Non-Normative Oxidation Stability Indication of FAME Produced from Rapeseed and Used Cooking Oil

Abstract

The article describes the mechanism of oxidation of polyunsaturated fatty acid esters and presents the effect of temperature as one of the factors accelerating this process. The consequences of aging for ester-based biofuels are discussed extensively. The article presents the results of aging of FAMEs obtained from frying oil and of FAMEs derived from unrefined rapeseed oil to examine the influence of temperature as a factor accelerating aging. Acid, peroxide and anisidine values were determined for each sample; additionally, IR spectra were measured. Based on the results, it was confirmed that temperature has a significant effect on the degradation of FAMEs. These changes are best represented by the anisidine value and the change in IR spectra. The paper presents the relationship between acid, peroxide and anisidine values. This paper also highlights the beneficial effect of natural antioxidants in the form of β–carotene, which is found in FAMEs derived from unrefined rapeseed oil.

1. Introduction

Fatty Acid Methyl Esters (FAMEs) have occupied a prominent position as biofuels and biocomponents in the last two decades. They are obtained by transesterification of vegetable oils and methanol [1]. One of the cheaper and more accessible feedstocks for FAME production is used cooking oil (UCO) [2,3]. Used cooking oil is produced from the use of vegetable oils and is considered waste. When oils are used during food preparation in the food industry, the fatty acids contained in triacylglycerols, which are the main component of vegetable oils, are degraded [2,4]. FAMEs obtained from frying oil have a number of unsatisfactory properties relative to diesel fuel (ON) in the context of use in compression–ignition engines. The most unfavorable are related to aging, which occurs at a faster rate in FAMEs than it does in diesel fuel. This raises issues regarding the use of fatty acid methyl esters as motor fuel or as a bio-component. The most important aspects to look out for in FAMEs and blends of FAMEs with diesel include susceptibility to microbial decomposition, low-temperature properties, and oxidative stability during storage [5,6]. The content of unreacted acids, methanol, metals, and free glycerol in FAMEs accelerates the degradation process. Unsaturated compounds contained in biofuels are easily oxidized and tend to polymerize, which leads to problems not only with storage, but also with the formation of polymerized deposits in the combustion chamber and injection system [7].

Fatty acid methyl esters contain structural elements that are particularly susceptible to chemical reactions. These include allylic hydrogen atoms, which readily undergo radical reactions such as oxidation and polymerization; double bonds, which are susceptible to oxidation, addition and polymerization reactions; and ester groups that are highly prone to hydrolysis and transesterification [8]. One factor that accelerates these reactions is temperature, which has a pronounced effect on the aging process of biofuel in that it accelerates its oxidation. The initial products of oxidation are peroxides and hydroperoxides. They then participate in subsequent transformations to form secondary oxidation products in the form of aldehydes, low-molecular-weight carboxylic acids, formic acid and its esters, and high-molecular-weight fatty acid oligomers formed by oxidative polymerization [9].

Unsaturated fatty acids can undergo peroxidation, which is a free-radical process consisting of three stages: initiation, propagation, and termination [10,11].

At the initiation stage, the products are free ester radicals. According to the principles of thermodynamics, oxygen in the ground-triplet state (positive resultant angular momentum) will not react directly with the double bond, which is originally in the singlet state (no unpaired electrons, no resultant angular momentum) [11]. In order for the reaction to proceed, the double bond must transition into the triplet state. Such a transition requires an amount of energy equal to 35–65 kcal/mol [12].

$$LH+O_{\mathbf{2}}\to L·+HO$$

(1)

$$\mathbf{2}LH+O_{\mathbf{2}}\to \mathbf{2}L·+H_{\mathbf{2}}{O}_{\mathbf{2}}·$$

(2)

In order to initiate an ester oxidation reaction, catalysts are required; these can be metals (e.g., Fe, Cu, Co, Mn and Ni, which have an oxidant effect and accelerate oxidized reactions), lipoxygenases, porphyrins, ozone, free radicals or initiators in the form of physical agents such as light or temperature (at temperatures above the storage temperature. The higher the temperature, the faster oxidation will be initiated) [10,13,14].

