Natural and Synthetic Antioxidant Additives for Improving the

Nov 21, 2011 - Natural and Synthetic Antioxidant Additives for Improving the ..... EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/. EC an...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/JAFC

Natural and Synthetic Antioxidant Additives for Improving the Performance of New Biolubricant Formulations Lida A. Quinchia, Miguel A. Delgado,* Concepción Valencia, José M. Franco, and Crispulo Gallegos Departamento de Ingeniería Química, Facultad de Ciencias Experimentales, Universidad de Huelva, Campus Universitario de El Carmen, 21071 Huelva, Spain, Campus de Excelencia Internacional Agroalimentario, ceiA3 ABSTRACT: Knowledge of the oxidative stability of vegetable oils for lubricant applications is a key point, because vegetable oil oxidation potential is the main disadvantage for its use as a lubricant. Oil degradation after an oxidation process can seriously affect its lubricating function and increase wear. In this work, two different methods for evaluating the oxidation stability of lubricating vegetable oils, the oxidation onset temperature, characterized through DSC measurements (ASTM E 2009-08), and the pressure drop in the oxygen pressure vessel (ASTM D 942-02), have been used. Additionally, thermogravimetric analysis and FTIR studies have also been carried out. High-oleic sunflower (HOSO) and castor (CO) oils were selected and blended with natural ((+)-α-tocopherol (TCP), propyl gallate (PG), L-ascorbic acid 6-palmitate (AP)) or synthetic antioxidants (4,4′methylenebis(2,6-di-tert-butylphenol) (MBP)), with the aim of formulating biodegradable vegetable-based lubricants according to REACH regulation.1 The results showed that the most effective biodegradable antioxidant is PG, comparable to MBP, whereas lower effectiveness was obtained for TCP and AP. In relation to the methods tested, DSC measurements achieve accurate data more quickly for evaluating the oxidation stability of these basestocks, showing a linear correlation with the traditional method based on the oxygen bomb test. The empirical equation obtained depends on the mechanism involved in the antioxidant activity. KEYWORDS: vegetable oils, oxidative stability, antioxidants, lubricant, biolubricant



mentioned byproduct give rise to physical and chemical changes, which may have a dramatic impact on the lubrication performance.6 Nevertheless, the use of antioxidant additives can delay or prevent the oxidation process by protecting the lubricant from oxidative degradation while also allowing the oil to properly fulfill the requirements demanded by the industry. 8,10 Antioxidants interrupt the autoxidation process in different ways, according to their structure and antioxidant mechanism. Different antioxidant additive groups, such as primary, secondary, and multifunctional antioxidants, hydroxylamines, and alkyl radical scavengers, can be found in the literature. 11 Most of them act as primary antioxidants, or chain-breaking radical scavengers, or secondary antioxidants, also named oxygen scavengers and peroxide decomposers, or through a combination of the functionalities of both of them. Several types of antioxidants are widely used to increase oxidative stability in mineral or synthetic lubricants. In the particular case of vegetable oils, natural, that is, tocopherols, propyl gallate (PG), and ascorbyl palmitate (AP), and synthetic antioxidants, that is, butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), mono-tert-butylhydroquinone (TBHQ), or 4,4′-methylenebis(2,6-di-tert-butylphenol) (MBP), may be used to formulate environmentally friendly lubricants.12,13 With the exception of the ascorbyl palmitate, which is used as an oxygen scavenger, all of these antioxidants act by quenching free radicals.6,14

INTRODUCTION Lubricants inevitably affect the environment and, for that reason, the development of new environmentally friendly lubricant formulations is required. It is well-known that lubricating oils consist of a mixture of different base oils and additives, oil being their main constituent.2 In this sense, the use of vegetable oils as basestocks represents an important alternative to achieve this need, providing higher levels of biodegradability and low toxicity. Besides this, vegetable oils have other advantages, such as very low volatility, good viscosity/temperature relationship, and high lubricity. On the contrary, they possess low thermal and oxidative stabilities, which represent the main problems for their use as lubricants. 3,4 Oils having poor thermal and oxidative stabilities tend to form deposits, sludge, and corrosive byproduct, which can strongly affect the lubricating process during their industrial in-life service.5 Therefore, a fundamental and practical knowledge of the oxidative stability of vegetable oil-based basestocks is necessary for their potential lubricant applications. Thermal and thermo-oxidative behaviors of vegetable oils are related to their fatty acid profile. Thus, the greater the level of unsaturation, that is, the greater the number of double bonds, the more susceptible the oil becomes to oxidation.6 The mechanism for the autoxidation of vegetable oils has been widely studied for decades.2,6−9 The process of oxidation requires an initiation process, when the first free radicals are formed. Free radicals rapidly react with oxygen to form hydroperoxides and, then, peroxides, which propagate the oxidation process. Finally, aldehydes, ketones, and acids are formed. The autoxidation process has, then, initiation, propagation, and termination reactions, which could be cyclical once started. These oxidation processes and the above© 2011 American Chemical Society

