreated SL, browned after heating at 200 for 15 min in oil e . However, when SL with acid value adjusted to 35 mg KOH/ g by adding soybean fatty acid and SL treated with magnesium silicate without acid value adjustment were heated in oil under the same conditions, the treated SLs also turned brownish and exhibited almost the same color f and g as intact SL e . These results showed that the addition of soybean fatty acids or treatment with magnesium silicate alone could not inhibit the heat deterioration of SL in oil. However, when soybean fatty acids were added to SL before treatment with magnesium silicate, the color improved marginally with increasing acid value h, i, j, k, l . On the other hand, when calcium silicate was used for the treatment of SL, the color improved significantly with increasing acid value, as shown in Fig. 2B h', i', j', k', l' . These results clearly show that soybean fatty acids are indispensable for inhibiting the browning of SL by metal silicate treatment, and that calcium silicate is more effective than magnesium silicate.
In order to confirm which fatty acid is involved in inhibiting the heat deterioration of SL, calcium silicate treatment was carried out after the addition of several fatty acids to SL. When saturated fatty acids, such as myristic acid, palmitic acid, and stearic acid, were added to SL before treatment with calcium silicate, the color indices after heating were 118, 125, 126, and the heat deterioration of SL could not be inhibited regardless of the carbon chain length Table 1 . On the other hand, when unsaturated fatty acids, such as oleic acid, linoleic acid, and α-linolenic acid, were added, the heat deterioration of SL was significantly inhibited and the color indices after heating were 23, 22, and 20, respectively. These results show that unsaturated fatty acids are indispensable for inhibiting the heat deterioration of SL by metal silicate treatment. It is well known that the main components of soybean fatty acids are unsaturated fatty acids, such as oleic acid and linoleic acid. It is, therefore, suggested that unsaturated fatty acids in soybean fatty acids are involved in the inhibition phenomenon. Furthermore, when myristic acid, palmitic acid, stearic acid, oleic acid, linoleic acid, and linolenic acid were added to SL before treatment with calcium silicate, the calcium concentrations of the mixtures were 9.80, 8.96, 7.94, 28.14, 29.76, and 31.05 mmol/L, respectively. The a. non heating of SL, b. non heating of SL which was adjusted to acid value 35 mg KOH/g, c. non heating of SL which was treated with magnesium silicate, d. non heating of SL which was adjusted to acid value 35 mg KOH/g before treated with magnesium silicate, e. after heating of SL, f. after heating of SL which was adjusted to acid value 35 mg KOH/g, g. after heating of SL which was treated with magnesium silicate, h. after heating of SL which was adjusted to acid value 15 mg KOH/g before treated with magnesium silicate, i. after heating of SL which was adjusted to acid value 20 mg KOH/g before treated with magnesium silicate, j. after heating of SL which was adjusted to acid value 25 mg KOH/g before treated with magnesium silicate, k. after heating of SL which was adjusted to acid value 30 mg KOH/g before treated with magnesium silicate, l. after heating of SL which was adjusted to acid value 35 mg KOH/g before treated with magnesium silicate.

Fig. 2B
Color changing by heating of SL treated with calcium silicate. a'. non heating of SL, b'. non heating of SL which was adjusted to acid value 35 mg KOH/g, c'. non heating of SL which was treated with calcium silicate, d'. non heating of SL which was adjusted to acid value 35 mg KOH/g before treated with calcium silicate, e'. after heating of SL, f'. after heating of SL which was adjusted to acid value 35 mg KOH/g, g'. after heating of SL which was treated with calcium silicate, h'. after heating of SL which was adjusted to acid value 15 mg KOH/g before treated with calcium silicate, i'. after heating of SL which was adjusted to acid value 20 mg KOH/g before treated with calcium silicate, j'. after heating of SL which was adjusted to acid value 25 mg KOH/g before treated with calcium silicate, k'. after heating of SL which was adjusted to acid value 30 mg KOH/g before treated with calcium silicate, l'. after heating of SL which was adjusted to acid value 35 mg KOH/g before treated with calcium silicate. calcium concentration with unsaturated fatty acids was more than three times that with saturated fatty acids. Thus, saturated fatty acids are probably less capable of extracting calcium from calcium silicate and producing fatty acid calcium salts. Based on these results, it is suggested that calcium is transferred to SL during calcium silicate treatment with unsaturated fatty acids, and calcium contributes to inhibiting the heat deterioration of SL.