During the propagation step, the removal of a hydrogen atom by the LOO∙ radical takes place. The reaction is actually a free and selective reaction that involves attaching hydrides with low binding energies, such as allyl-CH2- [10].

$$L·+O_{\mathbf{2}}\to LOO·$$

(3)

$$LOO·+LH\to LOOH+L·$$

(4)

$$LOOH\to LO·+·OH$$

(5)

This creates an opportunity to initiate alternative oxidation reactions by changing the composition of the oxidation-product mixture. Addition, cyclization, and elimination reactions compete with the attachment of hydrogen to the radical. There are many alternative mechanisms for the propagation step. LOO∙ enters into a hydrogen-transfer reaction in conditions with increased oxygen content [10,13].

$$LOO·+LH\to LOOH+L$$

(6)

Depending on the conditions under which the oxidation reaction is carried out, either regrouping or cyclization can occur [13,14].

(7)

In addition, it is possible to attach LOO∙ to double bonds [13].

$$·LOO·+R_{\mathbf{1}}-C{H}_{\mathbf{2}}-CH=CH-R_{\mathbf{2}}\to R_{\mathbf{1}}-C{H}_{\mathbf{2}}-\dot{CH}-CH\left(OOL\right)-R_{\mathbf{2}}·$$

(8)

The LOO∙ radical can also undergo disproportionation and β-splitting [13].

Propagation can also occur with the LO∙ radical, which initiates very rapid oxidation after the end of the induction period. LO∙ alkoxy radicals can be produced only from the decomposition of LOOH. They react much faster than LOO∙ radicals and become dominant in the oxidation. In the case of free-radical reactions with an alkoxy radical, hydrogen attachment is possible [13,14].

$$LO·+LOOH\to LOH+LOO··$$

(9)

Regrouping, or cyclization, is also possible [10].

(10)

In the case of cyclization, temperature has a negligible effect on the course of this reaction. It requires little energy for activation, so cyclization already dominates in pure esters at room temperature. As temperature increases, hydrogen transfer and cleavage become the dominant reactions [13]. Attachment to double bonds is another possible reaction that proceeds with alkoxyl radicals [13].

(11)

β–cleavage leads to the breaking of the C–C bond on both sides of the LO∙, with the formation of a mixture of carbonyl products and free radicals, aldehydes, and alkanes [10].

(12)

Cleavage is promoted by an increase in temperature, making it one of the dominant reactions of the propagation process [13]. Secondary stable oxidation products are then formed, as can be determined by the anisidine value.

The last stage of the free-radical oxidation reaction is termination, in which the peroxidation process nears its end. Free radicals form end products that do not contain free radicals in their composition.

$$L·+L·\to L-L$$

(13)

$$L·+LOO·\to LOOL$$

(14)

$$LOO·+LOO·\to LOOL+O_{\mathbf{2}}·$$

(15)

The termination process consists of four possible reaction pathways, one of which is radical recombination, which leads to the formation of products such as dimers, aldehydes, ketones, alcohols, ethers, and polymers [8,10,13].

$$R_{\mathbf{1}}O·+R_{\mathbf{2}}O· \to R_{\mathbf{1}}OO{R}_{\mathbf{2}}·$$

(16)

$$R_{\mathbf{1}}O·+ R_{\mathbf{2}}· \to R_{\mathbf{1}}O{R}_{\mathbf{2}}·$$

(17)

$$R_{\mathbf{1}}·+R_{\mathbf{2}}· \to R_{\mathbf{1}}{R}_{\mathbf{2}}·$$

(18)

(19)

During termination, β–cleavage is also possible. This reaction is the main source of aldehyde formation in the final products [13,14]. Another type of reaction that occurs is the oxidation of radicals with ester radicals and elimination reactions [8,13].