Received: Revised: Accepted: Published: 12917

September 7, 2011 November 21, 2011 November 21, 2011 November 21, 2011 dx.doi.org/10.1021/jf2035737 | J. Agric.Food Chem. 2011, 59, 12917−12924

Journal of Agricultural and Food Chemistry

Article

ranged between 0.5 and 2% (w/w). Blends were prepared by stirring the sample during 1 h, after antioxidant addition, at 100−120 °C, depending on the antioxidant melting point. This thermal protocol was required to completely solubilize the antioxidants in the vegetable oils. Afterward, samples were cooled to room temperature. A homogeneous single phase was obtained in all cases. However, antioxidants PG and AP showed a certain cloudiness with time in HOSO at concentrations >0.5 wt %. Analytical Methods. Thermogravimetric Analysis. The thermal degradation of vegetable oil/antioxidant blends was determined in a Q-50 thermobalance (TA Instruments Waters). Samples (10 ± 1 mg) were analyzed under both nitrogen and air atmospheres (flow of 60 mL/min) and applying a heating rate of 10 °C/min, in a temperature range of 35−500 °C. Oxidation Onset Temperature (OOT) Test. The OOT test was used to evaluate the effectiveness of antioxidants in vegetable oils, according to ASTM E 2009-08.16 The tests were performed using a differential scanning calorimeter (DSC) Q-100 (TA Instruments Waters) under an oxygen atmosphere. Oxygen flow was maintained at 50 mL/min. Oil samples (3.0−3.3 mg) were placed in open aluminum pans and heated from 100 to 250 °C, at a heating rate of 10 °C/min. All of the experiments were replicated. The OOT was obtained by using the TA Instruments Universal Analysis software, according to ASTM E 2009-08. Oxidation Bomb Test. Oxidation bomb tests were done according to ASTM D 942-02,15 using a Hoffmann bomb Stanhope-Seta (U.K.). Four grams of sample was placed on five Pyrex dishes inserted into a stainless steel bomb. The bomb was sealed and pressurized with oxygen at 758 kPa. The whole bomb was immersed in an oil bath, at a temperature of 99 °C. The pressure drop was continuously registered as a function of time. Tests were carried out in duplicate, until the samples were completely oxidized (up to constant pressure values). Fourier Transform Infrared Spectroscopy (FTIR). The infrared spectra of vegetable oil/antioxidant blends were recorded with a Fourier transform infrared spectrometer, Digilab FTS3500ARX (Varian Inc.), before and after the bomb oxidation tests (ASTM D 942-02) had been performed. Samples were placed between two KBr disks (32 × 3 mm), and the set was placed into an appropriate sample holder. The spectra were obtained in a wavenumber range of 400− 4000 cm−1, at 4 cm−1 resolution in the absorbance mode.

A variety of methods have been reported to assess the oxidation and thermal stabilities of lubricating oils. Basically, a series of parameters, which allow the overall degradation process to be detected, are controlled. These include peroxide formation, oxygen absorption, and the heat of the reaction involved.7 The ASTM D 942-02 standard,15 usually known as the “oxygen bomb” method, has been commonly used in an industrial environment. It entails measuring the net change in pressure resulting from the consumption of oxygen by oxidation. However, a recent standard, ASTM E 2009-08, 16 has been proposed. It is based on the evaluation of the oxidative stability of the substance at a given heating rate and oxidative environment by determining the oxidation onset temperature (OOT) from differential scanning calorimetry (DSC) tests. 9,14 In this work, the thermal analysis technique and the oxygen bomb method, complemented with the FTIR technique, have been used to evaluate the oxidation stability of two vegetable oils used for lubricant purposes, high-oleic sunflower oil (HOSO) and castor oil (CO), with and without the addition of natural or synthetic antioxidants. Moreover, the advantages and disadvantages of each technique are emphasized.



MATERIALS AND METHODS

Materials. Two vegetable oils were used as basestocks, a refined HOSO (85 wt % oleic acid), kindly supplied by the Instituto de la Grasa, CSIC (Seville, Spain), and a CO, purchased from Guinama (Spain). HOSO was selected because of its high content of oleic acid and, consequently, a lower level of polyunsaturated fatty acid. CO was selected because of its high kinematic viscosity. The main physical properties of these two oils are collected in Table 1.