3.2 Inhibition of heat deterioration of SL with addition of fatty acid metal salts It is suggested that the heat deterioration of SL is inhibited by calcium from calcium silicate, indicating that calcium exists as a fatty acid salt. Therefore, several fatty acid metal salts were added to confirm whether heat deterioration could be inhibited while heating SL. When SL was heated with sodium stearate, calcium stearate, calcium oleate, magnesium stearate, barium stearate, and aluminum tristearate, the color indices after heating were 93, 22, 23, 25, 22, and 63, respectively. It was observed that the heat deterioration of SL could be inhibited by fatty acid metal salts, regardless of whether the fatty acids were saturated or unsaturated and whether the metal was monovalent, divalent, or trivalent, as shown in Table 2. We newly clarified that the heat deterioration of SL could be inhibited by adding fatty acid metal salts to SL.
The heat deterioration of SL could be inhibited only when metal silicate treatment of the mixture was carried out after adding fatty acids to SL. This result shows that metal silicates and fatty acids form fatty acid metal salts, which inhibit the heat deterioration of SL. Furthermore, the addition of saturated fatty acids could not inhibit the heat deterioration of SL even after treatment with calcium silicate because the saturated fatty acids could not extract calcium to form fatty acid metal salts. In contrast, when saturated fatty acid metal salts were used directly, the heat deterioration of SL was clearly inhibited. This shows that the metal salts of fatty acids have an inhibiting effect on deterioration, and the type of fatty acid is irrelevant.
Among sodium stearate, calcium stearate, magnesium stearate, barium stearate, and aluminum tristearate, the divalent fatty acid metal salts had stronger heat deterioration inhibiting effects than the monovalent and trivalent salts. These results show that divalent fatty acid metal salts strongly inhibit the Maillard reaction of PEs and sugars. In addition, even if the oil had a phospholipid content of 1 , heat deterioration could be significantly inhibited by blending 0.3 calcium stearate, as shown in Fig. 3.
Hayashi et al. 12 described that upon heating SL in octane, the browning of the solution was accompanied by the production of four novel pyridinium derivatives and an increase in UV absorption at 350 nm. These pyridinium derivatives were formed by the Maillard reaction of 1 mol of any sugar, except 2-deoxy sugars, with 2 mol of any PE. It is believed that fatty acid metal salts inhibit this Maillard  reaction. Therefore, we examined whether SL and calcium stearate reacted in octane to produce these four pyridinium derivatives. When SL or a mixture of SL with stearic acid were heated in octane, the resultant solution turned black and the UV absorption at 350 nm increased Fig. 4 ; peaks of four compounds Rt, A: 26.4 min, B: 27.9 min, C: 35.6 min, D: 37.6 min were also observed in the HPLC chromatogram Fig. 5 . These retention times were consistent with the four pyridinium derivatives described by Hamaguchi et al. 10 . Based on the UV data and the retention times in the HPLC chromatogram, we determined that the four pyridinium derivatives were produced when SL or a mixture of SL with stearic acid were heated in octane. Judging from their peak areas in the chromatograms, similar amounts of pyridinium derivatives were produced in these experiments. In contrast, when a mixture of SL and calcium stearate was heated in octane, the color hardly changed and no increase in the UV absorption at 350 nm was observed, as shown in Fig. 4. Furthermore, peaks corresponding to the pyridinium derivatives were not observed in the chromatogram, as shown in Fig. 5. These results clearly show that calcium stearate inhibits the Maillard reaction between PEs and sugars.
We newly found that the addition of fatty acid metal salts can inhibit the heat deterioration of SL and that divalent fatty acid metal salts strongly inhibit the heat deterioration of SL. The novel treatment described herein may be highly useful in the food processing industry for preventing the heat deterioration of lecithin.