(20)

In the process of peroxidation, the phenomenon of reinitiation can occur. Reinitiation involves the decomposition of peroxides and leads to the reformation of free radicals. The reinitiation phenomenon is favored by the presence of transition-metal ions such as iron or copper ions [8].

$$LOOH+F{e}^{\mathbf{2}+}\to LO·+O{H}^{-}+F{e}^{\mathbf{3}+}$$

(21)

$$LOOH+F{e}^{\mathbf{3}+}\to LOO·+H^{+}+F{e}^{\mathbf{2}+}·$$

(22)

Subsequent transformations of the products of the peroxidation process, which occur as a result of β–elimination, lead to the breakdown of polyunsaturated fatty acid residues and the formation of aldehydes, hydroxy aldehydes, and hydrocarbons [8,12].

The consequences of aging of fatty acid methyl esters are the formation of deposits, which are deposited on engine components, thereby increasing emissions of harmful substances in the exhaust gas and contributing to increased fuel consumption during operation; increased viscosity and acid number; and corrosion and rusting of the surfaces of structural materials [15]. Alcohols in FAMEs, are products of hydrolysis, and their presence leads to a decrease in flash point [16].

Based on the studies of Sampaio et al. [17] and Chung [18], it can be observed that degradation of carotenoids and tocopherols occurs very rapidly when the temperature is high. This demonstrates the need to study the behavior of FAMEs from frying oils. When used for catering purposes, oils are heated to elevated temperatures, very often for a long time [17,18]. Based on the studies of Sabliov et al. [19] and Morales et al. [20], it can be concluded that temperatures above 60 °C should not be used to accelerate oxidation in studies aimed at oxidative stability. The hydroperoxides formed are easily broken down, causing changes in the oxidation mechanism. The type of natural antioxidants can affect the overall oxidation rate and direct lipid-oxidation pathways.

The research is novel due to the determination of the kinetics of adverse oxidation and deterioration of ester biofuels. According to standard methods, the aging of biofuels is studied using a method based on accelerated oxidation and by evaluating changes in the specific conductivity of the aqueous solution (Rancimat) or by a method involving changes in pressure, the PetroOxy method. The third method involves the measurement of the content of precipitates produced. These methods show changes in samples in a macroscopic way. Often, measurements taken in succession produce a result that is not statistically significantly different. Therefore, none of these methods can be used to observe substitutions of either primary or secondary oxidation products over time. The authors propose adapting methods used to test the quality of vegetable oils to study the kinetics of oxidative changes in ester biofuels such as FAMEs.

The purpose of this study was to investigate how elevated temperature affects the formation and distribution of oxidation products. Based on the results, the correlation between the content of primary and secondary oxidation products was determined, and a simple infrared spectroscopy technique was used as a method to observe the changes after oxidation. Observation based on the study of changes in anisidine value could allow the elimination of competing effects on the oxidation mechanism in the presence of natural antioxidants. The development of rapid methods to assess aging by oxidation is particularly important for ester fuels deprived of natural antioxidants as a result of previous thermal degradation of the transesterification feedstock. Previous research has mainly focused on the formation and measurement of peroxide content in oxidized FAME samples. Secondary stable oxidation products have not yet been analyzed.

2. Materials and Methods

The raw materials used in the study were FAMEs obtained from two different types of rapeseed oil. To obtain the research material, unrefined rapeseed oil and used cooking oil (UCO) (obtained from a local food service) were transesterified.

Transesterification of each oil was carried out in the presence of an alkaline catalyst (NaOH at a concentration of 1 mol/dm³ in relation to a given methanol) at the boiling point of methanol. The reaction mixture was stirred using a magnetic stirrer at 1000 rpm for 90 min in a glass laboratory reactor. The ester phase was separated from the resulting reaction mixture in a separatory funnel and neutralized to pH = 7. The FAMEs were then dried using a silica gel with a moisture indicator and grain size 2–4 mm delivered from PPH STANLAB. Silica gel was added until the color change disappeared [21]. The basic properties of the obtained FAMEs, such as density [22], kinematic viscosity [23], total methyl ester concentration [24], peroxide value (LOO) [25], anisidine value (LA) [26] and acid value (LK) [27], as well as color, were examined organoleptically before the start of the study.