Table 1. Densities, Kinematic Viscosities, and Viscosity Index (VI) Values for Neat High-Oleic Sunflower Oil (HOSO) and Castor Oil (CO) kinematic viscosity (mm2/s) vegetable oil

40 °C

100 °C

density (g/cm3), at 25 °C

VI

HOSO CO

38 242

10 21

0.9035 0.9559

257 116



RESULTS AND DISCUSSION Oil Thermal Degradation. Thermogravimetric analysis was performed to evaluate the minimum temperature at which thermal decomposition of neat oils, or their blends with antioxidants, appears. Thermal decomposition profiles for both castor and high-oleic sunflower oils, under nitrogen or air atmosphere, are shown in Figure 2. As can be seen, the thermogravimetric curves obtained using nitrogen or air atmosphere are rather different for both oils. Thus, under air atmosphere, oil degradation begins first, as a consequence of oxidation. For instance, HOSO thermal decomposition occurs in a two-stage process. The first thermal event (Tmax = 371 °C) is attributed to the decomposition of monounsaturated fatty acids, such as oleic acid. During this reaction, the double bonds are broken, and the molecules become saturated. The second step, occurring at around 415 °C, corresponds to the thermal decomposition of saturated fatty acids, such as palmitic acid. 17 With regard to castor oil, its thermogravimetric curve presents three stages, which are consecutive and have been mainly attributed to the decomposition and/or volatilization of the ricinoleic fatty acid18 and subsequent degradation compounds at temperatures of 371, 410, and 432 °C, with weight losses of 35, 59, and 70%, respectively. On the other hand, Figure 3 shows the thermal decomposition profiles, under oxidative environment, for both vegetable oils and their blends with 0.5 wt % of one of

(+)-α-Tocopherol (TCP), propyl gallate (PG), and L-ascorbic acid 6-palmitate (AP) were selected as natural and biodegradable antioxidants, respectively. 4,4′-Methylenebis(2,6-di-tert-butylphenol) (MBP) was used as a synthetic antioxidant. The chemical structures of the different antioxidants are shown in Figure 1. All of these additives are commercially available and, in this case, were obtained from Sigma-Aldrich (Spain).

Figure 1. Molecular formulas of the antioxidants used. Blends Manufacture. Vegetable oils were blended with each antioxidant additive in batches of 250 cm3. Antioxidant concentration 12918

dx.doi.org/10.1021/jf2035737 | J. Agric.Food Chem. 2011, 59, 12917−12924

Journal of Agricultural and Food Chemistry

Article

Figure 3. Thermogravimetric analysis curves for high-oleic sunflower oil (HOSO), castor oil (CO), and their blends with 4,4′-methylenebis(2,6-di-tert-butylphenol) (MBP), under air atmosphere. (Thin lines correspond to weight loss curves and thick lines to derivative weight loss curves.)

Figure 2. Thermogravimetric analysis curves for high-oleic sunflower oil (HOSO) and castor oil (CO), under nitrogen and air atmospheres. (Thin lines correspond to weight loss curves and thick lines to derivative weight loss curves.)

the antioxidants tested, MBP. By comparing their respective weight loss versus temperature plots, different antioxidant effects depending on oil nature can be noted. Table 2 shows the main characteristic temperatures, obtained from TGA curves, for vegetable oils containing the different antioxidants used. As can be observed, the onset temperatures for thermal decomposition are usually lower for HOSO/antioxidant blends, because only one of the antioxidants (PG) improves oil thermal stability (Tonset increases from 331 to 347 °C). Similar comments can be made for the temperature at which the maximum decomposition rate takes place (Tmax). On the contrary, three of the antioxidants used do delay CO thermal degradation onset. Thus, only α-tocopherol decreases Tonset (Table 2). However, in this case, Tmax, which defines the main thermal event, is always higher for all of the antioxidant/CO blends. Therefore, most antioxidants have higher affinity with CO than with HOSO, due to the presence of functional polar groups (−OH). Consequently, although the onset temperature for antioxidant thermal decomposition is lower than for the neat oils studied, interactions with the −OH group of ricinoleic fatty acid in castor oil prevent volatilization of the antioxidant molecules. Oxidation Bomb Tests. The oxygen pressure bomb method (ASTM D 942-02) enables the analysis of the resistance to oxidation of vegetable oils, in a very intuitive way, under normalized conditions of temperature and oxygen atmosphere. This technique is widespread among lubricant manufacturers and assesses the amount of oxygen that reacts with the sample as a function of time. Figure 4 gathers the