The unrefined oil FAME samples were aged at 80, 100, 120, 140, and 160 °C for 4, 8, 12, 16, and 24 h, while FAME samples from used cooking oil (UCO) were aged at 120 and 140 °C, also for 4, 8, 12, 16, and 24 h. Aging was carried out in a glass beaker in a thermostat-equipped laboratory dryer without light.

The stability of the fractions before and after aging of the samples was evaluated by determining the content of LOO [25], LA [26] and LK [27], from which the total content of primary, secondary, and acidic oxidation products, respectively, were determined, and by subjecting the samples to infrared (IR) spectroscopy. The IR spectra were produced by a method using a flow cuvette made of KBr in the FT-IR spectrophotometer BRUKTER INVENIO S. All tests were carried out in triplicate, and the results are presented as an average.

3. Results and Discussion

3.1. Properties of the Tested FAMEs

Table 1. Comparison of the values of the properties of the FAMEs prepared from different vegetable oils.

Parameters	Norm Values28)	FAMEs from Unrefined Oil	FAMEs from Used Cooking Oil
Density in 15 °C, kg/m3	860–900	883 ± 3	886 ± 4
Kinematic viscosity in 40 °C, mm2/s	3.50–5.00	4.31 ± 0.11	5.32 ± 0.18
Total concentration of methyl esters, %	min. 96.5	95.7 ± 0.7	94.5 ± 1.2
Peroxide value, meq O2/kg	Non-normative parameter	9.85 ± 0.31	7.40 ± 0.24
Anisidine value, AnV	Non-normative parameter	3.183 ± 0.307	15.790 ± 0.531
Acid value, mg KOH/g	max. 0.	2.38 ± 0.21	22.98 ± 0.46
Color	Non-normative parameter	Orange	yellow

shows the initial physicochemical properties of FAMEs obtained from unrefined oil, which met the requirements of the standard for fatty acid methyl esters (FAME) for use in compression–ignition engines [28]. The requirements were not met by FAMEs obtained from used cooking oil, which is a direct result of the nature of the transesterified material. This FAMEs was characterized by increased kinematic viscosity and acid value and had a lower-than-acceptable content of fatty acid methyl esters.

3.2. Change in the Values of Acids, Peroxides and Anisidine in the Samples Tested

As shown in Figure 1a for FAMEs from unrefined rapeseed oil, it was observed that peroxide values increased over time at all temperatures analyzed. The greatest increase in peroxide value relative to the initial value was found in samples aged at 100 °C and 120 °C. This indicates the formation of a large number of peroxides as primary oxidation products in FAMEs. The lowest values were observed at 160 °C and were lower than the initial peroxide values. At this temperature, after about 8 h, a maximum can be observed on the graph of changes in the peroxide content of the samples; after that point, the graph shows a decrease in the content. This indicates the decomposition of these forms and their rapid transformation into secondary oxidation products. The decomposition of peroxides accelerated as the temperature increased above 120 °C.

In Figure 1b, which shows results for FAMEs from UCO, it can be observed that over time, the LOO values increased when they were stored at 120 °C and were almost constant when they were stored at 140 °C. The largest increase in LOO with respect to the initial value is characteristic of samples stored at 120 °C. For samples stored at 140 °C, LOO values reached a small maximum after 12 h. The nature of the changes is due to the origin of the rapeseed oil used for transesterification.

Despite the samples having been heated at such high temperatures, the changes are rectilinear, with high regression coefficients (R² above 0.8).

Table 2. Analysis of kinetic equations based on changes in peroxide value.

Temperature, °C	Zero-Order Kinetic Equation	R2	First-Order Kinetic Equation	R2
FAMEs from unrefined rapeseed oil
80	c = 0.8489t + 8.2459	0.9449	log(c) = 0.0255t + 0.9279	0.9276
100	c = 1.6972t + 11.216	0.9737	log(c) = 0.0241t + 1.2891	0.9117
120	c = 1.6656t + 13.824	0.9555	log(c) = 0.018t + 1.3835	0.9639
140	c = 0.5392t + 12.741	0.8099	log(c) = 0.0071t + 1.2378	0.906
160	c = −0.1039t + 11.315	0.0898	log(c) = −0.0095t + 1.0777	0.2514
FAMEs from rapeseed oil used for cooking
120	c = 0.3293t + 8.5391	0.5331	log(c) = 0.0096t + 0.9793	0.3106
140	c = −0.0238t + 7.4333	0.1271	log(c) = −0.0016t + 0.8729	0.0960

shows a comparison of kinetic-change equations according to zero- and first-order reactions. Based on the analysis performed, it could be observed that the formation of primary oxidation products (mainly peroxides) was associated with changes in reaction order with increasing temperature. In the range from 80 to 120 °C, reactions proceeded for which the best fit was observed for the relationship c = f(t), indicating zero-order kinetics. At temperatures above 120 °C, the best fit was observed for the relationship log (c) = f(t), indicating a first-order reaction. The reaction proceeding at 120 °C was mixed in nature, and the obtained simple coefficients were on the same level. In the case of changes in FAMEs derived from rapeseed oil used for cooking, different types of changes that were directly influenced by the origin of the samples were observed. Post-frying oil is formed by heat treatment of rapeseed oil, which causes structural changes. These changes result in easier transformation into secondary oxidation products.

An examination of the change in kinetics based on the study of the content of primary oxidation products in FAMEs from frying oil used the fit analysis for zero-order and first-order kinetics shown in Table 2. When higher-order kinetics were analyzed for this test material, worse fitting coefficients were obtained. Fitting according to the zero-order equation yielded the best fit. This indicates the complex nature of the reactions involving primary oxidation products in FAMEs from frying oil. Peroxides, as primary oxidation products, are converted very quickly, and their concentration was difficult to measure compared to that of FAMEs from unrefined rapeseed oil. In FAMEs from frying oil, the double bonds showed a higher vibrational frequency, which was observed from the broadening on the IR spectra presented in the Supplementary Materials. The faster decay could be confirmed by comparing Figure 2a to Figure 2b for the FAMEs tested at the same temperatures. At 120 °C, a maximum of about 15 meq O₂/kg was obtained for FAMEs from frying oil, while about 45 meq O₂/kg was obtained for FAMEs from unrefined rapeseed oil. The concentration of primary oxidation products in the former was almost three times lower after a 20 h exposure at the same temperature.

Figure 2a shows the changes in anisidine value for FAMEs from unrefined rapeseed oil, while Figure 2b shows these changes in post-frying oil. In both cases, it was observed that there was an increase in the anisidine value over the entire range of temperatures tested. The greatest increase in anisidine value compared to the initial value was seen in samples stored at 140 °C and 160 °C. This indicates the formation of a large amount of secondary oxidation products in the tested FAMEs with the passage of time from the beginning of the test. The lowest anisidine values were observed at temperatures of 80 °C, 100 °C and 120 °C for FAMEs obtained from unrefined rapeseed oil. Based on this study, a direct relationship was observed between peroxide value and anisidine value. The highest anisidine value occured at 160 °C, while the peroxide value was lowest at this temperature. While the anisidine value was lowest at 100 °C and 120 °C, for peroxide value, the value is highest at these temperatures. This indicates the rate of transformation of primary oxidation products into secondary, stable oxidation products. The anisidine value is the parameter that best reflects the effect of temperature as a factor in accelerating the aging of ester biofuels, even when using higher temperatures, i.e., temperatures above 80 °C. In the case of FAMEs from unrefined rapeseed oil, natural inhibition of the formation of secondary oxidation products has been observed. This can result in a change in the color of FAMEs that indicates the presence of carotenoid pigments (mainly beta-carotene), which are natural antioxidants. When the limiting temperature of 120 °C was exceeded, discoloration of the samples occurred, which meant complete conversion of carotenoids into oxidized forms. Above this temperature, there was a rapid increase in the anisidine value.