Table 2. Thermogravimetric Analysis of High-Oleic Sunflower Oil (HOSO) and Castor Oil (CO) in the Presence of Natural and Synthetic Antioxidants (0.5 wt %) sample

Tonset (°C)

decomposition temperature range (°C)

Tmax (°C)

residue (%)

331 323 347

331−446 323−459 347−421

371; 417 371; 410; 432 382; 410

7.3 6.6 0.3

323

323−433

358; 376; 417

8.5

322

322−424

356; 417

HOSO CO HOSO + PG HOSO + MBP HOSO + AP HOSO + TCP CO + PG

312

312−424

351; 416; 434

8.3

338

338−420

6.8

CO + MBP

336

336−449

CO + AP CO + TCP

336 314

336−431 314−405

370; 407; 432; 462 371; 400; 430; 468 370; 429 363; 426

10.1

7.4 6.7 7.0

pressure drop versus time for the two vegetable oils studied and their blends with 0.5 wt % of different antioxidants. Table 3 collects the oxidation times, evaluated according to this standard. The oxidation time is calculated from the intersection of the tangent line to the initial baseline and the tangent line to the point of maximum decrease, corresponding to the inflection point of the oxidation peak. 12919

dx.doi.org/10.1021/jf2035737 | J. Agric.Food Chem. 2011, 59, 12917−12924

Journal of Agricultural and Food Chemistry

Article

Figure 4. Oxidation stabilities of high-oleic sunflower oil (HOSO), castor oil (CO), and their blends with different antioxidants (0.5 wt %), using the oxidation bomb test.

Table 3. Oxidation Onset Temperature (OOT)a Values and Oxidation Times (OPV)b for High-Oleic Sunflower Oil (HOSO), Castor Oil (CO), and Their Blends with Several Natural and Synthetic Antioxidants, at Different Concentrations HOSO antioxidant

antioxidant concn (wt %)

antioxidants tested, the results obtained with TCP and AP were rather similar in both cases. Thus, AP slightly increases the oxidation time, whereas TCP induces a higher consumption of oxygen. In addition to this, this standard method suggests that these antioxidants are more effective with HOSO than with CO, taking into account that neat CO was more resistant to oxidation than neat HOSO (Table 3). This is most probably related to the amount of natural antioxidants present in the vegetable oils and the synergism with the antioxidants added. Thus, for instance, 0.05 wt % TCP, in its natural form, is found in castor oil, whereas sunflower oil contains 0.07 wt % antioxidants, mainly as α-tocopherol (in vivo activity).19 Furthermore, some modified high-oleic varieties contain γand δ-tocopherols, which possess antioxidant properties in vitro and could even have synergism with other antioxidants.20,21 It is worth pointing out that the mechanisms involved in the antioxidant activity of these additives are quite different. For instance, MPB, PG, and TCP are primary or chain-breaking antioxidants, which act by delaying the initial autoxidation step.6,8 As was previously mentioned, MBP and PG showed very striking results. Thus, a delay of the oxidation step and, consequently, an improvement of vegetable oil stability against oxidation were always noted. On the contrary, much worse results were found by using TCP. This may be explained by taking into account the low activity of tocopherols, especially at high temperature and when added to oils and fats.8,22 On the other hand, AP is a secondary antioxidant, which acts as an oxygen scavenger.12 As a consequence, a slight influence was observed in the oxidation bomb test (Table 3). Thereby, the main difference between primary and secondary antioxidants is that the latter do not convert free radicals into stable molecules,8 which can continue consuming oxygen. In summary, antioxidant effectiveness is affected by chemical properties and environmental conditions, whereas the oxidation bomb test results are influenced by the mechanisms involved in the antioxidant activity. To evaluate the physical and chemical changes induced by the oxidation bomb test, oxidized vegetable oil samples, with and without antioxidants, were analyzed visually and by FTIR. The first visual indication of oil oxidation is the change in color due to the oxidative reactions occurred. Different studies have found that the sequence of compounds formed during

CO

OOT (°C)

OPV (h)

OOT (°C)

OPV (h)

none

0

191

14

197

48

PG

0.5 1 1.5 2

225 233

337 407

233 234 240 242

80 69 105 114

AP

0.5 1 1.5 2

206

21

208 212 200 198

47 60 20 22

TCP

0.5 1 2

190 206 214

3

195 201 203

33

MBP

0.5 1 2

226 232 242

217

212 222 234

105

a

Oxidation onset temperature of hydrocarbons by differential scanning calorimetry (ASTM E 2009-08). bOxygen pressure vessel (oxidation stability of lubricating greases by oxygen pressure vessel method, ASTM D 942-02).

The results obtained for HOSO (Figure 4a; Table 3) show strong evidence regarding the effectiveness of PG for delaying oxidation, which occurs 323 h later when 0.5 wt % of this antioxidant is added to the neat oil. MBP also shows an important effect in delaying HOSO oxidation. Similar results were obtained for castor oil (Figure 4b; Table 3), although, in this case, the largest antioxidant activity was found for MBP. Thus, oxidation occurred 57 h later when 0.5 wt % of this antioxidant was added to the neat oil. PG addition also yielded large increases in oxidation time. With respect to the rest of the 12920

dx.doi.org/10.1021/jf2035737 | J. Agric.Food Chem. 2011, 59, 12917−12924

Journal of Agricultural and Food Chemistry

Article

Figure 5. FTIR spectra of oxidized and nonoxidized high-oleic sunflower oil (HOSO), castor oil (CO), and their blends with different antioxidants (0.5 wt %), after the oxidation bomb test. The peak height ratios between characteristic bands are included.

oxidation bomb test. Besides this, the band at 3006 cm −1 is typically associated with the C−H stretching vibration of the cis-double bond.28 The band position is almost unaltered or undergoes just a very low shifting as a result of the oxidative process.27 The reason for this is the disappearance of cis double bonds, as well as the isomerization of cis to trans groups, along with hydroperoxide generation. As a consequence, the frequency and absorbance of the band close to 3006 cm−1 suffer slight changes as the oxidation process proceeds further.23 For nonoxidized HOSO and its blends with MBP and PG, this band remains constant at 3006 cm−1, but, for oxidized HOSO and its blends with AP and TCP, this band is shifted to lower wavenumbers (3003 cm−1), which may reveal evidence of an oxidation process, although, in this case, it could not be considered to be significant. Something similar occurs for castor oil samples. Another characteristic band appears at 970 cm−1. This band is usually assigned to the acyl groups of linoleic or linolenic fatty acids, but can be also attributed to secondary oxidation products, such as aldehydes or ketones. For HOSO, with a low content in these fatty acids, this band cannot be detected. However, once the sample is oxidized, it is clearly noted. This band can be also seen in samples containing AP and TCP; however, it is suppressed by the other two antioxidants tested (MBP and PG) (Figure 5a). These findings are consistent with those found from other tests, reflecting that additives AP and TCP show poor antioxidant performance with both oils. With castor oil, this band is not detected nor for nonoxidized or oxidized samples. Figure 5 also shows the intensity ratios between the absorbance peaks at 2853 and 3450 cm−1 for neat HOSO and CO, as well as for their respective blends with antioxidants. The ratio of the absorbance of the bands at 2853 cm −1, due to symmetric stretching vibration of CH2 groups, and at 3450 cm−1 has been selected as an index of the oxidation process. As can be seen, the ratio P2853/P3500−3400 is much higher for the nonoxidized oils (58.8 and 23 for HOSO and CO, respectively), whereas lower values were found for oxidized oils (15.7 and 5.2 for HOSO and CO, respectively). With regard to their blends with the antioxidants, the lower values of this ratio were found again for TCP− and AP−oil blends. On

oxidation is as follows: hydroperoxides, then peroxides, and, finally, aldehydes, ketones, and acids.6,7 In this sense, the samples become dark due to accumulation of the abovementioned products. In addition to this, it is well-known that the chemical changes produced in vegetable oils during the oxidation process can be suitably analyzed using FTIR.6,23 As previously reported,24 the FTIR spectra of vegetable oils change due to the presence of oxidation products. As can be seen in Figure 5, the FTIR spectra of CO and HOSO are characterized by the presence of several peaks in the studied wavenumber range, some of them typically found in neat vegetables oils, 25 whereas others appear only after the oxidation process. 23,26 In this work, the main interest has been focused on the range of 3000−3500 cm−1, where the activity of stretching vibrations of fatty acids and the presence of hydroxyl groups and hydroperoxides formed in oxidation steps appear.26 In addition, special attention has been paid to the range between 700 and 1500 cm−1, where the vibrational activity of conjugated bonds and bending vibrations of aliphatic compounds can be detected. In this range, some bands can be also assigned to secondary oxidation products, such as aldehydes or ketones. 23,27 In this sense, the band at 3450 cm−1 is assigned to oxidation products, such as hydroperoxides, that are formed in the initial phase of the oxidative process. In neat vegetable oils, this band can overlap with the one at 3470 cm−1, related to the overtone of the glyceride ester carbonyl group, which is widened and intensified after the oxidation process. In this particular case, for oxidized additive-free HOSO, the band at 3450 cm−1 is clearly produced for hydroperoxide formation and, as shown in Figure 5a, all of the antioxidants assessed inhibit the appearance of this particular band, which means that they are able to avoid the formation of hydroperoxides. In the case of CO, the hydroperoxide band is overlapped by the typical hydroxyl group band (at 3500 cm−1), although, in this case, the band intensity clearly changes. Thus, we can conclude that CO containing MBP has the lowest absorbance, which suggests that this additive protects the oil more significantly than the rest of the antioxidants tested. On the contrary, CO containing AP shows the highest absorbance, although lower than for the oxidized additive-free CO, followed by the sample containing TCP. These results are in agreement with those obtained in the 12921