As shown in Figure 2b for FAMEs from UCO, it can be observed that the anisidine values increase over time at both temperatures tested. FAME samples from this type of oil were yellow, indicating a direct dependence on natural antioxidants. The largest increase in LA with respect to the initial value was seen in samples stored at 120 °C. The different nature of the changes is directly attributable to the initial feedstock used for transesterification.

Based on the analysis of the kinetics of changes in the formation of secondary oxidation products shown in

Table 3. Analysis of kinetic equations based on changes in anisidine value.

Temperature, °C	Zero-Order Kinetic Equation	R2	First-Order Kinetic Equation	R2
FAMEs from unrefined rapeseed oil
80	c = 0.5167t + 2.1449	0.9114	log(c) = 0.0353t + 0.4401	0.9328
100	c = 0.462t + 1.6607	0.9000	log(c) = 0.0471t + 0.1982	0.8975
120	c = 0.6974t + 1.2477	0.9418	log(c) = 0.0511t + 0.2652	0.8975
140	c = 5.0258t + 19.949	0.8754	log(c) = 0.023t + 1.6249	0.7749
160	c = 17.403t + 53.317	0.8991	log(c) = 0.0252t + 2.1071	0.7971
FAMEs from rapeseed oil used for cooking
120	c = 6.7035t + 19.887	0.9676	log(c) = 0.0455t + 1.3857	0.8630
140	c = 4.4838t + 23.747	0.9523	log(c) = 0.0379t + 1.3851	0.8214

, it can be observed that when the equation c = f(t) was used, higher coefficients of straight-line fit were obtained, while the straight-line equations for the relationship log(c) = f(t) were characterized by lower correlation coefficients (the inverse relationship was observed only at 80 °C). For FAMEs obtained from unrefined rapeseed oil, the rate of change was higher with increasing temperature, rising from about 0.5 to 17.4. At 120 °C, the reaction rate for the formation of secondary oxidation products in FAMEs from used cooking oil was almost 10 times higher than that in FAMEs from unrefined oil. The rate of this reaction for FAMEs from used cooking oil decreased with increasing temperature, which is due to the type of feedstock used for transesterification. The used cooking oil had been subjected to previous thermal treatment when it was used in catering. In tests after transesterification, the resulting product, when subjected to successive temperature treatments, showed low resistance to oxidation. Thus, other quality parameters will also deteriorate very quickly, despite meeting standards immediately after production.

Figure 3 shows the changes in acid value for both FAMEs tested. For FAMEs from unrefined rapeseed oil, it was observed that the acid values increased with time for each of the temperatures used. The greatest increase in this parameter, with respect to the initial value, was characteristic of samples aged at 140 °C and 160 °C. This indicates the formation of a high number of carboxylic acids in FAMEs. The smallest changes in acid value were observed at a temperature of 100 °C. The acid values obtained for these samples did not show statistically significant changes, which means that these changes cannot be used to elucidate changes in the behavior of the samples tested. In the case of FAMEs from UCO, it was observed that acid values decreased with time. The decreasing acid values relative to the initial value indicate a difference in the chemistry of the effect of temperature on FAME samples from unrefined rapeseed oil compared to that on FAME samples from UCO. Therefore, it becomes reasonable to question the utility of the acid value as an indicator of FAME aging. The carboxylic acids formed may react with alcohols or hydroxy aldehydes formed during oxidation. The different course of fatty acid formation in FAMEs is indicated by the low fitting factor found when zero-order reaction fitting was used, which may indicate parallel esterification reactions.