dx.doi.org/10.1021/jf2035737 | J. Agric.Food Chem. 2011, 59, 12917−12924

Journal of Agricultural and Food Chemistry

Article

Figure 6. Oxidation stabilities of high-oleic sunflower oil (HOSO), castor oil (CO), and their blends with different antioxidants (0.5 wt %), using the oxygen onset temperature test (OOT).

the contrary, the higher P2853/P3500−3400 values were found for MBP/HOSO and MBP/CO systems. Oil Oxidation Onset Temperature. OOT values, according to ASTM E 2009-08 standard, were determined for CO and HOSO, with or without antioxidant addition. Different antioxidant concentrations were evaluated (0.5, 1, and 2 wt %). Figure 6 shows selected heat flow curves for both oils and their blends with the antioxidants studied (0.5 wt %). Their respective OOT values are listed in Table 3. The oxidation onset temperature has been calculated from the intersection of the tangent line to the initial baseline and the tangent line to the point of maximum increase, corresponding to the inflection point of the oxidation peak. The results obtained from ASTM E 2009-08 demonstrate that, in general, these antioxidants delay vegetable oil oxidation, as can be deduced from the increase in OOT values. The effect of 0.5 wt % TCP is almost negligible. However, larger concentrations yield higher OOT values for oils containing αtocopherol than those found for neat oils (mainly HOSO). On the other hand, the most outstanding results were obtained with propyl gallate and castor oil. Thus, for instance, castor oil OOT was enhanced from 197 to 233 °C by adding 0.5% PG. Besides this, the effects of PG and MBP on HOSO are very similar, whereas the effect of MBP on CO is more limited. These differences can be explained by considering the different chemical compositions of HOSO and CO.14 Finally, AP slightly influences the delay of the oxidation process. Therefore, on the basis of the OOT, the antioxidant activity of the additives evaluated can be ranged as follows: PG > MBP > AP > TCP. Correlation between DSC and Oxidation Bomb Results. The results obtained from both DSC and oxidation bomb tests have been correlated for two of the antioxidants used (PG and AP). They have been chosen by taking into account the different mechanisms involved in their antioxidant activity. Moreover, the influence of antioxidant concentration has been evaluated by using blends of these additives with castor oil, in a concentration range of 0.5−2 wt %. DSC and pressure drop curves are shown in Figure 7. The comparison between both sets of data reveals important differences concerning the performance of both antioxidants. Panels a and b of Figure 7 collect the results obtained with PG. The

Figure 7. Oxidation stabilities of castor oil and its blends with PG and AP, at different antioxidant concentrations, using both OOT and oxidation bomb tests.

oxidation bomb curves (Figure 7b) show that castor oil oxidation stability always increases with additive concentration. However, pressure drop occurs more rapidly for the lowest concentrations (0.5 and 1 wt %) and is retarded for larger additive concentrations (1.5 and 2 wt %). However, pressure versus time plots for oil samples containing 1.5 and 2% PG are very similar. These results are in agreement with those found in the oxidation onset tests (DSC), which allows a linear correlation between OOT and OPV time values to be obtained (Figure 8). With regard to the antioxidant AP (Figure 7c,d), OPV results are quite different from those found for PG. Thus, the lowest concentrations (0.5 and 1 wt %) do not yield any significant improvement in oxidation performance. By increasing further the AP concentration (1.5 and 2 wt %), a faster pressure drop is 12922

dx.doi.org/10.1021/jf2035737 | J. Agric.Food Chem. 2011, 59, 12917−12924

Journal of Agricultural and Food Chemistry



REFERENCES

(1) Regulation 06/1907/EC of the European Parliament and of the Council of 18 December 2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency, amending Directive 1999/45/EC and repealing Council Regulation (EEC) No. 793/93 and Commission Regulation (EC) No. 1488/94 as well as Council Directive 76/769/ EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/ EC and 2000/21/EC. Off. J. Eur. Communities 2006, L396. (2) Luna, F. M. T.; Rocha, B. S.; Rola, E. M.; Albuquerque, M. C. G.; Azevedo, D. C. S.; Cavalcante, C. L. Assessment of biodegradability and oxidation stability of mineral, vegetable and synthetic oil samples. Ind. Crops Prod. 2011, 33, 579−583. (3) Erhan, S. Z.; Sharma, B. K.; Perez, J. M. Oxidation and low temperature stability of vegetable oil-based lubricants. Ind. Crops Prod. 2006, 24, 292−299. (4) Campanella, A.; Rustoy, E.; Baldessari, A.; Baltanás, M. A. Lubricants from chemically modified vegetable oils. Bioresour. Technol. 2010, 101, 245−254. (5) Duangkaewmanee, S.; Petsom, A. Synergistic and antagonistic effects on oxidation stability of antioxidants in a synthetic ester based oil. Tribol. Int. 2011, 44, 266−271. (6) Fox, N. J.; Stachowiak, G. W. Vegetable oil-based lubricants − a review of oxidation. Tribol. Int. 2007, 40, 1035−1046. (7) Hassel, R. L. Thermal analysis: an alternative method of measuring oil stability. J. Am. Oil Chem. Soc. 1976, 53, 179−181. (8) Wanasundara, P.; Shahidi, F. Antioxidants: science, technology, and applications. In Bailey’s Industrial Oil and Fat Products, 6th ed., Vol. 1, Edible Oil and Fat Products, Chemistry, Properties and Health Effects; Shahidi, F., Ed.; Wiley: Hoboken, NJ, 2005; pp 431−489. (9) Pardauil, J. J. R.; Souza, L. K. C.; Molfetta, F. A.; Zamian, J. R.; Rocha, G. N.; da Costa, C. E. F. Determination of the oxidative stability by DSC of vegetable oils from the Amazonian area. Bioresour. Technol. 2011, 102, 5873−5877. (10) Garcia-Perez, M.; Adams, T. T.; Goodrum, J. W.; Das, K. C.; Geller, D. P. DSC studies to evaluate the impact of bio-oil on cold flow properties and oxidation stability of bio-diesel. Bioresour. Technol. 2010, 101, 6219−6224. (11) Thomas, R.; Dexter, M.; King, R. E., III. Antioxidants, polymers. In Encyclopedia of Chemical Technology; Kirk, R. E., et al., Eds.; Wiley: Hoboken, NJ, 2002; Vol. 3, pp 102−134. (12) Hras, A. R.; Hadolin, M.; Knez, Z.; Bauman, D. Comparison of antioxidative and synergistic effects of rosemary extract with αtocopherol, ascorbyl palmitate and citric acid in sunflower oil. Food Chem. 2000, 71, 229−233. (13) Merrill, L. I.; Pike, O. A.; Ogden, L. V.; Dunn, M. L. Oxidative stability of conventional and high-oleic vegetable oils with added antioxidants. J. Am. Oil Chem. Soc. 2008, 85, 771−776. (14) Artiaga, R.; López-Beceiro, J.; Tarrío-Saavedra, J.; Mier, J.; Naya, S.; Gracia, C. Oxidation stability of soy and palm based biodiesels evaluated by pressure differential scanning calorimetry. J. ASTM Int. 2010, 7, 1−10. (15) ASTM D 942-02. Standard test method for oxidation stability of lubricating greases by the oxygen pressure vessel method. In Annual Book of ASTM Standards; ASTM International: West Conshohocken, PA, 2002. (16) ASTM E 2009-08. Standard test method for oxidation onset temperature of hydrocarbons by differential scanning calorimetry. In Annual Book of ASTM Standards; ASTM International: West Conshohocken, PA, 2008. (17) Souza, A. G.; Oliveira, J. C.; Conceiçao, M. M.; Dantas, M. C.; Prasad, S. A thermoanalytic and kinetic study of sunflower oil. Braz. J. Chem. Eng. 2004, 21, 265−273. (18) Conceiçao, M. M.; Candeia, R. A.; Silva, F. C.; Bezerra, A. F.; Fernandes, V. J.; Souzab, A. G. Thermoanalytical characterization of castor oil biodiesel. Renew. Sustain. Energy Rev. 2007, 11, 964−975. (19) Bailey, A. E. Aceites y Grasas Industriales; Ed. Reverté: Barcelona, Spain, 2001; pp 743.