Based on a study of the numbers characterizing various oxidation products, a relationship was observed between the acid value and the peroxide value. The highest peroxide value occurred at 100 °C, while the acid value was the lowest at this temperature. The largest acid value and the smallest peroxide value occurred at 160 °C, indicating the transition of peroxides as perishable forms of organic compounds into stable forms, including carboxylic acids.

The relationship between the acid and anisidine values was present for both FAMEs used in the study. The acid values decreased as the anisidine values increased, which may indicate additional recombination of the oxidation products or their over-reactivity, as indicated by Reactions (5) and (9), respectively. In Reaction (9), alcohols are formed from fatty acids and consequently may additionally undergo esterification reactions with the resulting fatty acids.

3.3. Application of IR Analysis in the Study of FAME Aging under the Influence of Temperature

Samples for infrared spectrum studies were selected based on analysis of acid, peroxide and anisidine values. Samples heated for 12 h and 20 h at all temperatures were analyzed. All IR spectra were included as Supporting information for publication. Particular changes can be observed in Figures S1b and S2b, that is, in the band characteristic of O-H bonds (3000–3500 cm⁻¹)). In Figures S1c and S2c, changes can be observed in the bands characteristic of C=O bonds (1645–1850 cm⁻¹) and C=C bonds (1580–1700 cm⁻¹). In Figures S1d and S2d, changes can be observed in the bands characteristic of methylene groups (800–1000 cm⁻¹) and C-O bonds (1000–1260 cm⁻¹). It can be observed that with increasing temperature, there was a decrease in the transmittance value. In Figures S1d and S2d, the overlap of IR spectra in this range of wave numbers can be observed, indicating the lack of effect of elevated temperatures on hydrocarbon groups. In the range of wavenumbers characteristic of the ester group, especially at lower wavenumbers, a broadening of the spectral lines for the characteristic groups can be observed. This indicates the effect of temperature on changes in the vibration range of these groups. This is particularly marked for samples obtained after 20 h of exposure to elevated temperatures. The broadening of the peaks increases with increasing temperature, indicating changes in the arrangement of ester bonds caused by prolonged exposure to elevated temperatures. This confirms the conclusions obtained from the peroxide value and anisidine value, which changed with increasing temperature. IR spectra confirmed the presence of higher vibrations in the structures, which promotes the occurrence of oxidation reactions.

Analysing the IR spectra obtained for FAMEs from unrefined rapeseed oil after 12 h and 20 h of exposure to elevated temperatures, it was observed that the range of wave numbers 3200–4000 cm⁻¹ is particularly important. In this range, especially at around 3300 cm⁻¹ and 3550 cm⁻¹, clear differences in transmittance are noticeable. The most intense changes were observed for samples after 20 h at elevated temperatures. These changes may indicate the presence of secondary oxidation products. For FAME samples that were not subjected to elevated temperatures, no changes were observed. The intensity of the changes increased with an increase in the temperature to which the sample was exposed. An analogous mechanism of temperature-related effects was observed for the anisidine value. Accordingly, the following section compares the transmittance changes in this range with changes in anisidine and peroxide value.

Figure S3 shows IR spectra for FAME samples obtained from UCO at 120 °C and 140 °C after 12 and 20 h. Particular changes can be observed in Figure S3b in the range of wave numbers between 3200–4000 cm⁻¹ and in the band characteristic of O-H bonds (3000–3500 cm⁻¹). In the case of FAMEs from UCO, it can be observed that after 12 h at 120 °C, there was already a significant reduction in transmittance. This reduction proceeded with increasing time and increasing temperature. In Figure S3c, changes can be observed in the band characteristic of C=O bonds (1645–1850 cm⁻¹), where the change also, analogously, occurs in the range of lower wave numbers. This also indicates changes in the range of ester groups. Figure S3d shows changes in the band characteristic of methylene groups (800–1000 cm⁻¹) and C-O bonds (1000–1260 cm⁻¹). It can be observed that with increasing temperature, there was a decrease in the transmittance values, but only in the range of C-O bonds. This indicates a different effect of temperature on FAMEs obtained from unrefined rapeseed oil versus used cooking oil.

Based on the analysis of IR spectra, structural changes could be observed with an increase in the temperature acting on the types of FAME studied, as well as the effect of the time of its action. In the following part of the work, the authors attempt to determine whether the changes observed in the IR spectra could depend on changes in the studied peroxide and anisidine values, which determine changes in the content of oxidation products. For this purpose, wave numbers of 1049 cm⁻¹, 3300 cm⁻¹, and 3545 cm⁻¹ were selected in the range in which the greatest changes were observed. The relationships are shown in Figure 4 and Figure 5 for the changes in the peroxide and anisidine value, respectively, depending on the transmittance at the indicated wavelengths. Based on the fit coefficients (R²), it can be observed that the IR spectra changed with the change in anisidine value, particularly at wavenumbers of 1049 cm⁻¹, and 3300 cm⁻¹. Changes in this range confirm the formation of secondary oxidation products. At a wavenumber of 3545 cm⁻¹, the lack of linearity confirms the change in the content of hydroxyl groups, which can react with the formed carboxylic acids, directly affecting the impossibility of unambiguously determining the thermo-oxidation of the ester biofuel on the basis of changes in the acid value. Transmittance changes at selected wave numbers show a low dependence on the primary oxidation products. This indicates the possibility of observing changes in the presence of secondary oxidation products by IR spectrometry.

4. Conclusions

On the basis of the study, it was determined that as the temperature increases, the aging process of fatty acid methyl esters is accelerated. The increase in temperature initiates a series of reactions that lead to the formation of carboxylic acids and secondary oxidation products. On analysis of the results for FAMEs from unrefined rapeseed oil, an increase in acid and anisidine value was observed, aling with a decrease in the peroxide value. The decrease in LOO is due to the transition of peroxide forms into carboxylic acids and secondary oxidation products. On analysis of the results for FAMEs from UCO, a decrease in acid value and peroxide value was observed, while anisidine values increased. The difference in the behaviors of acid and peroxide values may be due to the fact that the initial content of carboxylic acids was much higher in FAMEs from used cooking oil compared to FAMEs from unrefined rapeseed oil. In FAMEs from UCO, the content of carboxylic acids was high from the beginning, so they were undergoing aging, while in FAMEs from unrefined rapeseed oil, these acids were just beginning to form. These changes were confirmed by IR spectra, in which changes could be observed in the wave number range 3200–4000 cm⁻¹, which corresponds to the vibration band of O-H bonds. Based on the comparison of changes in anisidine values at selected wavelengths, it can be concluded that the aging of ester biofuels can be monitored by IR spectrometry. Changes in characteristic vibration bands of functional groups indicates progressive oxidation processes.

The paper confirms the impossibility of observing changes in ester biofuels based on the study of peroxide forms of the resulting products at elevated temperatures. When conducting accelerated studies at elevated temperatures, it is necessary to analyze changes in secondary rather than primary oxidation products, which, despite the presence of a different mechanism, are an indicator of biofuel aging. These changes can be observed using IR spectrometry.

The presence of natural antioxidants in the form of carotenoid pigments and tocopherols may be a particularly important aspect based on the research performed. In view of the above, further research is planned to determine the effect of natural antioxidants on the rate of oxidation of FAMEs obtained from high-temperature-treated catering frying oil. Currently used antioxidant additives are artificially produced chemical compounds. During storage, it would be especially important to avoid overheating of ester biofuel tanks, especially in the context of the use of this type of biofuel by farmers who can produce FAMEs from oil crops such as rapeseed on their own [29,30]. The biofuel that they store in mauser-type containers, especially during the summer, can be exposed to elevated temperatures and light [29]. Accelerated oxidation would occur regardless of the type of oil farmers use to produce FAMEs.