Figure 8. Relationship between experimental oxidation bomb times and OOT onset temperatures for castor oil and its blends with PG and AP, at different antioxidant concentrations.

noted. This result was unexpected and suggests that the antioxidant mechanism plays an important role. As previously discussed, AP is a secondary antioxidant, which acts as an oxygen scavenger and decomposes oxidation products such as hydroperoxides.8,12 Therefore, AP consumes oxygen by itself and, because of this, OPV time is shorter at high concentrations and, also, in comparison with other additives, such as PG. In addition, it is worth remarking that FTIR results also highlight this discouraging behavior. Furthermore, these results are also in agreement with OOT tests, as can be observed in Figure 8. Conclusions. The oxidation stability tests used in this work, that is, the oxidation onset temperature and the oxidation bomb tests, are useful to determine the effectiveness of several natural and synthetic antioxidants, when blended with HOSO and CO, to delay oil oxidation. The experimental results obtained reveal that MBP and PG antioxidants, even at low concentration, defer the onset of oxidation or slow the oxidation rate, thus improving the thermal stability of the vegetable oils studied. On the contrary, TCP and AP seem not to be very effective antioxidants for these oils. A linear correlation between the main parameters obtained from both types of methods (OOT and OPV time) has been found. The oxygen bomb test, complemented with FTIR analysis, provides a full description of the chemical changes occurring in high-oleic sunflower and castor oils upon oxidation and also provides evidence on the mechanisms involved in the antioxidant activity. Temperatures for both thermal degradation onset and maximum decomposition rate generally increase with the addition of antioxidants to castor oil. On the contrary, only PG delays HOSO thermal degradation.



Article

AUTHOR INFORMATION

Corresponding Author *Phone: +34959219997. Fax: +34959219983. E-mail: miguel. [email protected]. Funding This work is part of two research projects (PSE420000-2008-4 and CTQ2010-15338) sponsored by MICINN-FEDER programs (Spain). L.A.Q. received a Ph.D. Research Grant from the FPU-MEC program (Spain). We gratefully acknowledge their financial support. 12923

dx.doi.org/10.1021/jf2035737 | J. Agric.Food Chem. 2011, 59, 12917−12924

Journal of Agricultural and Food Chemistry

Article

(20) Silva, F. A. M.; Borges, F.; Ferreira, M. A. Effects of phenolic propyl esters on the oxidative stability of refined sunflower oil. J. Agric. Food Chem. 2001, 49, 3936−3941. (21) Velasco, L.; Fernandez, J. M.; Begoa, V. Aceite de girasol con elevada termoestabilidad. Sp. Patent 52220464, 2010. (22) Becker, R.; Knorr, A. An evaluation of antioxidants for vegetables oils at elevated temperature. Lubr. Sci 1996, 8, 96−117. (23) Guillén, M. D.; Cabo, N. Usefulness of the frequency data of the Fourier transform infrared spectra to evaluate the degree of oxidation of edible oils. J. Agric. Food Chem. 1999, 47, 709−719. (24) Khan, M. N.; Sarwar, A.; Wahab, M. F. Chemometric assessment of thermal oxidation of some edible oils. Effect of hot plate heating and microwave heating on physicochemical properties. J. Therm. Anal. Calorim. 2010, 102, 369−374. (25) Guillén, M. D.; Cabo, N. Relationships between the composition of edible oils and lard and the ratio of absorbance of specific bands of their Fourier transform infrared spectra. Role of some bands of the fingerprint region. J. Agric. Food Chem. 1998, 46, 1788− 1793. (26) Lai, Y.; Kemsley, E. K.; Wilson, R. H. Potential of Fourier transform infrared spectroscopy for the authentication of vegetable oils. J. Agric. Food Chem. 1994, 42, 1154−1159. (27) Navarra, G.; Cannas, M.; D’Amico, M.; Giacomazza, D.; Militello, V.; Vaccaro, L.; Leone, M. Thermal oxidative process in extra-virgin olive oils studied by FTIR, rheology and time-resolved luminescence. Food Chem. 2011, 126, 1226−1231. (28) Guillén, M. D.; Cabo, N. Characterization of edible oils and lard by Fourier transform infrared spectroscopy. Relationships between composition and frequency of concrete bands in the fingerprint region. J. Am. Oil Chem. Soc. 1997, 74, 1281−1286.

12924

dx.doi.org/10.1021/jf2035737 | J. Agric.Food Chem. 2011, 59, 12917−12924